site banner
birb_cromble  4d ago (text post)   545 thread views

Weekly Finance Thread 2026-06-06

A weekly thread to discuss financial matters - from personal all the way up to global.

Ground Rules

  • Remember that we're all just Internet randos. Don't bet your life savings on a hot tip from this thread.
  • Keep culture war in the culture war thread. Yes, global events may impact our personal finances, but that does not mean we have to incessantly harp on culture war aspects here. If you are going to discuss it, please stick to the practical impacts of it on an individual level.
  • Be kind. Remember that everyone here comes from different circumstances. We all have different resources available and different risk tolerances.
  • Don't let the perfect be the enemy of the good. Better is better. Celebrate people when they take a step up and work to move their finances in the right direction. Don't flame out because they haven't followed what you consider the optimal path. Everybody has to start somewhere.
1
Top Bottom Old Controversial Comments
Jump in the discussion.

No email address required.

There is a certain investment thesis that I see increasingly on LW / AI Twitter / adjacent spaces online. Let me summarize it:

  • If AI destroys the world, or someone uses it to, money doesn’t matter and probably even the most hardcore preppers are toast. There is no use preparing for this scenario other than maybe bringing forward some bucket list items if you’re a Yud level doomer.

  • If AI leads to some incredible abundant utopia with a FOOM / hard takeoff scenario, money won’t matter and we’ll all be trillionaire emperors of our respective limitlessly abundant space kingdoms and/or VR wirehead heavens. Maybe Elon Musk will become some overlord, but the average equity investor won’t know or care.

  • If things don’t change, and/or AI gets retrofitted onto the existing economy in a fake scenario to preserve social and economic status stratification and/or private ownership of property / means of production, people who own big tech / AI companies will be best off. Everyone else will either be a UBI peasant at best or Elysium / Manna underclass at worst.

If the realistic possibility space consists of solely these scenarios, the only logical investment strategy unless you’re retired or retiring in under 5 years is to YOLO everything on AI/tech/etc stocks. But that is a big if. This thesis is especially enjoyed by tech industry workers who argue that either AI will replace them, in which case this must make money, or it won’t, in which case at least they still have a high paying jobs.

I don’t think it takes a genius to see why this is more than a little flawed.

I think Point 1 really obscures a lot of bad AI outcomes by cloaking them in the most dramatic one. It is much, much more likely that AI does (or is used to do) something "mundane" like a superengineered virus or destroying the internet. (The latter may already be happening - it won't be a one-time event, just a slow but steady string of incidents that make the internet as we know it insecure and see the migration of core functions away from it.)

Lumping all other bad AI outcomes under "AI destroys the world" is a terrible idea, particularly when considering investing. Destroying the world is very hard to do, it's much more likely that any damage done by AI is far short of "destroying the world" and there is a considerable overlap between "world not destroyed" and "very bad outcomes for human flourishing, civilization and society." If you accept what should be very obvious - that it's much more likely that AI ends up creating the conditions for a catastrophic scenario that does not instantly kill you than one that destroys the world or at least you personally - then there are pretty decent investment strategies at your fingertips.

If you accept what should be very obvious - that it's much more likely that AI ends up creating the conditions for a catastrophic scenario that does not instantly kill you than one that destroys the world or at least you personally

I don't think "instantly kill you" is a prerequisite for "destroys the world". Sure, a hostile AI that goes full Skynet is unlikely to get everyone in the first pass with bioweapons, but if the AI is not destroyed or crippled beyond repair in that chaos then you're just the last light to go out; the cleanup robots will break open your bunker months or years later and there's fuck-all you can do about it. Skynet isn't like a plague or an asteroid impact, because it doesn't naturally end - indeed, if it wins, it gets worse over time as the robots build more robots.

I think a key point here is that most of the scenarios you're thinking of where there's a standard catastrophe are subsets of "AI fizzle", where AI does not hit a perfect 10. If AI is a 10, then either it's [OPPOSED TO LIFE] (and you're dead), it's aligned to someone who'll take all your stuff away/kill you (what use is wealth you cannot retain?), or it's aligned to someone who'll give you utopia (and possibly even resurrect the recently-dead).

Sure, a hostile AI that goes full Skynet is unlikely to get everyone in the first pass with bioweapons, but if the AI is not destroyed or crippled beyond repair in that chaos then you're just the last light to go out; the cleanup robots will break open your bunker months or years later and there's fuck-all you can do about it.

Yeah but

  1. this is yet another scenario where having a hedge helps you live longer and makes it more likely that you do survive, and
  2. this scenario isn't, like, real - Skynet is fictitious and the real dangerous capabilities that AI has, although probably currently overblown (remember when GPT 2 was too dangerous to release?) are more likely to generate a catastrophic scenario that you can hedge against than one you can't. This is just a probability curve.

I think a key point here is that most of the scenarios you're thinking of where there's a standard catastrophe are subsets of "AI fizzle"

It's true that I think it's more likely than not that large language models "fizzle" in the very specific sense that they don't become physics-defying literal deus ex machinas.

If you expect an AI fizzle, sure, the rest follows.

(TBC, I want an AI fizzle, at least for neural nets, but I suspect that we will need to ban it to make it fizzle. And I am a prepper, just not particularly for AI - non-AI GCRs are definitely a thing.)

Take this with a grain of salt because it isn't my area of expertise and I am dashing this off to you as a quick reply rather than doing a half-hour's worth of research to verify, but I tend to expect a slowdown in the nearish future for two reasons:

  1. The major developers (Anthropic, OpenAI) are running a fundamentally unsustainable business model and need insane amounts of revenue generation to be sustainable. They've hiked prices recently and the response from Big Tech was to crack down on the AI compute spend. That's not a good sign - who's going to spend millions on AI if Uber or Zillow won't? Not people making HTML browser games and memes for kicks, that's for sure.
  2. Despite all the breathless takes about how good The Next Model is, from what I've seen (very casually, so I might deserve correction here) the direction the models are going isn't "Anthropic's Gigamonster: now hallucinates half as much time as Gila Monster for only 110% of the inference," it's "Anthropic's Fable now hallucinates 10% less than Sonnet for only 300% of the inference." On the one hand, hallucinating only, say, 10% of the time rather than 20% of the time is a big difference, I'm not suggesting there's no gains being made. But from what I can tell, the so-called "leading edge" models are using tremendously more compute to make relatively minor absolute gains. Everyone focuses on the trendline going up but if you model it as performance-per-cost from what I can tell there's an argument the trendline is going down.

Now note that the trendline can be qualitative as well as quantitative - so for instance Haiku is just straight-up better at performing simple math than Opus or something (it will get the answers right much more cheaply) but if you're getting into something that Opus can't do but maybe Fable can - okay, maybe you're okay paying twice as much. But that general trend, I think, suggests that pushing the frontier is growing more costly, which is not ideal for an infinite recursion scenario. If you get to the point where "sure okay we can make Deus Ex but it will cost the moon if it was made of gold to run" that's another way of saying it won't get made and if it is nobody will run it.

Note that I'm not an "AI denier" or something, and I don't think that anything that I am talking about will cause AI to go away like it never existed. But I think that a world where we have AI slowdown because the demand for compute at prices that make it solvent isn't there to the extent to support the overhead to keep pushing the frontier forward is pretty plausible. A world where the US government bails out AI and then nobody ever sees Mythos again also seems pretty plausible. But both of those scenarios will look - at least temporarily - like an LLM fizzle (unless the .gov makes Skynet real, I guess - my guess is that they are more likely to use it to gundeck mandatory paperwork, hacking, and programming, though, not stick it in a bunch of robots.).

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