site banner
PaperclipPerfector  1mo ago (text post)   23552 thread views

Culture War Roundup for the week of June 8, 2026

This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.

Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.

We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:

  • Shaming.

  • Attempting to 'build consensus' or enforce ideological conformity.

  • Making sweeping generalizations to vilify a group you dislike.

  • Recruiting for a cause.

  • Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.

In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:

  • Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.

  • Be as precise and charitable as you can. Don't paraphrase unflatteringly.

  • Don't imply that someone said something they did not say, even if you think it follows from what they said.

  • Write like everyone is reading and you want them to be included in the discussion.

On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.

6
Top Bottom Old Controversial Comments
Jump in the discussion.

No email address required.

Does anyone else feel like we're heading for a good old-fashioned, 2008-tier, financial crash? I recall people bringing up the possibility ever since Covid bucks started rolling out, but even though we were due for one, and even though the money printer was going brrrrr, the crash has so far failed to materialize.

At the time, I was of two minds about this. On one hand, all the libertarian theory I used to subscribe to said money-printing => boom, boom => bust. On the other hand, the problem for me was it never felt like a boom, and I think this is changing now. A key feature of the pre-crash boom is "malinvestment"; capital going into often downright deranged projects, that are later abandoned half-finished. Well, I feel like the datacenter craze qualifies, and with the wave of AI IPOs that are coming just as major indices are changing their rules, to allow for these companies' near-instantaneous inclusion (an investment so good, you can't pass up on it. Literally, if you're American), it seems like we're solidly in the "irrational exuberance" phase.

Add this to the list of things I hope I'm wrong about, because if we get a proper crash, the political fallout is going to be massive. The script writes itself: Trump / tech bros / capitalism bad, even more gay race communism now.

To whitepillers: is there an argument for why I'm wrong that doesn't boil down to "you don't get it, chud, it's the New Economy! The Singularity is just around the corner! All the rules are obsolete!"? This argument is verboten, because this is pretty much what people say with every bubble.

To fellow blackpillers: any ideas on how to brace for impact? Any IT guys here old enough to make it through the dotcom bubble? How did you do it? Any advice you would have given your past self?

No, I don't really think we are. It's definitely possible we could get a 2001 style sector driven crash if AI growth slows, but I don't think there's any structural leverage for a 2008 style devaluation. The two inflated assets are 1) The US stock market. and 2) The US dollar. Yeah, nobody in 08 thought the housing sector could crash, but I don't see any scenario where the dollar falls more than 20% right about now, and the US stock market's insane valuations are driven by a number of companies with genuine chances of reshaping much of the world economy. There's still plenty of cash on the sidelines willing to move in if valuations drop.

If you wanna blackpill, my bigger concerns would be structural unemployment for the bottom 60% of people in the coming decades, and inflation hollowing out middle class wages following increasing salary pressure from the capital and well earning professional classes. I don't think the bubble explodes. I think it keeps growing and largely pulls money upward out of wage earner hands.

Anyway, if you wanna be prepared, the advice is basically the same as it's always been. Have a big pile of cash, your housing situation locked in, and be able to weather a sudden job loss (Back up plan, second household income, or savings).

Does anyone else feel like we're heading for a good old-fashioned, 2008-tier, financial crash?

We've been heading towards a crash since 2008 because we never really got over it. We've been kicking the can down the road just delaying the inevitable.

I recall people bringing up the possibility ever since Covid bucks started rolling out, but even though we were due for one, and even though the money printer was going brrrrr, the crash has so far failed to materialize.

It didn't materialize because policymakers did everything in their power to delay it, but it's just a bet against the future. It's like saying John managed to avoid bankruptcy by getting another credit card. No, he is still bankrupt, it's just not official yet.

On the other hand, the problem for me was it never felt like a boom, and I think this is changing now.

There was most definitely a boom since the 90s up until 2008. It was palpable.

To fellow blackpillers: any ideas on how to brace for impact? Any IT guys here old enough to make it through the dotcom bubble? How did you do it? Any advice you would have given your past self?

I'm an old IT guy that went through the dotcom bubble. I never really bought into it. It was obvious to me that no revenue means no revenue. The difference now is that somehow people know it's a bubble, but there's money to be made if you guess correctly when it's going to pop.

I'm not interested in that kind of speculation.

Either way, I'm beyond blackpilled. I don't think the inevitable upcoming crash is going to be "good old-fashioned 2008-tier", I don't even think The Great Depression would make it justice. I think this is going to be the mother of all crashes and the entire capitalistic system is going to collapse. The younger adult generation is already antagonistic towards capitalism, they see they can't afford a house, while BlackRock owns obscene amounts of real estate. When the system collapses capitalists will have no leg to stand on, arguably they already don't.

But ultimately I see it as a good thing. There's only so much time you can pretend the emperor has no clothes. At some point civilization has to face reality, and that point is coming... fast.

Does anyone else feel like we're heading for a good old-fashioned, 2008-tier, financial crash? I recall people bringing up the possibility ever since Covid bucks started rolling out, but even though we were due for one, and even though the money printer was going brrrrr, the crash has so far failed to materialize.

Of that nature? Not really, since the financial crisis was mechanically distinct. A big economic doldrums? Sure, but then I would have thought most people had that priced in even before the Trump tariffs last year, let alone the Strait of Hormuz closure this year. What did people think the western demographic trends implied for global economics?

We're in the start of the broader first world demographic retirement wave, where many Western (and 'western' Asian) countries aren't just older on average, but the generation than makes it older is starting to retire and leave the work force. For example, here was the demographic pyramid of Germany in 2023.. The retirement age in Germany is 67. That disproportionately wide bar that was 60-64 three years ago is now 63-67 now. They are starting to transition out of the work force, taking their accumulated expertise and economic inputs with them and turning into retirees instead who draw more from the public purse. That next ten years will be even larger cohorts of the demographic-industrial base. Germany is not unique.

The implications of demographic age-out have never been optimistic. Fewer workers contributing to taxable economic growth, higher expenditures for the welfare state dominating public spending that could otherwise go to economic investments, loss a large amounts of highly-trained and experienced knowledge to be replaced by less experienced younger demographics who have often suffered from youth unemployment dynamics due to worker protections benefiting the now-retiring or policies to import specialists from abroad. I wouldn't go so far to say inherently recessionary, but if you remove part of the foundations of your economic growth, you are going to have less economic growth, which was the longer-term impact of the great recession in many places.

And this was before various macroeconomic disruptions and trade flow changes that have ended the neoliberal consensus. COVID 19 alone was enough that many countries were already relooking supply chains that had prioritized cost-efficiency over reliability. Well, that change was and is going to lead to slower economic growth. The Trump tariffs were part of an effort to fundamentally reshape the global economic system by reducing the role of the US as consumer of last resort for other country's exports, and it is still shaking out as the European Union gears up for a trade war with China as it doesn't want to absorb all the redirected Chinese exports that the US used to import. The US-Iran War has led to the eyebrow-raising twin blockades of the US blocking Iranian exports that mostly went to China, and Iran blockading everything else which mostly goes to also China (and also other asian countries). The consequence is global energy prices raising for everyone, but not raising the same for everyone, with the biggest cost-increases headed for the energy-import-dependent manufacturing-export countries, epecially when national stockpiles currently being released dry up.

And even this is before other, easily foreseeable, major macro-economic disruptions that could happen in the future. The Russia-Ukraine War has progressed beyond pushing Russian oil onto the black market as expanding Ukrainian long-range strike capabilities decrease what Russia can push onto the sea at all. A PRC invasion of Taiwan is liable to destroy the high-end chip economy for a decade, and that's before the PRC and US target eachother. The European Union could have significant structural crisis in the next few years if Euroskeptic parties on the rise in France and Germany take over in one or both. Expanded naval trade piracy / 'taxation' at various Eurasian maritime chokepoints, mass migration destabilizations... heck, even (or especially) AI, which may not be a net-negative, is going to lead a lot of economic disruption and/or ruin in its wake.

The global economy was already going to have to undergo uncomfortable economic dislocations of where growth occurred just from demographic age-out. It just so happened to be occuring concurrently with the transition away from the post-cold-war global neoliberal order, which prioritized cost-minimization over other considerations. The de-prioritization of cost-minimization was always going to further complicate global finances. There just so happen to be another half dozen economic asteroid events approaching as well, for various reasons.

The consequence is global energy prices raising for everyone, but not raising the same for everyone,

I guess I don't have a good view for what everyone else is seeing, but at least here (unspecified southern US), wholesale electricity prices today seem to be within 10% or so of the averages for 2024 and 2025. Gas prices rose probably $2.00 per gallon after the start of the Iran conflict, but checking today are only up about half that (probably a bit more if you average the last month), and have stayed lower than 2022 highs the whole time. (Not expressing an opinion on that, er, adventure here, just observing).

But the utility company does expect total grid energy usage to double within 5 years or so, mostly on the basis of unspecified "data center" usage. I have trouble believing all of that is actually going to manifest, but it's probably better to be proactively investing in infrastructure than the reverse.

We had a bit of a bust in Silicon Valley with the end of ZIRP.

Also while I'm biased from being in the industry, I don't assign high probability to the "datacenter craze" being some bubble phenomena. Companies including my employer are snapping up all the GPUs they can not because they're pretty tulips, but because we can make productive use of them. We don't have much in the way of idle GPUs sitting around.

Late to the party, but wanted to share a relevant essay written by Howard Marks, a well respected voice in the financial industry (cited by buffett, gurley, etc.) :

Is It a Bubble?

It was published in December, so some specifics are outdated (anthropic/openai valuations, etc.), but it's a comprehensive & thoughtful analysis.

It doesn't answer all the specific questions you asked, but you might find it useful & interesting. He's been writing for decades, so if you enjoy this piece, there's a deep archive to draw from.

In general it is impossible to predict when a crash will happen until it has already happened. So I am a greypiller. You should always try to be prepared for the possibility of a downturn (whether in the larger economy or in your own personal economic fortunes) but you're fooling yourself if you think you can predict them.

In general it is impossible to predict when a crash will happen until it has already happened.

I agree, and there's also a certain irony here. With most market trends, the first people to anticipate the trend end up doing great. Not so with market crashes. The first people to spot them (and act) frequently get screwed.

Yeah for every Big Short that works out there'll be a plethora of attempts that end up getting absolutely wrecked (even if they're directional correct) by funding costs and market realities.

The classic "the market can remain irrational longer than you can remain solvent." Never short!

Traditionally, the way to handle this was to reallocate to stable assets, where "stable asset" meant gold and treasury bonds.

With the crash we're headed for, I think canned fruits and bullets are the stable assets of choice.

Not shells?

Yep. Preparing for a crash with a crystal ball is very expensive, opportunity cost wise. I.e. the market was clearly heated in 1999, but you'd have missed some of the best months in history if you left early.

Unfortunately no. We are headed towards a depression. AI is essentially deflationary. What it is best at is reducing demand and destroying added value.

Think of the following scenario - lets just say that 3d printing advances a bit more, as is home cnc, sheet metal bending and laser cutting. And add a bipedal robot. Suddenly all you have to buy from there on is basic commodities and semiconductors.

Actually, no! I really like how this article, a book review, summarizes it:

[Goodspeed] wants to see if the data supports theoretical accounts of recessions that locate the cause of the downturn in the preceding expansion. His antagonists are two celebrated Austrians: first Friedrich Hayek… whose prominent account of business cycles rivaled Keynes’s in the 1930s; and second, Joseph Schumpeter, who saw in recessions the opportunity to sort out the mistakes and misallocations of the boom. Goodspeed’s main goal in the 200 or so pages of the book is to ask whether the data supports these theories, or their modern variants, or whether it is consistent with a much simpler story. The short answer is no.

…Goodspeed shows that British and American expansions do not resemble Dorian Gray, looking beautiful but hiding an inevitable accumulation of malinvestments (objectively bad investments that are destined to fail) and distorted decisions (mistaken economic decisions taken on the basis of bad regulation or flawed prices) that make a correction inevitable. If they did so, he argues, one would expect that as expansions get longer they get more and more likely to end. In his data, however, the relationship between the age of an expansion and the probability of death is essentially zero. Nor do measures of increased investment during the boom correlate with the severity of a downturn. Nor do longer expansions have longer recessions after them.

This is why recessions remain essentially unpredictable. Any perceived regularity is likely to be a statistical illusion.

On top of all this, he finds no evidence that recessions are corrective. Reallocations tend to happen more aggressively during expansions not contractions, contrary to the arguments that Joseph Schumpeter famously made. Similarly, he finds that contrary to common belief recessions tend not to be contagious but rather are mostly patriotic, usually confined to a single country like the modest 2001 downturn in the US (which did not spread to the UK).

I also find myself in strong agreement with the reviewer:

Personally, Recession strengthened my prior beliefs that policymakers simply don’t have enough information to even distinguish between a robust expansion and a speculative bubble in real time, let alone the tools to safely tame any bubbles that they did find. Rule-based monetary policy, which allows market participants to form stable expectations about what its response will be, while still allowing flexibility in the event of shocks, might be the best we can hope for. Here Goodspeed’s advice is sensible: policymakers should first do no harm before thinking that they have the ability to entirely tame the business cycle.

In short: crashes happen, but attempting to pop them early can only do more harm than good.

Goodspeed shows that British and American expansions do not resemble Dorian Gray, looking beautiful but hiding an inevitable accumulation of malinvestments (objectively bad investments that are destined to fail) and distorted decisions (mistaken economic decisions taken on the basis of bad regulation or flawed prices) that make a correction inevitable. If they did so, he argues, one would expect that as expansions get longer they get more and more likely to end

It sounds like he didn't understand the argument he's disputing. Malinvestments aren't necessarily "objectively bad investments that are destined to fail", their failure and "badness" stems from whether the current interest matches the market time preference, so I don't see how "one would expect that as expansions get longer they get more and more likely to end". I heard about Schumpeter, but I don't recall anything specific he said, so I don't know if he was portrayed accurately.

In short: crashes happen, but attempting to pop them early can only do more harm than good.

Maybe. I can easily imagine people doing even more harm, but I don't see what it has to do with what I asked. My question was about what I, personally, can do to brace for impact, not for policy recommendations.

The 2008 crash resulted in an excess demand for money. That's what catalyzes a crash in a particular market of the economy into a crash throughout the whole economy. The mechanism was that the mortgage backed securities were being used like money substitutes, so when they lost value there was a widespread shift into holding more money itself. When not matched by an increase in supply, this causes a systemic failure. Money is traded on every market, so a disequilibrium in the money itself screws up everything.

