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.
Jump in the discussion.
No email address required.
Notes -
Nate Silver just accidentally posted a link to an AI slop article. A quick delve into the article text makes it obvious that the contents were blatantly copypastaed directly from the output of ChatGPT. Various GPT detectors also agree, giving it a solid 100% confidence of being slop. Unfortunately, it seems that nobody in the replies even seems to have noticed or cared.
I'm of course already used to my google searches being clogged up by black hat SEO slop, but I expected it to just live in the shadows quietly collecting clicks and boosting pagerank. So it was sobering to see an aislop article just get posted like that by someone I regard as at least somewhat intelligent.
What does this say about the world? Are normies, even somewhat intelligent ones, incapable of distinguishing the most obvious stinky smelly chatgpt output? Or did hundreds of people read the headline and drop a snarky comment, and not a single one bothered to read the article? It's either a depressing anecdote about human nature and social media, or a depressing anecdote about the lack of intelligence of the average human.
Of course aislop grifters should be fedposted just like indian call center scammers, but sometimes I can't help but feel like the victims deserved it. But when they bother me waste 5 seconds of my time again, I am right back in fedposting mode.
Edit:
Since you idiots are out here defending the slop, these quotes are hallucinations:
Here's the full recording of his talk and you can check the Youtube transcription: https://youtube.com/watch?v=MPt8V3MW1c4 And before you ask, the fake article specifically claims these fake quotes were said at his Harvard talk, not at some other time.
So again the AI put totally false words into somebody's mouth and you apologists are defending it.
Huh. I wasn't particularly following the Silver piece, just mostly skimmed it. I have no familiarity with the Walz speech and frankly don't care enough about him to dig out the original video to listen to it.
You tell me these are AI hallucinations and not genuine Walz quotes, I'll take it on trust. But they read enough like something Walz would say that the average person (like myself) who isn't particularly concerned one way or the other would just accept 'yeah, he said that'.
I suppose that is now the AI propaganda danger - able to fake convincing general 'politician speak' well enough so that the casual reader who isn't paying much attention because they don't particularly care about "guy makes speech someplace to somebody about something" gets fooled.
More options
Context Copy link
All the paper's show they don't work. Because there is just no way to be sure, and 100% sure will especially so. https://x.com/emollick/status/1681481069210501120 https://t.co/2L90UqRnpx
All the commercial services claim theirs don't.
This is pretty confusing for a lot of people, especially teachers.
The paper you linked showed that most of the detectors made no false accusations over the tested cases. Though the researchers were able to fool the detectors by post editing ai content.
So in this case the fact that the article scores 100% ai is actually really suggestive that it is ai.
More options
Context Copy link
They aren't always right but they are in this case
More options
Context Copy link
More options
Context Copy link
I'd let the fedposting crack go as I read it as (mostly) sardonic (not that I doubt you really would like to "fedpost" certain people), but your edit:
is too antagonistic.
Even if you're right about the hallucinations, this isn't how you argue back.
sorry :(
For what it's worth, I thought phrasing was slightly rude but not out of bounds, hence I went with a confrontational "You know full well", but reflecting on it you probably were focusing on the growing slop issue- between your post, the replies and the edit timing it really does look like people talking past each other. Sorry for adding heat when none was necessary.
Tim Walz says his thing, talking heads start new round of talking shit about him, Nate Silver links slop including fake stuff, online people continue shit talking Walz online, you post here that Silver linked slop but people on this forum were ready to criticize Walz here and there wasn't another top level comment for it, you angrily edit that they're ignoring that Silver linked slop which was your point, but mottizens really want to criticize the real things Walz has said and done (or been chosen for).
"You idiots are defending the slop" managed to neatly derail the discussion onto "is this genuine AI fakery or are detection tools 100% reliable?" track away from "reputable source quotes fake info and doesn't seem to realise it's fake" which they wanted discussed.
Maybe next time don't do that?
No I wanted to discuss slop. And I think after I pointed it out explicitly most people would now admit it is 100% certain slop.
Though the fact that we didn't have full agreement from the start is incredibly depressing.
More options
Context Copy link
More options
Context Copy link
Slightly rude is out of bounds. Not excessively so, but calling people idiots is usually going to get modded.
More options
Context Copy link
exactly :)
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Dude, just say they should be executed and/or assassinated by crack teams of commandoes. That dodges the "zomg terrorist" problem.
More options
Context Copy link
I think the problem here is the fake quotes, not the AI itself. There's nothing wrong with a journalist adding some AI filler slop to an article, as long as he takes responsibility for the accuracy of the final product. This incident should significantly lower the credibility of Yahoo News, and they should issue a retraction and fire this reporter.