If there is a data center crash that means people suddenly stop spending money on data centers and their inputs. That is bad for that particular market and the people who depend on that (like me), but for the economy as a whole the relevant issue is what happens instead. If people shift that spending into different markets, then there will not be generalized crash like 2008. If people simply increase their money holdings instead, and the Fed does not increase the supply to meet demand, then there might be a similar crash. However, the mechanism here will be less direct and seems less likely to me though not impossible.

2008 seems entirely the wrong reference point, the dot com boom would seem a much better analogue. There is clearly a lot of potential for value and the market may get ahead of itself trying to figure out how and where it will be and over invest in infrastructure from FOMO. But the general thesis seems pretty strong (like dot com boom having the internet taking over everything, they were not wrong! yes pets.com went bust, but everyone gets their dog food from chewy now, orders from Amazon watches youtube and door dashes their food, can you even book travel by phone anymore?) and very likely 5 years later it would all be bigger then the peak again. Maybe we'll end up with 'dark' data centers like we had dark fiber for a while, but I would bet not for that long.

The various tech giants are so in the long on this shit I just can't wait for it to crash. Gleefully, if possible even government action to slash and burn the whole field. Maybe we will luck out and some vibecoded bullshit project will cause an irl disaster leading to an explosion or other shit hits the fan disaster. Prompting a kneejerk legislation from the usual crowd trying to be seen as "doing something", and maybe that would pour some cold water on the whole thing.

So a few things are hitting at once.

US energy production has been basically flat for the past 25 years. Renewables have been replacing coal with the total largely unchanged. Progressives, thanks to funding from China, adopted anti growth strategies and blocked new power plants. Now there's suddenly huge strategic demand for power hungry data centres. They need to find places with enough surplus grid capacity to build them. China has switched to PR campaigns about water usage.

The typical small town mayor will eagerly sell out the towns water and power rights to get his daughter a Google marketing internship. The small town residents are well aware of this. Additionally data centres don't generate many local jobs, so most towns don't see the point.

We're in an era of high interest rates, so companies that need to raise big money to build big computer now need big IPO.

I haven't seen any actual craziness yet. The original dot com bust had things getting just silly... loose money on each sale and make up the difference in ads! SF had a proto door dash that would have someone deliver a single candy bar for no mark up. The 2008 real estate bubble was famous for it's NINJA loans (no income, no job, no assets).

In comparison data centres are genuinely useful and they should be able to pay for themselves with other tasks if AI doesn't pan out like they expect.

I'm not sure how wasteful the current model training is... eg how much each new version carries over from the prior training. It might be possible that they are currently doing too many releases and it'd be more long term efficient to do fewer. But it's R&D, so who knows.

What people tend to miss is that all of the money is being spent on the AI training arms race. It's simple to run a model profitably after it has been produced. The issue is recovering cost of development and funding new development.

So even if all of the AI companies go bust, we're likely going to still be running models in those data centres.

It doesn't exactly feel like a bubble to me because the average Joe isn't getting rich off of it yet. There are no FOMO investors getting into it because their neighbour made a ton of money. That may change with the IPOs.

Progressives, thanks to funding from China, adopted anti growth strategies and blocked new power plants.

That's a pretty big claim. Do you have any references for that so that I can read up on it?

The one I've most seen pointed out is that Code Pink has both been investigated for its ties to China and the CCP, and also opposed construction of data centers on the grounds of "imperialism". That one doesn't even seem particularly subtle.

I'd be surprised if China wasn't doing some level of meddling but trying to prove/demonstrate that it's actually a significant factor feels like a far longer bow to draw.

Reading the article, the main source for this is OpenAI, who is not an uninterested party.

Remember the furor about Russian election interference on social Media and it ended up being a few thousand dollars worth of ads for a Sexy Bernie coloring book?

This is a few thousand dollars worth of OpenAI tokens to create ads with pictures of farms and forests and lakes with text which says "No data center is worth losing a single inch of this" or similar.

Yes. Russians have used their absolute mastery of digital marketing in order to take over every single electoral project in the West for the last 15 years, according to some people.

More recent than Dradis's claim, but OAI claims to have discovered a similar influence campaign from China yesterday:

https://www.politico.com/news/2026/06/10/openai-china-ai-data-centers-report-00957612

Thanks!

US energy production has been basically flat for the past 25 years. Renewables have been replacing coal with the total largely unchanged.

This is true, but I think you put the cart before the horse: US electricity demand has been flat between 2005 and 2020. Renewables have been replacing aging coal plants that were not economic to run under those conditions - the most expensive generators went off-line first, starting with aging coal plants that would have needed major CAPEX.

Progressives, thanks to funding from China, adopted anti growth strategies and blocked new power plants.

This might have resulted in more environmental protection, making coal even less competitive than it is anyway. But make no mistake, coal would not have been competitive with natural gas plants either way. It's all around worse, not only because the tree hugging hippies getting their way politically. Coal is more expensive to produce (especially since the shale revolution), more expensive to transport, more expensive pre-treatement, the plants are more expensive to staff, the plants are significantly less thermally efficient and the plants are much slower to ramp. And, yes, pollution control is significantly more expensive for coal.

There's a reason why most data centers try everything to get their hands on a gas turbine or to even get a nuke back online, and nobody even thinks about building a mine mouth data center.

People could have built more aluminum/silicone smelters or arc furnaces in the US in 2005, paid market rates for electricity and keep those coal plants going for a little bit longer (until the next gas turbine - or later, solar panel - would have put them out of business anyway). But they built those in China, because steel workers and coal miners and power plant personell are much cheaper there. And yes, pollution control is also cheaper there.

It doesn't exactly feel like a bubble to me because the average Joe isn't getting rich off of it yet. There are no FOMO investors getting into it because their neighbour made a ton of money. That may change with the IPOs.

That's just the thing that pushed me over the ledge and made me post. It's not that it may change, with the IPO, it will. It's mandated by law. Major indices are changing their rules so that these new stocks are included nearly immediately, so anyone doing passive investment - your retirement fund likely included - will be buying them.

Major indices are changing their rules so that these new stocks are included nearly immediately

Dow Jones (for the S&P 500) actually decided against this just this week. Some of the other indexes, like the Nasdaq 100, did change, but it seems unlikely most new AI companies will be able to get in anyway; they're doing it for SpaceX.

they're doing it for SpaceX.

Interesting. I suppose that makes me more whitepilled on the economy, but even more bearish on Elon. Originally I thought this was a SpaceX only thing, and a sign of desperation, but than I heard about the other AI IPOs and thought Elon is just riding a bigger trend. Maybe my first instinct was right.

Yes, but I think the concern over this is way too overblown. Average indices buyer will have an exposure of something like 0.35% to SpaceX. There's plenty of dogshit companies in the indices. If that's your main signal for predicting imminent financial crash, I think you're wrong.

Booms and busts will always keep coming, because of human nature. So what if it does? What's the problem, specifically? How would a crash affect you in particular?

How would a crash affect you in particular?

I have a job. Based on my experience of 2001 and 2008, there is a very good chance that a crash would cause me to not have a job.

I rather enjoy not freezing to death in a gutter, so I am keeping an eye on things and trying to insulate myself from that outcome.

I have a job. Based on my experience of 2001 and 2008, there is a very good chance that a crash would cause me to not have a job.

Losing a job sucks. But in the aggregate, unemployment in 2008 only rose like 6 percentage points. I'd accept that the traditional unemployment metric has its flaws, but while the crash then sucked for a lot of reasons --- I know friends who couldn't get internships and entry-level jobs that year, setting them back a year or more in long-term career growth --- the numbers suggest that most people kept jobs and stayed employed. Some sectors certainly got it worse than others, too.

After a few friends and family members overdose or blow their brains across their living rooms, that kind of argument becomes abstract enough that it doesn't really impact my decision making.

Would you really freeze to death in a gutter if you lost your job? You sounded pretty insulated already, in a different comment.

To be honest, I don't know.

I spent my youth dirt-eating poor and have been homeless more than once. It had a pretty profound effect on me, and I probably put more effort than is healthy into making sure it never happens again. I intellectually understand that I'm better off than most people, but at a gut-level I don't know if I'll ever feel "safe".

I know the feeling. I've been working on reducing my poverty mindset in the last couple of years.

The work must be done inside, not externally at Sunday school or whatever. Building a strong, supportive social circle would probably help in many ways, but in other ways it'd be just akin to band-aids than healing the wound itself, I suspect. Emotional memory reconsolidation can help with the wound itself.

Build a social circle helps a lot. Tried and true way is to join a stable or growing church, be involved, go to Sunday school. Can find ways to do it secularly, building a high quality social circle is basically a part time job (and can happen through a job). Marry a person with a strong family/social circle.

It can take some time but it builds inertia and can build generational inertia.

Are you doing the meme? What sort of question is that? I'm relatively well positioned, but I wouldn't want to lose my job in an environment where a good chunk of the workforce has also been fired, for example.

Lol. Never seen that meme before.

It's a serious question. There's no use in just carrying a completely vague worry.

To whitepillers: is there an argument for why I'm wrong that doesn't boil down to "you don't get it, chud, it's the New Economy! The Singularity is just around the corner! All the rules are obsolete!"? This argument is verboten, because this is pretty much what people say with every bubble.

The AI boom is mostly funded by private capital so the blast radius should be limited. Does anyone need to bail out Google or Facebook or Microsoft or Oracle? Nah.

The 2008 crash saw banks imploding due to correlated defaults and the government needed to bail them out because the economy needs banks to exist. This was pretty bad but the crash was only about 50% from a 2007 high and recovered within 5 years.

A crash to pre-AI markets in 2023 is one possible regression, which would be about a 40% drop. But this would mean ignoring all utility provided by AI aside from speculative gains.

The biggest unknowns are if a mildly bad crash causes a general mood of fear and depression to set in and make things much worse. Also, a big implosion in private capital could reveal unanticipated systemic risks.

I can't really get a clear picture either way.

Meta (Facebook) actually stands out as the one who might lose the most. They’ve invested heavily in AI with almost nothing to show for it and I think it might substantially unbalance their books.

Meta's LLM offerings have been pretty crap but they're still making lots of money off AI in extremely unsexy algorithmic improvements, using the offshoots of LLM tech, using all those GPUs.

So they have something to show for all they've spent. Ad revenue grew about 22% in 2025, in large part due to AI improvements. Though it's kind of diffuse and hard to assign revenue growth to a certain piece of tech when it's all backend, under the hood stuff.

AI is existential for software/ad companies like Facebook or Google, I don't think they have a choice here.

Meta's biggest contribution to AI is, by leaking their llama models early on, sabotaging the moat other companies were trying to build around the AI field through proprietary tooling, enabling the creation of robust open-source tooling that has imposed itself as necessary to support.

This also affects their own ability to profit from AI.

Does anyone need to bail out Google or Facebook or Microsoft or Oracle? Nah.

When they start slashing overseas jobs to cut costs? Maybe pull out of foreign investment entirely? This is going to hit those countries (my own included) catastrophically.

When there are mass layoffs at home? When the tanking share price hits the stock market?

The companies may be able to hunker down and survive, but I think it's absurd to imagine "the blast radius will be limited".

There is no such thing as ring fenced private capital in a crisis. VCs are backed by private equity backed by family offices backed by stakes in public companies intrinsically tied to the public bond market. You can’t stop contagion. Banks are a central leveraged financing source behind the current private credit boom as the recent MFS collapse here in the UK showed. Big private equity and credit players like Apollo bought insurers and then funnelled 20%+ of net assets into private credit, if AI reveals those SaaS companies to be largely worthless and they can’t fund their obligations that’s another major systemic problem because that leads to asset fire sales and a repricing of the entire credit market, which is catastrophic. Some Silicon Valley VC yoloing a few million on seed rounds isn’t systemic but that’s a tiny fraction of private capital.

I can’t say it means anything or not, but I can tell you many companies I sell to have already burned through all their AI budget for the year.

If there's a crash, I predict it'll be due to energy prices and the war in Iran, not AI.

Anthropic is nearly profitable, or supposed to have been profitable in Q2. Some say that's phoney accounting and Anthropic says profits may not be maintained. But it seems that major investments are paying off. Furthermore, AI models are getting continually better as with Fable most recently. This trend will continue, bigger and better models working for longer need more compute to run them and so justify further intensifying investment. I guess that argument is forbidden by your post. But if it is The Singularity, if it is a New Paradigm then presumably that's good for stocks!

Anyway, energy is more important to the economy than AI right now. You can't just shut off a huge amount of oil and gas production without ramifications! The 1973 oil shock is a useful precedent. The market seems to have been expecting peace talks to advance more smoothly than they actually have been. Bombing has just resumed. Iran has announced they're re-closing the straits and possibly the other straits in the Red Sea too.

Anthropic is nearly profitable, or supposed to have been profitable in Q2.

I'm suspicious about that given stories like this.

Maybe they had a revenue-positive quarter, but if it was buoyed by various companies overshooting their AI budget, this might not be an actual trend.

I've been suggested this was even an intentional move by an Anthropic investor to goose the company's valuation by artificially increasing their ARR (claims, not substantiated).

The 1973 oil shock is a useful precedent.

Not quite. The U.S. has been a net energy exporter since 2019. So domestic energy production can more than satisfy our needs for the foreseeable future.

That's before we get into the possibility of nuclear coming online.

Yes, higher oil prices impact the U.S. too, but we're not going to be the first ones to tap out, in either a literal or metaphorical sense.

Remember we just secured a massive source of nearby oil.

One thing about markets, they respond to shifts in supply and demand.

A company spending a crapton of money on Claude is good news for Anthropic, it means they have an addictive product, financially speaking. I don’t get why you’re spinning it as evidence against them.

I'm not under the impression that these companies are HAPPY to shell out this money, and if they start to more aggressively cap their employees AI use (Already happening). Other headlines say it accidentally spend that much.

This particular spend doesn't seem like it was received well. So if Anthropic being profitable (in the short term) requires a bunch of companies to recklessly/ignorantly run up AI bills well beyond what they expected or intended, that's not a sustainable model.

If you get companies ratcheting down their AI spend, then the numbers we saw for Anthropic in the last quarter are so will likely not repeat.

I'm suspicious about that given stories like this.

I basically don't believe this at all. That's literally a million instances on the fast/more expensive version of the highest cost model running for the entire month straight. The only way this is at all plausible is if their keys got leaked and millions of people were using them. doesn't pass the back of the envelope smell test.

Company Blew $500M On Claude AI In One Month Due To No Usage Limit On Licenses For Employees

How do you even do that, are we supposed to believe that some company with $500 million to spend didn't know how to use cost tracking? Cost tracking comes with the pleb-tier management tools I get!

And if they were trying to goose the valuation then why would they admit it was unintentional, it makes Anthropic sound expensive? (They are expensive). It sounds more like an anti-Anthropic story to me.

The story sounds greatly exaggerated or misleading IMO. Who is this mystery company too, what are the details on this? More likely some company just consciously spent a lot of money on Claude AI and then some reporter fluffed it up into a narrative we all hear instead of a boring article that goes nowhere.

Regarding energy, even though the US is a major producer and isn't as badly affected as other countries like Australia, Europe or poor countries, the world economy is global. Problems in Asia will spread to America. The US is busily exporting oil, including the strategic reserve releases, to take advantage of price gaps and stabilize markets. But this is a temporary fix. The price gaps will narrow. The invisible hand of the market will slowly but surely squeeze the US economy if this price pressure remains. Even if oil producers profit, much more of the US economy relies on truck transport, plastics, feedstock, jet fuel, lubricants and all the other chemicals which support industrial civilization.

Venezuelan oil is low quality and requires years of patient investment and capable administration to realize much net gains from. Crude oil production there is barely half of what it was in 2016 and shows no signs of making up for the current supply loss. Nobody is a bigger fan of nuclear than me but nuclear energy is not going to lower fuel prices in a matter of months.

https://www.ceicdata.com/en/indicator/venezuela/crude-oil-production

are we supposed to believe that some company with $500 million to spend didn't know how to use cost tracking

One of my tiny clients was billed over $10,000 recently because their customer-facing chatbot got locked in a loop with somebody's openclaw agent.

An acquaintance works for a small to mid sized private software company, and he alleges that they put restrictions on AI use after they already hit seven figures this month.

Overall, I don't know if that story is true, but what I'm seeing on the ground strongly suggests that executives really don't have the slightest fucking clue what's happening until the bill lands on their desk.

Why didn't they set a spend limit? Can't get tinier or more amateurish than my projects. I have a spend limit set up for precisely this situation, or if I get my API keys stolen. I'm pretty sure spend limits are on by default!

I guess there's just a huge gulf in discipline between businesses of similar sizes, the small companies I work with take pains to track all the dollars in their accounting.

I am less familiar with the AI space specifically, but cloud compute providers like AWS are famous for making setting up spending limits difficult generally. It's certainly convenient for them from a business perspective, but to be fair it is often unclear what "limit" would mean in the general case: nobody likely wants "we deleted all your data in S3 so you didn't exceed your spending limit on data storage", for example.

The small client didn't set a limit because another salesman/contractor set it up for them without a limit. They are not technical people.

The other company was pushing a tokenmaxxing scheme and got burned by their own choices.

The small client didn't set a limit because another salesman/contractor set it up for them without a limit. They are not technical people.

That seems to represent severe negligence on the part of the person who set it up, with foreseeable consequences. The limits are AFAIK turned on by default so the salesman must have turned them off. There's a decent chance they could get their money back in the courts IMO. Especially if it was a salesman for a larger company not a contractor.

If you went to court the company would probably just say that it's their policy to have the limits on by default, which would of course be true. You would then have to prove that it's the salesman who turned off the limits instead of you. Good luck doing that.

This raises the question- how much does a token cost, and how much does it do? I'm given to understand the standard software engineering productivity line is 1 line of code/hour, is a token more or less than that(and does it cost more than an hour of a software engineer's time)?

This discussion of tokens is sort of like 'the carpet costs 40,000 Kazakhi tenges(to pick random foreign money I'm not familiar with)'. OK, sounds like a lot, but how much is a tenge? If it's worth as much as yen, that's not a bad deal.

[apologies if I'm answering a rhetorical question]

A 'token' is the ML equivalent of a syllable: the word, portion of a word, or symbol that represents the smallest viable input or output unit of effort. The exact value (and cost) depends on model, as well as whether it's input or output.

I'm ... skeptical to endorse line-of-code as a measure of programmer time -- I've spent days planning out business-critical logic that ended up five lines of code and needed to be absolutely correct, and spat out thousands of lines of text in an hour before when it was just boilerplate -- but the output side you can give a pretty good average. Depends on the tokenizer and your output language, but I'd expect less an average of less than 20 tokens per line in C++ or TypeScript, and I'd get worried if a human coder was regularly writing >80 tokens in a single line.

((edit: less so in java.))

So pessimistically, 12.5k LoC per million tokens, more realistically 25k.

Input is the high-variance part. If you're writing from scratch, the input is a few paragraphs and some design documents, maybe some scribbled image files if you feel spicy and the model supports it. I've done a few personal projects like that where it's been <2k tokens to get 20k line-of-code. If you have an existing codebase you want the model to adjust to, or an API document you need the model to learn, that can burn through a lot of tokens fast; the only real restriction is context window size, and most of the corporate APIs obfuscate that (tbf, often because they have an automated store and search strategy). I've blown through 50k in a single search once (thanks, Atmel, love your manual layout too). Input is typically cheaper and there's some strategies to reduce the cost of repeated input hits with the same content, but they're complicated and pretty specialized.

For some examples:

Model Input (USD/million-tokens) Output (USD/million-tokens) Output (USD/thousand-line-of-code)
Claude Mythos $10 $50 $2
Claude Opus 4.8 $4 $25 $1
ChatGPT 5.5 $5 $30 $1.2
Grok 4.3 $1.25 $2.50 $0.10
Qwen3.7 Max $1.25 $3.75 $0.15
Qwen3.7 Plus $0.32 $1.28 $0.05
Qwen3.6 35B-A3B $0.15 $1.00 $0.04

For smaller or more efficient models, inference is pretty cheap: Qwen3.6 35B's probably the weakest coding model I'd use in a professional environment (and borders the point where it might be better to run it locally, if only for privacy/security reasons), but there's a lot you can do.

That said, all of that can go out of the window when you start getting agentic options involved. Someone made a fun experiment of trying to let a local model figure out a display protocol by hooking a camera, an LLM, and a microcontroller together, and they got it mostly there overnight, which is really cool. It also probably burned tens of millions of tokens on output for an interface code block that should have ended up in the <1.5k line-of-code level.

I'm ... skeptical to endorse line-of-code as a measure of programmer time

Hell, I'm at negative LoC for the year so far.

Deleting code feels a million times better than writing it

So TL;DR, tokens are cheaper than basically all white collar workers in the modern west, but they get overused for fun projects?

Largely (though Mythos approaches white collar wages in terms of dollars per hour at API rates). But it's not like all tokens go straight to code written. Tokens are more like measuring thought.

When I give Mythos/Fable instructions it first goes 'I'll explore the codebase' and so it searches for relevant things (those search commands are output tokens). Then it reads files which have the relevant data, more input tokens. Then it thinks for a while (that's output tokens). Then it makes its to do list. Then it reads some more, thinks some more. There are pages and pages of just reading and thinking before it goes 'i have a full picture'. Then it starts editing code!

Then it'll try and test if it actually works, often writing some test cases, so that's more code. Then it tells me everything it did in summary and adds stuff to its memory files.

So a lot of thought is happening even if it only adds a few pieces here and there for a new feature.

More comments

More that they're cheaper than a code monkey, only weakly expensive if you're doing something hard or novel (or novel-to-you), and they can get ludicrously expensive if you just start firing the slop cannons, either to solve a problem by volume or by producing a lot of useless or specialized lines-of-code.

And because they allow brute-forcing problems in ways that weren't possible before, or tackling new problems. Or because the user cocks up, as in the case where they fail to notice their two bots getting into an infinite loop.

The problem is that a token is cheap, but the amount of tokens you need to do useful things can be very high.

For agentic programming, the agent needs to hold a non trivial amount of the codebase in the context. That can easily be millions of tokens. Then you have whatever pile of "skills" (read: markdown files) you use, then add the various layers of prompts, then add reasoning chains. It adds up very quickly.

Once you start adding parallel agents and loops, it can get insane.

More comments

A token is broadly a word. 'Tokenisation' is what happens because AI is fundamentally mathematical, and so it only works on numbers, so we have to turn language into numbers.

So if a line of code looks like:

def my_function(my_variable):

Then that's

def| |my|_|function|(|my|_|variable|)|:

where each | is a split between tokens.

So that line is 11 tokens: one for each common word, one for each element of punctuation. In practice there's a big table with each word and each punctuation assigned to different numbers, and it looks up the numbers, so that line of code gets turned into the numbers (tokens)

234 756 32423 56 789789 2334 54 56747 35423 2354 213

and each of those numbers costs a certain amount to process. Each new token, i.e. each new word and punctuation mark it produces costs another (considerably larger) amount. There's a certain amount of complexity around how to represent numbers for maths, and some fairly commonsense rules about how to split up words that are uncommon, so "unfireable" might become the three tokens for un|fire|able and "garbleflarg" might have to be spelled out letter by letter with each letter being a separate token.

(Note: this is how the little guys like me do it. Tokenisation for Anthropic might be something much more advanced.)

How much a token is worth depends on what it's doing for the customer. If it's a word in my romance novel then it's priceless not worth much, if it's a word in the code for my startup, it's either worthless or very valuable depending on how that startup turns out. Or if you like you can decide it's worth what it would cost to get a suitable skilled human to do it, which is generally how the big companies value it.

A good mental model for a token is 3-4 characters. I'm assuming somebody will come in with an example of how it's wrong, but it's not a terrible heuristic.

A line of code is usually between 1 and 200 characters, with a fat part of the curve sitting around 80-100 characters.

their customer-facing chatbot got locked in a loop with somebody's openclaw agent.

I'm sorry for the company because that kind of money can hurt a small company, but this is just too perfect. Do away with even the callcentre human element for the sake of "save money, more efficiency" and you end up with no real humans in the loop, just bots jabbering at one another.

These are the days of progress towards the Singularity, just believe in and hold fast to the shining vision of paradise on Earth!

These are the days of progress towards the Singularity, just believe in and hold fast to the shining vision of paradise on Earth!

A whole lot of people died due to boiler explosions in the early days of the industrial revolution, but in the end it all turned out quite awesome. (In the old sense of being grand and powerful. Don't @ me, Uncle Ted fans.)

This pattern of argument is not convincing -- where people bring up anecdotes about most spectacular fuck-ups and then use them as basis for sarcastically mocking the idea that anything useful could come from this.

The sarcastic mocking was more about "this is where the grand dreams of artificial minds that are better than human has ended up, money-scamming machines communicating mindlessly with each other to burn time and money". This is what we do, we turn visions into "how can I scrape pennies out of this?"

Great things may come later, but we'll still be trying to scrape pennies.

Overall, I don't know if that story is true, but what I'm seeing on the ground strongly suggests that executives really don't have the slightest fucking clue what's happening until the bill lands on their desk.

Anecdote from a previous job that you just gave me a PTSD flashback to, back when I worked on the help desk and sysadmin side of things instead of development:

The company I was working at got acquired by another company (most Americans here would probably know the acquiring company but I won't get any more specific than that). Our accounting department was mostly laid off and we now had to send our invoices to accounting in the acquiring company to get them paid.

My boss starts sending them our invoices for the phone company (and internet, and a bunch of other important things) up to corporate accounting every single day because accounting isn't responding to him and isn't paying our bills. He's also calling them multiple times a week, but no one is answering. Also our long distance phone service is separate from our main phone service (this will be important) for complicated reasons I never bothered to learn because I was a help desk grunt at the time.

This goes on for months, and the phone company is getting pissy and threatening to cut off our phone service. My manager is forwarding the service cut off threats to accounting too. Finally long distance service actually gets cut off (but local phone service still works), and me and the other help desk grunt got flooded with about 200 calls from pissed off users that day.

This causes enough of a stink that corporate catches wind of it and ask my manager why he wasn't paying the phone bill. After all, they showed him how to send invoices to accounting, etc., how could he be so irresponsible? My manager whips out 60+ emails and his phone call logs and corporate immediately apologizes and presumably goes to bite off someone's head in accounting.

Perhaps this is a stupid question, but why couldn't your manager escalate to his boss until it gets to someone who can do something about it?

Yeah, I assumed reading the story that something must have been communicated up the chain because without that, it just looks a lot like malicious compliance.

If I recall correctly, my manager's new manager was someone in IT at the corporate HQ and they got some blowback over it too because my manager had reached out to them several times about accounting not responding.

I believe your war story, because I've seen the likes of it in a previous job. Being as vague as possible, two separate entities were amalgamated into One Big Happy Family for the sake of efficiency and cost-cutting and other fun management notions. It was all going to be peachy, the matters handled by both halves would now all be handled in the post-amalgamation blob and this would mean Better Customer Service and More Responsiveness and the other buzz words.

First thing to blow up was the annual Christmas party (I hadn't worked there long enough to have gone to previous parties). Before, both places had their own parties and management of both places threw a few bob in the respective funds for it. Afterwards, entity A (based in the city) would not come down to our town, and entity B (based in our town) would not go to the city to host it, because transport (everyone wanted to get blotto on free-ish booze because it was the Christmas party, nobody wanted to have to be sober enough to drive home). Plans to arranging hire of private buses (so people would be collected and then dropped back home or near enough) foundered on "yeah, but where will we hold it?" There wasn't any compromise "here's a nice restaurant or hotel halfway between both places", so it ended up no Christmas party at all for anyone.

That was the kind of co-operation and mutual understanding which developed between the two halves, which post-amalgamation continued on with "we do our stuff, you do your stuff, we don't interact or co-operate any more than absolutely necessary".

I imagine the reason for the accounting snafu was a combination of acquiring company accounting department going "Well nobody told us we were supposed to pay their bills" so the bills coming in got shoved into a pile on somebody's desk and ignored, and "who the hell are these guys, never mind, go to the bottom of the heap while we deal with the really important payments for our main office".

I agree that the story is very vague, some mysterious unnamed company ran up that much of a bill? But the plausibility lies in if the company didn't understand what they were buying - they thought the licence worked like the usual software-as-a-service subscription and you just buy a licence for everyone who is going to use the product. They didn't realise they were being charged for token usage or how tokens worked.

As a cautionary tale, this hit my own workplace where unexpected heavy usage of photocopying meant big charges for excess monthly, to the utter surprise of the boss who thought we were paying for all that with the monthly service charge anyway. Nope, this is why you always read the fine print (because the buggers selling you the crap are always going to present the best face of it then hit you with the "and also if you do this, this, or that, which you will do because that's what you need to do, we charge you an excess" after you've signed the contract).

According to social media, it seems that several companies have ordered their employees to maximise token usage, thinking that this would launch their company into the future. I don't know about the $500M figure specifically, but it seems entirely plausible that this practice would lead to a ton of money being wasted.

I've not seen a single proper source for that 500m claim either, but everyone seems to just accept it at face value.

The one time I saw something like an original image for the claim, it was from an Indian subreddit (or twitter, I can't remember exactly) and had no currency attached. So if it were 500m rupees, that would only be 5m USD. That would make a lot more sense to me

Indians tend to use Lakhs and Crores though

I'm actually quite positive about it although I hope the madness doesn't cause massive political instability worldwide. Destruction is necessary for new growth to happen.

Most of the problem comes from the current energy crisis, the way growth has been artificially pumped due to oversupply of (dumb) money, and too long a tail on ZIRP that we're feeling the hangover from now. The AI pump is because most of the money has nowhere to go and line must only go up; they're betting on the future of AI being significantly more valuable than it is today, because those valuations can't be justified any other way.

The energy crisis I consider a good thing as long as it doesn't lead to major geopolitical conflict over the scramble for resources. More than anything else, Trump may be remembered over the long term as the president who put the cornerstone on the long term energy transition to nuclear. The current pressure on energy prices has a lot of the whiners shut the fuck up real fast about alternative energy, and I hope that people move fast enough to avoid the next-memory holing of the crisis.

For the first time in decades, in my entire life, the US seems to actually be excited about real huge scale projects. We lead the world undeniably in a new state of the art, we have energetic and ambitious figures fighting each other in a titanic race to innovate on projects at never before attempted scale. A site the size of Manhattan! that's small beans lets do it in space! It's just such an incredible bummer how cynical and bitter so many seem to be about it. Truly it could all go wrong, scaling could stop it's definitely within the distribution of possible future that tokens become a commodity in such a way that these investments don't pay off. But if you wanted to make America great again, if you wanted to actually build impressive physical things here at home and show the world what we're capable of, then this is our ticket. Were you under the impression regeneration was going to be a sure thing?

I feel like I'm watching people sabotage nuclear energy all over again over bullshit hallucinated fears of disaster or made up sour grapes about how huge amounts of energy from fission wouldn't even be that useful. We made computers talk like people and they've progressed in only a few years from barely coherent chat partners to genuinely useful junior programmers. I think there's a fair chance this whole thing ends up with all of us dead as well as a fair chance that progress peters out but how can you not be excited?

I think there's a fair chance this whole thing ends up with all of us dead

as well as a fair chance that progress peters

Are you really surprised why a technology where its proponents and developers are saying that it has a solid chance of killing everyone and/or consigning the median person into permanent unemployment at best, or creating a titanic financial bubble that's going to bring everyone's retirement with it if it doesn't pay off at worst, is extremely unpopular?

As a software engineer I enjoy using coding agents and I do think with demographics and fertility being what they are, the only real choice as a society is leaning into AI and automation as much as possible even with the commensurate risks, but frankly the median person has really not seen proportional consumer surplus from the trillions invested into post GPT-4 or so LLM development, and in return all they get is dealing with a tsunami of online slop, vibe-coded software that gets worse even as engineers boast about how productive they are, and higher prices for power and RAM.

At least in China they have the message discipline to tell people that AI is going to be used to improve people's quality of life, but all SF can do is jerk off about the AI-induced permanent underclass happening any month now and how dangerous AI is is going to be; it's obvious why the median person hates the AI build-out.

At least in China they have the message discipline to tell people that AI is going to be used to improve people's quality of life, but all SF can do is jerk off about the AI-induced permanent underclass happening any month now and how dangerous AI is is going to be;

I have a snarky Freudian theory that the rise of the tech right and a lot of SV comms missteps have the same root cause - that the founder/VC elite are failing to contain their white-hot rage at having to pay upper-middle class salaries (and extend the respect the upper class traditionally extends to the upper-middle class) to mere codemonkeys who lack upper-middle class social graces.

In this model, the "permanent generational underclass" meme is founders, VCs, and people who see themselves as future founders and VCs expressing glee at the codemonkeys losing undeserved social status. From the point of view of someone who can afford a family-sized house in the Bay Area, nurses, plumbers etc. are already part of the underclass.

If that was the issue for them they’d have fired all the expensive American software engineers making $300k years ago and replaced them with French, British and Germans making a quarter or a third of that.

I think both sides of the culture war agree that the Valley did try to do this, but with Indians instead of Europeans. They disagree on whether or not it succeeded, and whether or not it would be a good thing if it did.

They tried. Didn't work. The French, British, and Germans at home weren't good substitutes, and the ones who came to the US demanded the $300k.

But I would have pegged you (along with BC) as one of most likely to be experiencing that white-hot rage at the wrong people making the big bucks.

I think nobody deserves to make one cent less than the amount they can fairly negotiate. IQ and demographics do not indicate a great superiority for American tech workers. US companies achieved dominant global positions because of larger domestic markets, far more readily available capital, network effects and superior entrepreneurial culture. But none of these suggest US engineers are far better or smarter than their European peers. That would be unlikely.

The pay is likely load-bearing for the engineers just as it is for the founders. Pay a top engineer a normal middle class salary, and you'll get good, solid work out of them. Give them a chance to get rich and you'll likely get a lot more.

In this model, the "permanent generational underclass" meme is founders, VCs, and people who see themselves as future founders and VCs expressing glee at the codemonkeys losing undeserved social status. From the point of view of someone who can afford a family-sized house in the Bay Area, nurses, plumbers etc. are already part of the underclass.

What is the downside to just throwing all these people off a cliff again? I'm actually asking.

The right-wing misanthropic VCs, the no-longer-needed codemonkeys, or the plumbers? Tossing any of the three is a plausible sentiment for at least one Motteposter to hold.

But if, as I suspect, you mean why can't we throw Elon Musk and Marc Andreessen off a cliff, the answer is that Musk is too valuable to the work of society given that he can build space rockets 1-2 orders of magnitude cheaper than anyone else. Andreessen, on the other hand...

Why would anyone wanna throw the plumbers off a cliff? I’m not trying to be dense here that just seems like a weird group of people to be angry at.

Childhood trauma from Super Mario?

In the UK, because too many of them are Polish.

It's not that weird for committed shitlibs. Union plumbers are ultramaga republicans.

Granted we don't have a lot of people on the motte that want to clear out all the magats on the grounds that if they can do it we can do without it. But the bluesky crowd being horrified at the politics of the experts at keeping their lights on and their water running is a semi-regular occurrence.

Marc Andressen was the first person to come to mind when I wrote my response to the OP. The guy was born to middle-class parents in Iowa and raised in Wisconsin. He was himself a computer programmer. He's not the kind of elitist described. (Also you can't throw him off a cliff because the man is huge.)

It's been a very long time since Andressen was checking in code. It is obvious from reading e.g. Paul Graham's essays that VCs and successful founders see themselves as a different kind of person to top technical talent (tl;dr - Altman said that James Bond was fundable but Q would not be) - and the things that Paul Graham is saying are conventional wisdom in the valley. My impression is that both founders/VCs and top technical talent see the gap between top technical talent and mid-tier technical talent as so large that the average FAANG Senior is closer to QA and HR than to themselves.

My take is that Andressen is one of the main people pushing the theory (which I think originated with Musk) that FAANG were roughly 2x overstaffed and that this was a bad thing because they were outbidding startups for talent. I remember him tweeting "Nature heals itself" in response to a post about tech layoffs and coming to the conclusion that he resented his (indirect via portfolio companies) employees.

I will admit to also disliking Andressen because a16z seems to be disproportionately likely to invest in ethically dubious startups like crypto scams, Adam Neumann's Flow and the AI "cheating" app Cluely.

I (a former FAANG engineer) would agree with him that founders are different sorts of people than top technical talent; at least most of them are. (Steve Wozniak is one obvious exception: it was Jobs who had the non-technical talent in spades). And that Google, at least, is ridiculously overstaffed. That's different than resenting "code monkeys" making a lot of money; wanting technical talent to be available to startups is valuing them highly, not considering them lower organisms of some sort. It might be better for the engineers to phone it in and collect a paycheck on the latest obviously-doomed FAANG project rather than bust their butt at a startup, but Andressen's desire for them to do the latter is based on greed, not resentment.

More comments

Also you can't throw him off a cliff because the man is huge

Huge as in obese or jacked?

I’ve only seen pics or interviews of him sitting or from the waist up. From what I could tell he always came across as a pudgy short computer nerd type. With a head shaped exactly like an egg.

I think he's like 6-7.

Okay actually an in inch or two shorter but still a giant.

Which people? There's a lot of groups mentioned there. The least valuable are probably "people who see themselves as future founders" (yeah, yeah, ideas are a dime a dozen)... and the people who have that rage. Who I don't think are actually SV founders and VCs; while some are traditional business types, and a few even have multi-generational money, most are noveau-riche just like the "code monkeys" they're paying. They may not like paying them, but that's pretty much for financial reasons. There are definitely those who hate that gauche technical people get paid big, but it's not that group.

Are you really surprised why a technology where its proponents and developers are saying that it has a solid chance of killing everyone and/or consigning the median person into permanent unemployment at best, or creating a titanic financial bubble that's going to bring everyone's retirement with it if it doesn't pay off at worst, is extremely unpopular?

The alternative is that we become Europe, a pathetic retirement home telling itself unending stories of its greatness or the populist have us tear ourselves apart on the pathway to becoming so. The chance of it killing us all is coming either way unless we can get yudd style international treaties to pause/slow down(which I do support but am not expecting to happen). And the likely results of trying it and it crashing don't look terribly different to me than the trajectory we're on if we don't take the swing. A bunch of retirement counts but especially silicone valley investor types get hosed and we all get ridiculously cheap cloud compute for a generation.

frankly the median person has really not seen proportional consumer surplus from the trillions invested into post GPT-4 or so LLM development, and in return all they get is dealing with a tsunami of online slop, vibe-coded software that gets worse even as engineers boast about how productive they are, and higher prices for power and RAM.

I mean of course they haven't because those dollars haven't even been meaningfully spent yet because we haven't built the datacenters. The models people are using are the results of more like the tens to low hundreds of billions of dollars and even that takes time to actually build into products people would like. It takes time for actual artisans to learn to use new tools to produce more than just slop.

At least in China they have the message discipline to tell people that AI is going to be used to improve people's quality of life, but all SF can do is jerk off about the AI-induced permanent underclass happening any month now and how dangerous AI is is going to be; it's obvious why the median person hates the AI build-out.

The median person who hates AI thinks a literal bottle of water is obliterated from the universe every time you query chat gpt for anything. Almost all of the concerns raised by the median anti-ai person are retarded propaganda produced by slopulists who, correctly, think of them as cattle that they can lie to with impunity in order to gin enough outrage to propel them into office, keep them supporting their slop rage social media account or because they're genuinely foreign spooks trying to retard our progress.

The median person who hates AI thinks a literal bottle of water is obliterated from the universe every time you query chat gpt for anything. Almost all of the concerns raised by the median anti-ai person are retarded propaganda produced by slopulists who, correctly, think of them as cattle that they can lie to with impunity in order to gin enough outrage to propel them into office, keep them supporting their slop rage social media account or because they're genuinely foreign spooks trying to retard our progress.

Actually, this exact perspective is the thing that the median person hates about AI. The water usage isn't the issue, the slop isn't the issue, the fact that AI doesn't produce any value for the median person and instead makes much of their life notably worse isn't the issue - the issue is that pompous 'tech elites' think they're better than us.

I'm personally of the opinion that the AI ecology in its current instantiation is evil in almost every single way, but the biggest way it is evil is that it's a transparent attempt to destroy the lives of everyone who isn't a part of that ecology. It's a predatory economic model, and the prey is that median person that you have such derision for.

AI doesn't produce any value for the median person and instead makes much of their life notably worse

Maybe I'm not a "median person" but AI has definitely improved my life in many ways and has made it worse in roughly 0 ways.

Basically every social problem AI is claimed to cause was a massive issue pre-2022.

"Oh no the internet is full of retards, slop and advertising!!" is it 2005, 2015, or 2025?

I think that last sentence exactly shows why you're out of touch. I'm not complaining about the effects on internet culture. I'm talking about the extreme damage to our education system, where students use LLMs to complete the assignments that were created by LLMs and will be graded by the same LLMs, and the data shows massive loss in basic knowledge and critical thinking skills. I'm talking about the environmentally catastrophic data centers being erected over the protests of the people living there. Real harm to actual people in the Real World.

extreme damage to our education system

So your stance is that the US education system was doing fantastically in 2021? No pre-existing problems or awful trends?

students use LLMs to complete the assignments that were created by LLMs and will be graded by the same LLMs

This is so depressing. I agree with you here. I hate that this is happening.

I don't blame LLMs though. I use LLMs to teach myself new things all the time.

I blame the absolutely shit school system, which has been shit and wildly unable to adapt to anything my entire life.

environmentally catastrophic data centers

How are they bad? Can you actually substantiate this? What makes them worse than giant suburban tracts of sprawl?

Of course education has been bad for a long time, but the fact that it's bad doesn't mean AI isn't making it worse.

As for the data centers, there's a lot of independent reporting on this from people living near them on immediate harms. There's damage to groundwater that makes the local tap water undrinkable, there's constant noise levels in residential areas.

On a larger scale, data centers are now responsible for 1% of global greenhouse gas emissions, and projected to rise. This paper estimates it at $25 billion in environmental damage per year.

There's also other more subtle impacts. For instance, the Utah data center project is expected to produce a heat island effect that would raise temperatures from 5 degrees during the day to up to 28 degrees at night, potentially preventing the dew point condensation effect and quite literally killing local ecologies.

All of these effects are being ignored in the name of progress, and the costs of these externalities are being paid by the people living in these areas who have no say in the matter.

More comments

People caught up in slop anti-growth and anti-human ideologies are worse than people who aren't. You weren't born this way, you can stop being worse than people who are better than you instantly by coming to your senses. No one is forcing you to identify with a hateful dead end black pilled ideology who's arguments are soldiers because they have no actual arguments. In every way that matters you're no different to the communists who cope with their own failures and salve their egos by inventing conspiracies of the rich. And the most pathetic element of it is this pure bullshit demand to have your feelings coddled without there being any condescension. There is no way to be this careful about hurting your feelings without condescension, it's not possible.

"It's not any of the things I say it is, it's that you don't pretend to like the people saying ignorant bullshit trying to stop progress convincingly enough"

Give me a break.

Thanks for reinforcing my point! Somehow, in your view, it's impossible to disagree with your take on AI without being 'pathetic' 'anti-human' who can't 'cope with their own failures'.

I rest my case.

Certainly it's possible to disagree with my position. But you've not disagreed with it, you've explicitly said the problem isn't my position, it's the rhetoric used, it's your feelings. I have a number of takes on AI that I think have very reasonable foils, I've gone over them at length. If you want to engage with them then please do so. But if your demand is that I show respect for slopulist liars who have contempt for the truth then I refuse.

Sorry, no. If you just go back and look, I originally responded to your claim that

The median person who hates AI thinks a literal bottle of water is obliterated from the universe every time you query chat gpt for anything.

by arguing that the median person's position on AI is essentially rooted in the (correct) belief that AI and tech bros in general are contemptuous of them and their concerns. This is not a 'slop' view nor is it a 'demand to have your feelings coddled'. It plays out in all areas of the AI ecology, from billionaires paying off city councils to build data centers that almost every constituent opposes, to companies firing large parts of their workforce because of 'AI efficiencies' that don't actually exist, to the degradation of our public education system by replacing teachers with AI tools that don't work... shall I go on? AI boosters are deliberately and with malice aforethought trying to destroy the lives of the 'median person' and treating that as just an unfortunate externality of their race towards godhood. It's the very definition of evil.

More comments

The chance of it killing us all is coming either way

I generally agree, in the sense that "WW2 was inevitable as a result of complex geopolitical factors post-WW1", that "if the fate of human civilization is to create the Torment Nexus, then the Torment Nexus will inevitably end up being created", but obviously that didn't clear Hitler's name and being on the vanguard of potentially creating the Torment Nexus is simply not going to make AI labs popular.

I mean of course they haven't because those dollars haven't even been meaningfully spent yet because we haven't built the datacenters

I mean, is that supposed to make it a more appealing thesis for the average person? "You should give us a trillion dollars to build out datacentres that aren't being built, and maybe if you give us enough and we finally succeed you'll be lucky to get something out of it and avoid the permanent underclass or being drone striked by Skynet" is not exactly stirring rhetoric. I'm not saying that AI is useless or that it couldn't theoretically generate a lot of consumer surplus, but that's clearly not what the public messaging on the topic in the West is pushing.

The median person who hates AI thinks a literal bottle of water is obliterated from the universe every time you query chat gpt for anything

While it's true that populist anti-AI arguments are not very good and mostly arguments-as-soldiers, that's distinct from not having good reasons to act that way in the first place. There's a dedicated core of people who hate phones, cars or fracking, but most people just enjoy the benefits to their lives from those technologies and move on with their lives; there isn't really such consumer surplus and many externalities from the current state of AI, so it's no wonder why people start turning to populists and arguments as soldiers.

I generally agree, in the sense that "WW2 was inevitable as a result of complex geopolitical factors post-WW1", that "if the fate of human civilization is to create the Torment Nexus, then the Torment Nexus will inevitably end up being created", but obviously that didn't clear Hitler's name and being on the vanguard of potentially creating the Torment Nexus is simply not going to make AI labs popular.

Yeah, we should get on those bilateral international treaties that Yudd wants. Until then the alternative is unilateral ceasing where we lose all the upside and delay the problem by optimistically 6 months to two years. I don't really see that as a reasonable choice, especially because approximately zero of the people you're talking about are taking x-risk seriously anyways. This subject is more about me hedging my excitement and has nothing to do with what the opponents actually care about.

You should give us a trillion dollars to build out datacentres that aren't being built

They are being built, just have not yet finished being built. and no one besides investors are being asked to give them anything. If people don't want to be personally invested in these projects then I have no real beef with those people. I haven't decided exactly what my own financial position will be in relationship to the IPOs.

maybe if you give us enough and we finally succeed you'll be lucky to get something out of it and avoid the permanent underclass or being drone striked by Skynet

You're conflating a number of different people and positions. none of the labs are pushing the "permanent underclass" line. That's an expressed anxiety of some people on social media who feel something big is coming and they don't know how to prepare for it. It's like when people get indignant at "tech bros" for the learn to code meme. It wasn't tech bros that proposed miners should learn to code, it was random journalists. Anthropic is currently in the process of getting blackballed by the department of war for taking a stand against their models being used for domestic surveillance and autonomous weapons. The labs and their staff have been unusually honest and forthright with their concerns about how we need to work out as a society what we're going to do if a singularity event happens. For the most part people like op have accused them of fear mongering to hype their product. No doubt if they downplayed the risks they'd be accused of lying about the potential downsides.

While it's true that populist anti-AI arguments are not very good and mostly arguments-as-soldiers, that's distinct from not having good reasons to act that way in the first place. There's a dedicated core of people who hate phones, cars or fracking, but most people just enjoy the benefits to their lives from those technologies and move on with their lives; there isn't really such consumer surplus and many externalities from the current state of AI, so it's no wonder why people start turning to populists and arguments as soldiers.

I don't know what to do here. Really. I have no idea what I'm supposed to do about people who make terrible and ignorant arguments as soldiers against fantom interlocutors. That's one of the most frustrating elements of this whole topic. There are civilizationally important conversations to be had. But all I'm presented with are people who somehow simultaneously believe that the whole industry is a scam that doesn't even work while also being an imminent threat to all jobs which it will somehow be able to remove without actually providing any value. And this incoherent position is expressed in the form of on their face ridiculous empirical claims. It's exhausting.

More than anything what I feel about these people is that they just want to be mad, half of them have been whining for years about American decline and incapacity. This is a field where we are excelling, if America is going to right itself and bring about a new age of American prosperity this is it, this is the chance. Should we squander it? Certainly one should be cautious and prudent but that isn't what these people are. They have already written us off and want things to get worse to validate their black pilled priors. They'd rather this all fall apart so they can smugly say "I told you so" in the ruins. I fear some might even, if they could, sabotage us to this end. I want to wake them up.

They are being built, just have not yet finished being built. and no one besides investors are being asked to give them anything

Money is fungible. To a first approximation, if a trillion dollars is going into trying to build data centers (and failing, in many respects) that's a enormous chunk of resources that isn't going into fusion research, hiring, fixing bridges, what have you - things that could directly and meaningfully improve people's lives.

FWIW I'm optimistic about AI as a normal productivity-improving technology for improving people's lives, but this is explicitly not what the frontier American labs are aiming for, and it seems extremely difficult to justify the trillion dollar investments in a AI-as-normal-technology world where tokens are inevitably commoditized and primarily inferenced locally.

Should we squander it?

I guess I'm just feeling a disconnect from your posts. If you want to be a hard-nosed historical realist, OK: in the same way that the Treaty of Versailles and the Great Depression made WW2 inevitable and Hitler was merely in the right place at the right time, I agree that race dynamics ensure in no realistic world does a bilateral pause happen, and that as soon as Attention Is All You Need was published realistically no force on the planet could have stopped the AI race from accelerating - Altman and Amodei were merely in the right places at the right times to ride the wave.

But then you go on about "how can you not be excited" and "[bringing] about a new age of American prosperity" and I'm just confused about how you can think we're being forced into a world where if anyone builds it everyone (probably) dies and think this is a good thing. At best, if by some miracle alignment is inherent, you're under the heel of God-Emperor AI Frontier Lab Leader or God-Emperor Democratic/Republican President (adjust for your outgroup party) for all time and at worst all America's done is won the "created the Torment Nexus first" award.

At the end of the day "American" and "Chinese" are merely constructs of society as we know it and not immutable laws of reality - I fail to see how "superhuman general intelligence" and "liberal democratic All-American values as we know them" or "superhuman general intelligence" and "socialist values with Chinese characteristics as we know them" could meaningfully co-exist, so what benefit does the median person really get from ultimate, unaccountable power being instantiated 2000 vs 6000 miles away from them?

Money is fungible. To a first approximation, if a trillion dollars is going into trying to build data centers (and failing, in many respects) that's a enormous chunk of resources that isn't going into fusion research, hiring, fixing bridges, what have you - things that could directly and meaningfully improve people's lives.

Their money is not fungible with our money or the state's money. much of it is debt that can only be created because of the promise of a return. If you want to be more sophisticated you could say the labor being used to build these things was bid up and would otherwise be able to be allocated more cheaply to other projects but even then what you're arguing for is that trades people get paid less which seems kind of counter to the rest of your position. Trillions are spent on much less worthy consumption than building durable data infrastructure.

it seems extremely difficult to justify the trillion dollar investments in a AI-as-normal-technology world where tokens are inevitably commoditized

Yes, the question of whether the tokens get commoditized is a major factor on whether these investments will pay off for the labs. But why is that your problem? So in one potential future a bunch of investors who aren't you get soaked and you get to enjoy the fruits of incredibly discounted inference capex. This is a gift directly from frontier lab investors to you, you should cheer on their generosity.

where tokens are inevitably commoditized and primarily inference locally.

Unless many breakthroughs happen this will not be the case any time soon. Batched queries are several orders of magnitude more efficient than single thread inference.

But then you go on about "how can you not be excited" and "[bringing] about a new age of American prosperity" and I'm just confused about how you can think we're being forced into a world where if anyone builds it everyone (probably) dies and think this is a good thing. At best, if by some miracle alignment is inherent, you're under the heel of God-Emperor AI Frontier Lab Leader or God-Emperor Democratic/Republican President (adjust for your outgroup party) for all time and at worst all America's done is won the "created the Torment Nexus first" award.

There's some difficulty of keeping multiple potential futures in your head at once. It goes something like:
Outcome Probability Thoughts
World where someone builds it and everyone dies but they wouldn't have if we told more blatant and ridiculous lies about the American data center build out <0.0001% whooops
Somebody builds it and everyone dies 15% This is indeed too high and we should seek bilateral global treaties to try and sort out alignment while we can. Unilaterally hamstringing our labs doesn't change this possibility, even us and china both hamstringing our labs only pushes the date back some number of years.
Scaling plateaus and inference is a commodity soon 25% The lab investments don't pan out, all go bankrupt, the models and datacenters get acquired and are served incredibly cheaply. This is bad for some investors but good for most consumers of tokens, which is practically everyone else. A gift from them to us. We should be broadly happy that these datacenters are on our soil and subject to our laws
Scaling plateaus and inference is a commodity in a few months or years when the models are so good that they can do most but not all jobs 30% The lab investments may or may not pan out as their commodity is still meaningfully scarce it just turns out more of that scarcity is in the data centers than the models, some go bankrupt. We see a lot of job displacement but also a tremendous amount of value creation. Robots advances in leaps and bounds. After a shake up this is potentially a very good future. It is critical that these data centers are on our soil and subject to our laws, that are incredibly important geopolitically.
Singularity event where we don't all die 25% basically by definition these are difficult to predict but I'm confident we'd rather have it kicked off on our soil than chinese soil. The marginal progress we make in alignment work between now and the kick off will be critical and china doesn't broadly seem to even understand alignment to be a meaningful concept. If I had more time I'd break this into several worlds but I think in nearly all of them there'll be pivotal moments that I prefer be performed by americans.
I'm a retarded hype monkey with deep ai psychosis. AI can't even do what I can plainly see it doing now. 5% Hey, at least I'll get some credit for acknowledging it's possible

Essentially across all of these worlds and acutely in some these data centers are pure upside epsilon the handful of upset caused by any construction project that should be easily sorted out locally.

Money is fungible. To a first approximation, if a trillion dollars is going into trying to build data centers (and failing, in many respects) that's a enormous chunk of resources that isn't going into fusion research, hiring, fixing bridges, what have you - things that could directly and meaningfully improve people's lives.

No, production is not zero-sum and in any case money is not the limit of these things. God could come down from Heaven and declare a halt to data center development, and we would not get one step closer to fusion (might actually slow things down), and not one more bridge would be fixed (possibly fewer, in that the data center people might need a bridge fixed).

Money is an abstraction over resource allocation in this context - I agree that production is not zero-sum, but many of the inputs to production are obviously rivalrous; land, power, raw resources, construction workers and chip fab allocation are all finite and rivalrous to name a few.

If you're just trying to be edgy about lack of state capacity being why shit doesn't get built then I agree, in many cases resource allocation is not the limiting factor as to why no progress is made, but it's farcial to claim that resource allocation isn't the limiting factor for anything that might directly improve people's lives.

More comments

You're conflating a number of different people and positions. none of the labs are pushing the "permanent underclass" line. That's an expressed anxiety of some people on social media who feel something big is coming and they don't know how to prepare for it.

I mean this in all sincerity and good faith, and not as a rhetorical question: are you hearing yourself talk right now? You do understand why many reasonable people would find this rhetoric offputting, right?

"Yeah there's something big coming. You'll lose your job. Everyone will lose their job, there will be no more jobs. Deal with it, crybaby. Adapt. Not my problem if you can't prepare. Move fast and break things."

Losing your job is generally a traumatic and unpleasant experience for people. It means they stop making money, and they need money to buy things. Everyone losing their job at the same time because there are no more jobs sounds even worse.

Now, if you want to simply take a hardline Darwinian stance and say that the fittest will win out and life has always been permanent flux and that's simply that, then I wouldn't really have any problems with that. You would at least be internally consistent. But you're both a) saying things that have extremely offputting implications, and b) you're confused about why anyone would oppose your position, so you deduce that they must all be brainwashed cattle. Do you see the issue here?

while also being an imminent threat to all jobs which it will somehow be able to remove without actually providing any value.

But how do we determine how that value gets distributed, and to whom? How do people accumulate resources if they're unable to sell their labor? These are reasonable questions to ask, no?

I mean this in all sincerity and good faith, and not as a rhetorical question: are you hearing yourself talk right now? You do understand why many reasonable people would find this rhetoric offputting, right?

I'm at a disadvantage because I have to defend a complicated process that must exist in the real world from people in the peanut gallery and I'm not willing to lie about my own worries. There are problems and things we are going to have to address in the near future around this rapidly changing technology. I've got my eyes wide open and probably take existential risk more seriously than any of the detractors here. I advocate for getting UBI set up as basic civic infrastructure for if this job loss things starts coming at us fast. The critics here are making no such concessions to reality. They aren't expected to actually have an answer to the "what about the fact that china is going to build this anyways just a few year later?" question. All they do is throw peanuts and moan about problems that those of us who see the potential in this industry have been talking about for years and have produced volumes of discourse on that they refuse to read or engage with.

You'll lose your job. Everyone will lose their job, there will be no more jobs. Deal with it, crybaby.

Literally who the fuck is saying this? Every lab that thinks job displacement is coming has produced documents and spoken at length about how we must prepare for this and create a future where everyone can benefit from the abundance. There is a great deal of pressure to lie about this. What is your solution exactly? Can you please come down from the peanut gallery and actually talk about what we're going to do productively instead of engaging in this exhaustive discourse about the feelings of people who are acting like toddlers. I'll be honest, I've completely lost my patience with this discourse.

But how do we determine how that value gets distributed, and to whom? How do people accumulate resources if they're unable to sell their labor? These are reasonable questions to ask, no?

Yes, these are indeed good questions to ask. I don't object to people asking these questions. In fact I think they're very important things we should politically work out, the labs seem to broadly agree that this is the kind of discussion we should be having. I object to people lying about projects. Lying about projects and talking about what we're going to do if things go well and no one has to work anymore are very different types of behaviors.

This is a field where we are excelling, if America is going to right itself and bring about a new age of American prosperity this is it, this is the chance.

We have spent trillions of dollars on AI tech, not turned a profit, and not been able to build even a modest moat around the technology.

What if we had spent those trillions of dollars on literally anything else? We could have the best automated manufacturing in the world, or advanced medical technology, or any of a dozen other things that would bring actual prosperity to the nation as a whole.

You seem to be conflating 'we spent a ton of money on this' with 'we are excelling'. The whole point is that AI is a bad investment!

It takes more than a couple years to see the fruits of spending trillions of dollars. Building data centers takes actual years itself, and those data centers are used to do research which itself takes time. I would definitely argue that we're already seeing returns for previous investments and that the perspective that they haven't paid off is hopelessly short sighted. If we want to be great we have to be willing to make investments that take more than a quarter to pay off. I'm reminded of people who doubted amazon or uber because they invested aggressively rather than trying to immediately turn a profit.

China, whose one ruling party can have a plan spanning many decades, has the advantage in long term developments. Like when they decided 40 years ago to corner the rare earths market.

Aren't there, like, more practical ways for America to right itself? Like, I dunno, kicking out criminal illegal immigrants and purging the commies?

Would those actually result in America "righting itself", even ignoring the expected self-harm from the backlash due to the large number of Americans against those two things? Criminal illegal immigrants don't really seem like a problem that is big enough except psychologically, unless the illegal immigration itself is considered a sufficient crime, in which case kicking them out would probably kick the bottom out of agriculture and other menial labour heavy areas; purging the commies might entail destroying much of the tech industry and university research, because a lot of the most productive people there are about the most proudly self-identifying commies around.

"Sure, destroying the things I hate will cause a lot of disruption at first, but what will grow back afterwards will be so much better because it will not be encumbered by the things I hate" is a line of argumentation that has a bad enough track record (including, in particular, from assorted commies!) that one should not just accept it on faith.

Are you not expecting self-harm from the backlash due to the AI revolution?

More comments

AI is going to be a massive, world historical disruption--yes, even the Singularity--but that doesn't mean we're not in a bubble. Railroads, the Internet, canals, even arguably finance itself were born in bubbles: investors take huge losses, and society and technology continue on, faster than ever.

So, how to avoid being one of those investors getting reamed? Hell if I know. I've moved to more conservative investments myself, but that only very partially reduces my exposure. I'm expecting a pretty nasty correction in the next year that hurts all of the economy, government budgets, etc. Then, will buy things on fire sale.

It can be useful to divide the real world from the stock market when trying to assess bubbles. Both can be unhinged, but in different ways/degrees. The stock market variety of bubble can pop and be permanently gone. That doesn't mean the real world changes have stopped happening. The dotcom bubble pop didn't stop the internet from taking over the world.

I don't know if it'll be 2008-tier, but I'm expecting an ugly correction by the summer of 2027.

I've been expecting a correction since 2021 or so, but I keep being wrong. It feels like we go from one hype to the next without any response from the market when the hype doesn't pan out. Cryptocurrency was going to change everything. NFTs were going to change everything. The metaverse was going to change everything. 3d printing was going to change everything. Hyperloop was going to change everything. Full self driving was going to change everything by 2017!

Sitting here in 2026, I'm about to get in my car that I drive myself. I'm going to buy a new spatula with cash instead of 3d printing a replacement using feedstock that I bought with Bitcoin. The Hyperloop is an expensive joke, metaverse has been wound down amid record-setting layoffs, and NFTs are basically dead.

It seems like the market is fully decoupled from the economy at this point. Everything moves up and down on vibes, and the idea of "uncorrelated assets" shit the bed and died in 2022.

At this point the only solid indicators I have left are interest rates and debt, and both of them are looking increasingly dicey. A lot of very important loans (eg: SoftBank) will by due by the Summer of 2027, and debt collectors are not known for their understanding and gentle forebearance.

I have lived through both the dot com boom and global financial crisis. The claims of AI boosters are not identical to the people caught up in those manias, but it sure rhymes. I'm told that this time it's different. I'm told not to worry about troubling financial signals, because Line Goes Up, and it goes up more than enough to cover the risk. I'm sold an increasingly rosy picture of a hypothetical future mixed into a cocktail with a stark fear of being priced out forever... sorry, a stark fear of "becoming part of the permanent underclass". I was back in 2008 for a minute.

In both cases, I just kept my head down, kept saving, kept investing, and held a death grip on my job. It probably set me back in terms of total lifetime career prospects, but I was also never homeless or counting individual beans to see how long I could go before malnutrition kicked in either. I didn't buy a house until 2012, when I had saved enough to put down a full 20%.

Right now, I'm still contributing to my 401(k), Roth IRA, and HSA investments for the tax advantage, but I'm being more conservative in my taxable brokerage. I like Dimensional and Avantis a lot. The expense ratios are higher than pure passive choices like VTI, but they have a few additional filters that seem to help against the absolute worse of the current debt bonanza. I'm also increasing my bond and securitized debt exposure. It has its own risks, but debt is senior to equity when shit hits the fan. I'd get into metals, but I don't understand those markets well enough for it to be anything other than a gamble.

Beyond that, I'm trying to keep my spending down, and trying to make myself Highly Visible to management. It's pretty clear at this point that almost all layoffs are done on vibes, so if I can make myself part of the c-suite's emotional in-group, I think I have better odds. Hopefully I can speed run a fully funded retirement, then just keep working until I can't anymore.

It feels like we go from one hype to the next without any response from the market when the hype doesn't pan out.

If you zoom out a bit, you can also see cycles of overzealous investing in technology that was hyped to change everything, and eventually did, just not fast enough to keep the investments solvent. Home computing and video game companies crashed in the early 80s, but I'd be hard-pressed to argue that investors in early Apple or pre-IPO Microsoft were wrong about computers on every desk. Or Cisco investors in the late 1990s thinking about that this internet thing would sell lots of networking hardware. It's perfectly possible for the hype to be right without the expected profits following from it.

It feels like we go from one hype to the next without any response from the market when the hype doesn't pan out. Cryptocurrency was going to change everything. NFTs were going to change everything. The metaverse was going to change everything. 3d printing was going to change everything. Hyperloop was going to change everything. Full self driving was going to change everything by 2017!

Surely you can't really be surprised about any of this. What is the market supposed to do about NFTs not panning out? To what extent were public companies investing in monkey pngs with the hope of selling them to a greater fool? Why should the markets crash because some retards (technical term) were wrong about NFTs?

Except for the metaverse this goes for all of your examples. And Facebook stock took a 64% haircut in 2022 (the market overall fell only 33%) so it was hardly without consequences.

This sounds like:

It feels like we go from one hype to the next without any response from the market when the hype doesn't pan out. Flat billed caps were going to change everything. High rise ladies' pants were going to change everything. Low rise ladies' pants were going to change everything. Streetwear was going to change everything.

I'm told that this time it's different. I'm told not to worry about troubling financial signals, because Line Goes Up, and it goes up more than enough to cover the risk.

Oh, yeah. Same old song. I lived through the Celtic Tiger era in Ireland and that time too it was going to be different. This wasn't like the last times we had economic boosts following after slumps and then petering out and being followed by the next slump. This time the good times would last forever, this time we would keep on getting richer and richer, this time we were finally a modern economy with permanent high-skilled, high-paying jobs and we'd all be middle to upper-middle class and eating avocado toast as we commuted to our office jobs in the Big City forever.

That it was undergirded by the construction boom, and that the subsequent property bubble inevitably burst, was handwaved away as pessimism. And then it all came tumbling down, triggered by the 2008 US crisis with its own property bubble and mortgage lending, but the house of cards at home was exposed as built on sand. Enter the austerity years, begging for bailouts, and the return of the emigrant ship as the solution to "too many people at home, not enough jobs". This time had not been different, after all.

The US economy is generally robust and may not suffer as much during a downturn or even a crash, but the ripple effects will be felt globally.

Is the Irish economy that bad? I thought they were still doing pretty decently in the context of EU countries, albeit with an American leash and collar.

We're much too over-dependent on the tax take from the multinationals, and for the employment in Big Tech. When the likes of Meta do global layoffs, that hits us too, and there isn't really the slack for "okay all the laid-off software engineers can instead go work for the other IT companies" because they're laying off or not hiring, too.

The government is also looking to borrow to keep programmes going, instead of socking away the surplus for a rainy day as had originally been the plan, as well as the usual budget over-runs (the health service is a constant black hole of sucking up money and running out and needing more):

IFAC chair Seamus Coffey said: "The key concern is that corporation tax that was planned to be put in these funds is actually being spent."

He added €5 out of every €6 which is collected by the State in corporation tax is now being spent.

The warnings come as concerns rise that Ireland is becoming increasingly dependent on volatile corporation tax payments.

The top taxpayers have been pharmaceutical company Eli Lilly and technology giants Microsoft and Apple.

The three groups have been responsible for half of Ireland’s corporation tax payments.

...He noted that the Irish economy is performing very, very well at the current time.

"Unemployment is low, wages are at least at the moment above inflation," he said.

"So the Government is stimulating the economy now when it doesn't need it. The risk is that we won't have the resources available in a downturn we might need," he cautioned.

"We have a bad history with this over 40, 50 years of spending the money when it comes in during the good times and then leaving ourselves exposed when the downturn hits," he said.

He said that when we talk about declining surpluses, we are in a very different position to other EU member states.

"The issue is that we're maybe not leaving ourselves sufficiently insulated against what might be coming down the track," he said.

"So, we must understand the Irish economy is in a strong position. When we talk about the public finances, we're talking about declining surpluses," he said.

"We're not borrowing money to run the government on a day-to-day basis. We are talking about borrowing money to save, which is maybe an unusual position to be in," he added.

We're not in crisis yet, but a bad day for the USA and/or the global economy is going to hit us hard, and the one lesson we should have learned is that the rainy day always comes:

A Government analysis on the impact of Donald Trump's proposed tariffs warned of a decline in GDP of up to 4% and a decline in employment by more than 3%.

The research modelled multiple possibilities including unilateral tariffs, a trade war with the European Union, or a wider trade war.

It said every single scenario would have "significant negative implications" for the Irish economy.

The unpublished analysis said GDP could be expected to fall by between 2.75% and 4% over the medium term.

The employment decline would be between 2.5% and 3.25% while domestic demand would drop between 1.75% and 2.5%.

It also warned of price level increases of between 2-3% and said the modelling may underestimate the actual impact.

The government is also looking to borrow to keep programmes going, instead of socking away the surplus for a rainy day as had originally been the plan, as well as the usual budget over-runs (the health service is a constant black hole of sucking up money and running out and needing more):

In other news, Captain Renault is shocked-shocked!-to find that gambling is going on in here!

Irish politicians boom-and-bust spending and borrowing to buy votes? I cannot believe it! Even though I've seen this exactly playing out since the late 70s when I became old enough to understand what was going on and then old enough to vote in the 80s.

You should see the saga of our new National Children's Hospital. Will be complete God knows when, will cost God knows how much, and the builders claim it's not their fault at the same time they're sending record profits home to their parent company. Claimed it would be open in 2026? That's not going to happen.

Almost all US states manage to keep "rainy day funds", although the sizes and average balances (per capita) range quite a bit. IIRC it's not quite a typical partisan divide on which are which.

Add on "Prediction markets are going to change EVERYTHING" (mostly used for sports gambling, unfortunately), the Internet of Things is going to change EVERYTHING (I will now spend EXTRA money for a toaster that isn't wifi connected), OMG we might have found a room-temperature superconductor, VR/AR tech is going to blow your mind, Crypto hype moved from Blockchain hype, to DApp Hype, to DAO Hype, to NFT Hype, to Stablecoin Hype, and it was all illusory.

WHAT THE FUCK WAS WEB3?

Its funny, I MOSTLY could spot when a trend was more hype than real.

I sold all my crypto in 2019. Never bought an NFT. I thought Metaverse would become something but nobody had a good vision for it, 3D printing is indeed a cool, useful tech but don't bet the farm on it.

But I feel sketchy now because all those hype cycles leave you suspicious if the next big thing is actually arriving or there's just people who make a lot of money if you believe it is.

SpaceX seems to be real, but maybe still overhyped.

GLPs ARE the real deal, and there's some other neat drugs in the pipeline.

Smartphones continue to be a boon... if you can avoid all the gamified apps designed to suck your attention and money.

Quadcopter tech is pretty damn mature now, and is finally achieving some market penetration.

Right now, I'm still contributing to my 401(k), Roth IRA, and HSA investments for the tax advantage, but I'm being more conservative in my taxable brokerage.

It feels absurd to focus on that right now, when the way things are going we're either going to get crash and hyperinflation wiping out a ton of that nest egg, or a new industrial revolution that goes exponential and so even modest market holdings will make you wealthy over the next 10 years.

The ONE thing that doesn't seem likely to happen is just steady 7% growth in the market (on average) over the next 25 years.

I'm old enough to remember the first VR hype in the early nineties. They made it sound like totally realistic VR sex with partners you created yourself, was only two or three years away. Seems like that isn't even on the farthest horizon now.

Around 5 years ago, I saw part of a huge tech webinar where a woman, presumably SF based, was talking in a very emotionally charged tone about how she would be really down on the world and its future if not for the promise of Web3 and its salvationist glory. I had no idea what that could possibly mean at the time. What difference could it make to real lives? Was I missing something important? Time would tell, I figured.

I don't think I was missing much. The promise of web3 was "decentralization" and getting away from the control of the biggest tech companies by using blockchain networks to build sites. But the companies operating the stablecoins and whatever, were mega companies like Coinbase. And the movement was marked by really low ease of use (complex seed phrases, "gas fees," etc.), and massive grift. Much like nfts, a few early ponzi schemers made obscene sums while many suffered.

But the companies operating the stablecoins and whatever, were mega companies like Coinbase.

Or they were inherently unstable. RIP Luna.

Its interesting that while the fundamental tech behind Bitcoin has proven utterly sound, the inability of the tech to compensate for the human/cultural element of the community we build around it has made it less effective at its generally intended purpose. I mean, it still functions as a currency, yet that's arguably the least common use case for it these days.

Either way, there was absolutely a tipping point when the core operators in the environment stopped being a majority of those who understood and supported the technology and got outnumbered by those who were about the price and keeping it pumped even at the expense of the tech, or just straight up scamming.

It feels absurd to focus on that right now

What else am I going to do with it? My house is just about paid off. I've got several months worth of food and water. My car is in good shape. I'm pretty well set for guns and guitars.

Hookers and blow aren't really an option either. I don't do drugs, and I live out in the sticks where the quality of prostitute is so low that I'd pay them to not fuck me.

Charitable giving? Precious metals? More expensive food?

I already donate to my local homeless shelter and animal shelter. I guess I could do more.

For now I'm just saving like crazy and trying to speed run retirement.

Solid question, in that case.

Could do what I did. Buy a classic muscle car. Gives you something to work on and fun as hell to drive, especially on backroads.

LARP as a moonshiner.

To whitepillers: is there an argument for why I'm wrong

Whitepill, I would say I recheck this graph every so often to see if this ratio starts going up again. (Blue is all aggregate USD debt/gdp, red & green are more usefully decomposing that into private (at least non-federal) debt and federal government debt, and the bold black line is the ratio of red:green). Maybe you could slice the sectors up a bit more finely between households, firms, local/state governments, foreigners, etc., but I think this one pretty much gets what it's trying to capture.

Basically I think many agree that a big caution point for an economic crash/collapse is ballooning private debt. Most analysts now recognize there is just a fundamental difference between that and 'public' debt from the government that issues the currency, and that public debt is at least not as bad. I'd go much farther and say that by the accounting, the national government is more akin to the overall economic score-keeper, and that it would be rather goofy to think you're tracking something negative at all when counting up everyone's total outstanding scores and calling that 'public debt'. The national public debt is essentially our net money supply (money can be increased by other private debt, but not in net); our financial wealth.

So then, tracking private debt to gdp (red line) is some kind of a 'debt to income' ratio term, while private debt to public debt (black line) is a kind of 'debt to wealth' ratio. So I see the red line still steadily coming down since 2008's high, and the black line falling off precipitously after it rose to a huge level in 07/08. I've posted this a few times before, but in the US, our only 6 economic depressions were each immediately preceded by the only 6 multi-year significant reductions in the outstanding government debt. The 7th most significant government surplus run was less dramatic, in the late 90s, and directly coincided with the private sector going much more into debt, causing that bold black ratio to spike, running up until the 08 crash when the private sector forcefully deleveraged, also forcing the government hard back into deficit. So the intuition that government debt is destabilizing or that government surpluses are a good thing, that's what I see as dangerous. If we start getting more politicians in places of influence who think 'quadrillion' is one syllable or zero too far for economic well-being, and who start pushing for austerity and surpluses, that's when I'm getting back into prepper mode.

As others have mentioned though, none of this necessarily has bearing on all stock-market bubbles or corrections. It's just to say that 2008 was a different beast, and I don't personally see any table-setting for that kind of financial crash & great recession/depression angle. I'm still plowing each monthly paycheck into the broad market index and trusting the longterm, while kind of wishing it will dip long enough at some point for me to harvest some on-paper losses to pay less taxes.

I’ve been feeling an oncoming crash for years, and it hasn’t paid off yet. From my COMPLETELY UNSCIENTIFIC, VIBES-ONLY perspective, the market has absorbed all sorts of rate changes and price spikes and geopolitical boondoggles which might have reasonably been expected to presage a crash. But then, if it was obvious, it would be priced in, right?

Speaking of which. The big AI companies are absurdly valued, which is not the same as being incorrectly valued. Pricing the possibility of a market-shattering payoff has historically caused markets to diverge from normal human expectations. This isn’t great, but it’s not inherently disastrous in the same way as, say, the subprime mortgage situation.

It only ends up a bubble if AI plateaus, stops picking up new niches, before debt outpaces investment. This is entirely possible and I don’t know what the leading indicators would be.

If AI expands into any other sector of the economy, it pays off. I suspect this would require a major change in robotics (to cover skilled trades, etc.) or ethics (to allow violence). Some of these routes are extremely goddamn dystopian, but at least they aren’t bubbles, right?

That said, AI isn’t the only source of optimistic IPOs. There is a lot of interest in manufacturing and defense investment. Onshoring stuff which might become contentious. I could believe that this is its own bubble; people want to believe that manufacturing will spring up. If they are underestimating the cost of local improvements, or overestimating the reliability of the current administration, there’s potential for some serious disappointment.

Add this to the list of things I hope I'm wrong about, because if we get a proper crash, the political fallout is going to be massive.

Ha. Norm MacDonald continues to be relevant.

Besides, if the techbros are revealed as delusional, if all the slush funds and gold plating in the country can’t stave off a crash…what are we supposed to think? Sorry, Trump was just holding the hot potato, blame the wreckers?

Ha. Norm MacDonald continues to be relevant.

I don't think that's fair. I'm not trying to make you feel sympathy for Trump, I'm trying to say what comes after him will be worse. See for example the thread after Labour won the most recent election in the UK, and compare to how eit panned out.

what are we supposed to think? Sorry, Trump was just holding the hot potato, blame the wreckers?

The busness cycle is usually beyond any president, so partially yes, but it's clear he also made many unforced errors here.

I don’t know what the leading indicators would be.

The Schiller P/E ratio has me spooked. It's over 40. I also don't love that Google has gone from using cash flow, to issuing debt, to issuing equity to fund build outs. Furthermore, I'm not fond of current private credit default rates right now. They're at a high water mark since we started tracking them after 2008.

I suspect this would require a major change in robotics (to cover skilled trades, etc.)

Have a look at what Hyundai is doing with shipbuilding and robotic welders.

Can't forget that AI is at present unlocking a lot of rapid improvement in the biology/pharma realm.

Even if we don't get true immortality-type stuff, keeping people healthy and youthful for longer is a massive productivity boom in and of itself. Curing obesity ALONE is trillion-dollar gain territory.

I've been feeling we are due for a crash since around 2018 basically. What do I know...

I’m fond of the argument that the 2018 correction represented the ‘natural’ end of both the post-2009 asset pricing boom and the broader (and obviously intrinsically related) credit cycle. That would be a relatively standard timeline for that kind of thing.

The response from the Fed (under Trump) but also other central banks was to supercharge the asset pricing bubble, cut or reverse planned hikes aggressively, and reinflate the bubble. Then came Covid and infinite money printing, which accelerated the real inflation that had been avoided through most of the QE era by falling prices for some goods that offset service price increases, plus slowing velocity of money and finally some rate hikes when they became impossible to avoid.

So we’ve been in a weird liminal era for almost a decade now, and it isn’t really clear how it’s going to end. I don’t think it’s viable for the American government to allow markets to experience a prolonged correction. A short, sharp, quickly-recovered-from 30% drop? Sure. But any longer, or any deeper, and big public sector retirement funds, 401ks and so on suffer to a degree no politician can survive. Thats different to previous generations in many ways. At the same time, you can arguably only fight gravity for so long.

"More money has been lost in anticipation of corrections than in the corrections themselves. " Or something to that effect.

I've felt that way for about 10 years now, so man I don't trust my judgment on that.

I do think we've drawn out a lot of 'slack' we might have had in the system, between Covid, the Ukraine War, Europe's general implosion, and now the Hormuz/Iran situation. Billions of dollars of fraudulent welfare payments don't exactly help, but at least they're an economic stimulant in the Keynesian view.

So a significant enough shock could definitely send us for a tumble.

For my money, it'll be a serious collapse in global trade that would be the catalyst.

ON THE OTHER HAND, we are energy-rich, and if we build out more nuclear there's virtually no mere economic shock that can send us into a long-term depression.

So I think its fair to keep the debate mostly in the range of "are we due for a market correction that will be relatively short-lived, or a long term depression where growth stagnates for want of any truly productive industries to invest in and an aging population that demands increased consumption spending."

I bought a house in 2016, thinking there was a 50% chance I'd be under water in it.

I did something similar with my first home. Thinking the housing market was overheated and there was a decent chance I'd end up selling it at a loss in order to move for work.

I was of course incredibly wrong.

LOL I bet we're close in age.

The 2008 crash left an indelible mark on me. Doubly so because I was in Florida, a major epicenter for the worst of it.

I viewed the increasing house prices post 2013 as a TRAP. But I was convinced to buy in 2019, and I am still fully mentally prepared for it to crash down to that level or below again.

I still do sincerely think that the stock market is overvalued (maybe specific companies are not). But if you aren't willing to put at least SOME of your long-term money at risk in it, you're objectively missing out. And the U.S. is the only country where the insane growth might be justifiable.

Ironically, owning the house makes me feel significantly more secure because even if I lost my job and my retirement accounts dropped by 50% tomorrow, I'm guaranteed to have a roof over my head for the next year (foreclosures take a LONNNNNNG time to process) which should give me time to figure something out.

The value of increasing white collar productivity by 10% permanently would be worth every penny spent on AI. It doesn't have to singularity to be worth it. There are nonobvious questions of how to capture the wealth as a company, but the basic "total addressable market" sanity check passes. TAM = 10% total white collar labor expense (well, value of marginal 10% of labor, but similar).

Typical GDP growth in the US is about 2%/year. That means just waiting 4 years doing normal stuff gets you ~10% productivity improvement. ChatGPT was released just about 4 years ago.

There's a lot of subtleties in the economic figures. But my back-of-the-napkin math above argues that we would have had this 10% permanent productivity gain without any investment in AI.

The release of ChatGPT was just the wake-up call that something was coming. No one used it back then for more than a trivial amount of real work. People will vary, but I'd put the dividing line for that about 9 months ago. It will take time to diffuse and for processes to adapt. So such conclusions seem premature.

Also, that 2% growth isn't some magic rule of nature, excepting population growth (which is covers about 1/4 of the economic growth over the last 4 years) the default growth is 0% (as it was epsilon close to for much of history) only by investing and inventing things does growth happen. AI is a thing human invented and are investing in, it's quite plausibly the thing that will keep the 2% going another couple of decades. We don't get to do nothing and magically get 2% growth as a baseline and AI is only measured as to what it adds on top of that. Computers, cell phones, internet, cloud services, etc. all of that was apparently worthless as we only grew 2% a year.

Reminds me of straight lines on a graph. More specifically:

If [societal controls driving steady improvements] were true, you wouldn’t see the effects of pollution-busting technologies on pollution. You’d see them on everything else. For example, suppose that absent any other progress on air pollution, politicians would regulate cars harder, and that’s what would make air pollution go down by 2% that year. In that case, the effects of inventing an unexpected new pollution-busting technology wouldn’t appear in pollution levels, they would appear in car prices.

If the productivity-boosting effects of AI were true but society controls the overall trend to a steady 2% growth rate, you could see the effects of AI on everything but productivity. Dark Leisure is one attempt at explaining it away.

Wasn't the "basic sanity check" for the Internet back in the 90's also passing? It's entirely possible to land on a piece of viable technology, and still crash the economy by investing in it too much and too early.

Right, that's the point of a basic sanity check. It's an easy check to rule out things that are insane, in this case "if this has a small TAM, it is not worth a large investment." AI does not have a small TAM, so this does not show that it is not worth a large investment.

But, as you said, it also doesn't show it is worth a large investment. It's a "fast rule-out" heuristic, not any kind of a "rule-in" one.

"Crash" means a very different thing if its a sharp correction that recovers inside a couple years, vs. getting stuck in a trough of low productivity growth.

It's entirely possible to land on a piece of viable technology, and still crash the economy by investing in it too much and too early.

Or consider railroads, a century prior.

This is actually my nuanced, enlightened centrist take: AI as a technology—or more precisely, a suite of technologies—is potentially as revolutionary as the internet, electricity, and steam power rolled into one, but current investments in AI may very well turn out to be myopic or premature. For one entirely plausible example, we may end up in a world where distilled models git gud and thus inference is mostly done locally on cheap, commodity hardware like smartphones, which would render the hundreds of billions being poured into data centers largely wasted.

I'm unironically rooting for the outcome where AI intelligence caps out below the super-intelligence level, but nonetheless gets more efficient and cheap, and distilled models become the general standard for everyday use. Like, your self-driving car will have one built in, your house will have one, and the absolute top tier models are mostly only used by Governments and large corpos for the highest level tasks.

Distributed, cheap artificial intelligence would be a boon, and without so much existential risk.

even more gay race communism now

At this point I would prefer to be ruled by the pink haired feminist communists than the AI techbros.

"They're the same picture."

I kid a bit. But not very much. Died hair, pride flags on chat flairs and hung on a couple walls, way more transwomen than you'd think, ICE warnings on group chats. We got it all in my office building. I'm not an AI techbro but I work for a major tech company.

I'd still take the AI techbros, you're likely looking back to the feminist communists with rose tinted glasses.

“Of all tyrannies, a tyranny sincerely exercised for the good of its victims may be the most oppressive. It would be better to live under robber barons than under omnipotent moral busybodies. The robber baron's cruelty may sometimes sleep, his cupidity may at some point be satiated; but those who torment us for our own good will torment us without end for they do so with the approval of their own conscience.” -- CS Lewis

I always wondered if Lewis considered the ultimate application that quote automatically suggests, being a card-carrying believer in the most supremely omnipotent moral busybody of all. I feel like the benevolent deity who fails to understand the wants and needs of His subjects is a somewhat exhausted trope now, but was it already back then?

I feel like the benevolent deity who fails to understand the wants and needs of His subjects

You are totally free to go to Hell if that's what you want. He won't turn off your brain and make you a drone that can only obey the rules. "We want what is bad for us" is also an exhausted trope, but still applicable. "Dear God, please make it okay that I want sterile sex with robots instead of marrying one wife and having children with her, that's not too much to ask now is it? And I would feel so much better if that term "sin" only applied to really bad people, like Trump voters, because it crushes my valuable self-esteem to be told I'm sinful too".

Look at all the talk on here about curing obesity, because obesity is a bad thing that costs the rest of us money. Not too much "understanding of the wants and needs" of the sinful there, just "they are a problem, we must fix that problem". Why, it's almost like society is a rule-imposing deity!

You are totally free to go to Hell if that's what you want.

Which version of it, the "separation from God" one or the "fire and brimstone" one? If it's the latter, "hey, I'm not forcing you to do anything, I will just torture you for{ever, _aeons} if you don't obey" doesn't register as a particularly compelling vision of freedom. I know that especially Evangelical Christians would consider "God is evil" to be very nearly an oxymoronic statement, but if their God created me, why did He equip me with a moral compass that says it?

If you want separation, you can have separation. I do find it intriguing that the complaint there by some atheists seems to be rooted in "well why didn't god make me do it, why didn't they make me be good and set me up in such a way that I couldn't do bad things" because that seems to clash with "we want perfect freedom, no more gods needed or wanted, humanity is sufficient for itself" views around atheism.

If you don't believe god exists, hell doesn't exist, death is the end - then why cry and carry on over the non-existent? It would be like me complaining I never got my share of the gold at the end of the rainbow. God is so mean! He doesn't exist, but He's still so mean! Why isn't He Santa Claus in the sky, giving me everything I want without consequences or responsibility on my part? Well when we get FairyGodmother post-Singularity AI that makes us all immortal and rich and transhuman and colonising the entire universe so we all have our own solar systems and can get anything we want for nothing with no effort on our own parts, that'll show Him!

I think this is misrepresenting the complaint (at least the one I implicitly voiced here). The problem is not along the lines of "God wants me to be good, but I'm evil and can't help it, why did God make me this way?", and I can only see this framing emerging as a Christian (mis)interpretation of a mindset that is alien to them. Rather, it takes the form that the Christian God, as described in Christianity, is evil from the Atheist's perspective, and moreover is claimed to command everyone to do evil (as understood by the non-Christian) to earn his approval. It doesn't seem to me like the Christians I interact with (actively or passively) ever have a useful response to this, only being able to say that God is definitionally good and so the degree to which His purported nature disagrees with my understanding of the Good just is prima facie evidence of the latter being mistaken.

I've seen enough atheists going on about how they would have totally created a much better universe not to think that there isn't a Christian interpretation lurking there. It's not all sweet scientific reasoning and logic, there's "I want the things I want to be guilt-free and cost-free".

I know that especially Evangelical Christians would consider "God is evil" to be very nearly an oxymoronic statement, but if their God created me, why did He equip me with a moral compass that says it?

Do you believe that human moral compasses are 100% reliable and accurate? If they are not 100% reliable and accurate, would you agree that one of the ways we observe them losing accuracy is when people double-down or psychologically entrench on a bad moral choice?

Separation from God has always struck me as a cheat too. You never give informed consent to separation from God; you are just told "doing X automatically separates you from God, but since it's your own action, you don't get to blame anyone for it".

As the wokesters and anti-wokesters like to say (in different contexts), God doesn't owe you his presence.

If being separated from God is eternal torture, and if he created you it is equivalent to the abortion issue except you are undeniably a person. In which case yes, he does owe you his presence.

You never give informed consent to separation from God

Alternatively, "informed consent" doesn't work the way you appear to be arguing it does. Does a 3-pack-a-day smoker give informed consent to the consequences of smoking three packs a day? Obviously not, since they didn't know for certain what the consequences would be or what the subjective experience would actually be like, right?

We've got a variety of fairly reliable evidence that smoking causes cancer. The evidence that unrepentant sin sends you to hell is, let's say, more on the hearsay side of things. If the best evidence for smoking causing cancer is that somebody said it totally does, it would be a different story.

More comments

If I only got told about lung cancer patients, but could never see one even theoretically because the only way you can look inside a cancer ward is to become a patient, I might think "smoking kills" was bullshit, too.

If this was an old forum that quote would be in my signature.

You can always add a flair under settings. Of course, people will only see it if they click on your profile, and you’re limited to 100 characters, but you could at least remind some people that “It would be better to live under robber barons than under omnipotent moral busybodies.”

You can always add a flair under settings. Of course, people will only see it if they click on your profile, and you’re limited to 100 characters

You're mixing up two different features.

  • Your flair shows up next to your username at all times, but is limited to 100 characters.

  • Your bio has a much higher limit of 1500 characters, but shows up only when someone clicks on your username.

@zoink

Thanks! You were actually the one I had in mind as I wrote that, as I have always found your flair quite memorable.

I can imagine a AI post scarcity possibility. The communists will just never work if they ever get meaningful power.

AI bros also have lower downside, though. Communists winning would likely only result in deaths in the order of 10s-100s of millions, maybe a billion or two, but likely not more. AI bros winning could mean literally every last human dead, or perhaps even worse, their consciousness stuck in eternity in a hell simulator.

I think AI bros still win out in terms of expected value, though.

aren't we scrathcing St. Petersburg paradox territory here?

Is that a realistic outcome? I could say that Communists winning could mean nuclear bombs in space and sun death and the blow-up of the entire galaxy. Maybe Communists winning means all our consciousnesses stuck in eternity in a hell simulator, plus one.

Communism is a hell simulator, but at least you get to die (sometimes pretty quickly).

I'm pretty bad at making predictions, especially about the future, but a lot of people in AI seem to think that the odds of AI development leading to human extinction is at least in single digit percentage odds, which is high enough to be concerning. Communists could likely build rockets and simulators, but I'm skeptical that such tech progress is likely in a Communist-run world before we're brought back to some post-apocalypse situation or just boring stasis in some barely tolerable dystopia. Certainly I'd bet that inventing immortal consciousness that can suffer for eternity seems far more likely in an AI bro world than a communist one. In the latter, the suffering would only last about a lifetime, and in the real world.

A hard nope from me, but yeah, it's a South Park-esque douche vs turd contest.

Does anyone else feel like we're heading for a good old-fashioned, 2008-tier, financial crash?

Yes. Maybe not exactly like 2008, but the economy seems over-heated (don't talk about Ireland and how we're increasingly over-reliant on multinationals tax take plus our fiscal prudence has gone out the window with prospects of borrowing more money to pay for government vote-buying programmes), I don't trust the splurging on AI being the thing making line go up, and there's enough discontent about "my grocery basket is X times more expensive" even with all the corrective think pieces about ackshully we have never been so rich and well-off, unemployment is low, look at economy how line go up.

Throw in all the wars driving up energy prices and we're due a landing, and a hard one, soon.

Trying to time shorts/market reversals is effectively impossible. Essentially the entire market structure is doing its best to pump at all times, and will happily keep running on air roadrunner-style until the occasional straw that actually breaks the proverbial back turns out to be something absolutely random.

Just make sure you're controlling your spending, keeping a decent amount of savings in relatively durable spots and don't overextend too hard.

To whitepillers: is there an argument for why I'm wrong that doesn't boil down to "you don't get it, chud, it's the New Economy! The Singularity is just around the corner! All the rules are obsolete!"? This argument is verboten, because this is pretty much what people say with every bubble.

Trump admin is engineering a massive restructuring of global capital with the intent of making America richer. It runs a spectrum from factories and reshoring to taking the oil from Venezuela. DEI Disparate Impact theories are being unwound so that productive talent that was locked up by politics will no longer be locked up. Welfare migrants are being deported and America's balance sheet is shifting toward productive economic activity again.

The global future looks very mixed as population decline will hit the entire world. Fewer people means less capital to spend and fewer things to spend capital on. Bad news if you're China or Russia or Europe or Brazil. But this will make America even more attractive for capital and investment, relative to everywhere else. We will have factories, oil, gas, tech, services, military strength, capital, youth, growth, everything.

The market is already starting to price in this future. AI and datacenters are the headline boom feeding a lot of optimism, and it might come crashing down. But the underlying economic picture is very potentially strong because the fundamentals of the American economy are stable, and getting stronger. And, to my mind, AI unlocks so many economic opportunities that it will find a purpose eventually. If it does come crashing down it will be more like the dot-com bubble than full 2008. Because a technology that transforms electricity into intellectual work is enormously productive in real terms, not just fake email job terms. The dot-com bubble killed billions of dollars of imagined wealth locked up in companies like Pets.com. But, fundamentally, there was nothing Pets.com was doing that isn't being done today. They were just ahead of the curve.

Besides, despite all evidence to the contrary, the government has learned from some of its mistakes, and I don't think they'll repeat the dot-com bubble fiasco. Washing and Bessent and DOD and Trump will funnel billions of dollars in government contracts to the AI companies, and that will keep everybody afloat. The consequences will be an entirely new set of problems relating to economic management out of Washington central offices as the greatest economic change in the means of production since the Managerial Revolution plays out. It won't even be entirely a good set of changes. But they will probably fuel a lot of medium-term growth in the terms we're talking about here, and the negative consequences will be something we will spend the rest of our lives disentangling.

Not really seeing it. The AI boom could fail, but I wouldn't expect a 2008-tier crash from it. The datacenters which were built will still be there and still be useful for 'conventional' computing. So more like the dot-com bust which left a bunch of dark fiber which became the foundations for a new boom.

So more like the dot-com bust which left a bunch of dark fiber which became the foundations for a new boom.

I'm not sure if this is a great comparison. Fiber can sit in the dirt for 25 years and light up just fine.

Even in an optimal operating environment, I don't think we'll get nearly that long out of a modern data center GPU.

If things crashed and frontier models and data centers with their chips and hardware were effectively free I can't imagine someone couldn't productively use it at greater then the operating cost. Anthropic is paying $12.5 billion a year for Collosus 1 and best estimates I've seen is their inference margin for it will be >50% so they will sell that for >$25 billion / yr. It takes about $200 million / year for electricity to run Colosus 1.

The datacenters which were built will still be there and still be useful for 'conventional' computing.

Isn't the specialized AI hardware being rolled out considered inefficient for general purpose computing, though?

Ironically, the frustrating pace of permitting and construction means the amount of overcapacity will be limited. An AI bust which clears out the dead wood will end up with a few large winners who will still be able to make productive use out of all the AI hardware they can get their hands on.

Most of it is basically GPUs with a lot of memory, and good for any massively parallel problem. Google's TPUs are more specialized, but I wouldn't be surprised if they could use them for something else.

The first generation TPUs were for Search and Ads, bread and butter moneymakers.

They were used for a lot of things, those among them. But their basic operation was a matrix multiply followed by a LUT, a lot more restricted than GPUs.

Depends on what you mean by "conventional". Just imagine the gaming rig you can build with that.

Aren't most of the high-end AI chips a bit unwieldy to build gaming rigs with? And I feel like the same people who would want to build a massive gaming rig would also not want to stream their game from a server in the middle of nowhere somewhere.

Although now that you mention it, perhaps the Sea Power fanatics would probably accept a little bit of lag if it could let them play through "The Dance of the Vampires" from Red Storm Rising...

There's a funny world where, in a couple years, ultra-cheap used AI chips make local game streaming a viable option. The software already exists and does a great job with a fairly arbitrary setup, so if you wanted to have a server on one side of the house supporting an ultra-thin client on the other, it mostly just needs someone selling it as a product.

But the limitations to AI 'GPUs' are more broad than just the form factor or output slots, but down to the architecture, power focus, and supported pathways. The economics don't make a ton of sense even if the lag can be kept to sub-frame periods.

Specifically, a lot of the GPUs designed for AI workloads don't even have video outputs, so they wouldn't work in your own personal gaming rig.

At some point GPUs aren't even GPUs any more, they're something else.

Wild to think that the lineage of the graphics accelerators with weird aliens on the box is now utterly revolutionizing the world.

It is still surreal to me that Nvidia went from a niche company that only true PC Gamer hobbyists 'cared' about, to like somewhere around the top 10 most important companies in history.

It'd be like, I dunno, Warhammer 40k became the most popular competitive sport on the planet, with top players commanding tens of millions of dollars in salary.

My first GPU was a PNY Verto (something older than the 6600 in that gallery, though I don't recall the exact model number). That really brought back some nostalgia haha

The datacenters which were built will still be there and still be useful for 'conventional' computing.

I think that was the case for every bubble investment. The factories, houses, shitty dotcom-era websites were also still there and available, the question is whether they can justify their cost. The wasted capital usually gets picked up eventually, unless the investment is something truly retarded like tulips.

Tulip Mania was more or less a wild exaggeration of the truth.

The houses often rotted(literally) after being left half finished.

It was largely not true in the housing crash associated with the GFC; many of the half-finished exurbs ended up simply being destroyed (before or after being scavenged for parts). And the risk-aversion it generated, along with allowing the anti-growthers to take political control, depresses housing to this day. I would not expect an AI crash to do this.

The Netherlands had world-leading systems of farming and flood-control in the Tulip areas which may have been atleast somewhat encouraged by the bubble.

What is this place?

This website is a place for people who want to move past shady thinking and test their ideas in a court of people who don't all share the same biases. Our goal is to optimize for light, not heat; this is a group effort, and all commentators are asked to do their part.

The weekly Culture War threads host the most controversial topics and are the most visible aspect of The Motte. However, many other topics are appropriate here. We encourage people to post anything related to science, politics, or philosophy; if in doubt, post!

Check out The Vault for an archive of old quality posts. You are encouraged to crosspost these elsewhere.


Why are you called The Motte?

A motte is a stone keep on a raised earthwork common in early medieval fortifications. More pertinently, it's an element in a rhetorical move called a "Motte-and-Bailey", originally identified by philosopher Nicholas Shackel. It describes the tendency in discourse for people to move from a controversial but high value claim to a defensible but less exciting one upon any resistance to the former. He likens this to the medieval fortification, where a desirable land (the bailey) is abandoned when in danger for the more easily defended motte. In Shackel's words, "The Motte represents the defensible but undesired propositions to which one retreats when hard pressed."

On The Motte, always attempt to remain inside your defensible territory, even if you are not being pressed.


New post guidelines

If you're posting something that isn't related to the culture war, we encourage you to post a thread for it. A submission statement is highly appreciated, but isn't necessary for text posts or links to largely-text posts such as blogs or news articles; if we're unsure of the value of your post, we might remove it until you add a submission statement. A submission statement is required for non-text sources (videos, podcasts, images).

Culture war posts go in the culture war thread; all links must either include a submission statement or significant commentary. Bare links without those will be removed.

If in doubt, please post it!


Rules


Recommended Posts And Communities

Recommended Realtime Chats