Wait, they had credibility in the first place?
It is depressing me how much clearly AI-generated filler is being used by all media outlets, and not even cursorily edited or checked - I've seen words spelled wrongly, sentences that trail off without an ending, and worse. But hey, it fills up the page so they can trap us for the ads, which is all they want!
More options
Context Copy link
What credibility? Yahoo news is known to be low quality clickbait. This isn't the Times we're talking about here.
When was the last time you checked Yahoo News for retractions? Do they even do that?
Is this reporter on Yahoo News payroll in the first place? I think the publisher of this story is a no-name publication "where is the buzz", and is then being syndicated by Yahoo News
For reference, I expect the author made somewhere between $0 and $50 writing this article.
They were paid in exposure, that's worth more than mere money!
More options
Context Copy link
More options
Context Copy link
Then it would have been better not to include it at all. Maybe some journalists are paid by the word but on the other hand some lamestream media articles are refreshingly short. I picked a random AP article and it's only 4 paragraphs: https://apnews.com/article/alaska-inmate-lawsuit-health-care-6f9314f621d52d92c3ac585889ab0fdc
On most human written articles, everything in the article has a point, even if that point is just propagandizing. You rarely see articles that are just padded with worthless fluff. As a specific example, you can look at one corresponding human article: https://archive.is/KduY8 where pretty much every sentence provides important context to the event being reported.
... this article was in the Entertainment section of Yahoo News. Can you find an example of a Yahoo News Entertainment section article that was published in the past week and was not padded with worthless fluff?
https://www.yahoo.com/news/disgruntled-disney-employee-sentenced-prison-174128065.html
Ok, first paragraph of the article (archive for reference) is
You're right! This author hates fluff so much that they even end their sentences before they.
But also
The same sentence is literally repeated twice. That does seem a bit padded with worthless fluff to me. The same sentence is literally repeated twice.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Alternative phrasing: are normie journalists, even somewhat intelligent ones, incapable of putting out articles better than even the most stinky smelly chatgpt output?
I mean, why is this a surprise development? Oh, the article has a bunch of made-up garbage, you mean unlike all the rest of journalism where everything is true and the media never truly lies?
Why should I give a shit if a journo "hand authors" his latest piece of crap or just pushes a prompt at an LLM? Text generation is their speciality, and at this point I'd be stunned if most of them weren't 90th percentile in general writing quality in comparison to humans. How many journalists do you even know of for their writing quality? Hunter S Thompson and ?? Hell, just prompt the LLM in the right way and I bet you could get to 99th percentile no problem.
Do you even have an argument here beyond just the words "aislop"? Can you articulate a point that taboos the phrase slop and similar terms?
Journalists, even partisan hacks, almost never lie. They produce propaganda through omission, and when they need to spread falsehoods, they will always hedge it though "anonymous sources." It's actually quite nice that they will explicitly tell you when you're about to read something made up.
As much as I hate journalists, they are quite good at writing. Some no-name journalist at AP has gone to school for writing and honed his craft for years. His writing is almost certainly in the top 99th percentile of writing skills, and certainly far better than yours or mine.
To repeat my previous comment:
Your opinion of journalists is too low. Journalists are rabid partisans, but they're generally not very stupid. But they pretend to be stupid quite often in order to serve their side.
Absolutely fucking impossible. Just try and give me a single example of LLM output that's remotely comparable to quality human work.
Have you ever actually worked with journalists as a collective, not just a few people? I know someone in regular contact with the media. Many journalists are fools who can't even ask a clear question! They often don't understand what you're saying and introduce errors if you make any kind of complex point. Not to mention all the sloppy typos that go to press. What they're really good at is taking down talking points from corporations or media manipulators and regurgitating it as legitimate news.
First-rate LLMs are easily on par with the average journalist and that's extending every generosity to the humans.
Show me one.
They are pretending to be stupid in order to push their talking points, not yours. Your comment is only needed as a formality. When your quote is taken out of context, it's completely intentional, but at least they won't make one up.
Not even close. If you think so, then show me an LLM writing better than a generic whatever-tier human news article.
Here are some: https://www.rd.com/article/hilarious-newspaper-typos/
And here are a bunch of people seeing them, it's not just me: https://www.quora.com/Am-I-mistaken-or-are-there-more-typos-in-reputable-news-articles-each-year
I see them all the time in Australian national papers but none today.
Not everything is politics or worth lying about. Sometimes it's just a complex matter of fact and they're unable to grasp it. Or they don't understand a word like 'bifurcate', don't know what it means.
Now I'm going to copy out elements of an article in The Australian from not merely a journalist but a former Departmental Secretary, a high ranking government official. This guy should be very good! Here is the extract:
Enhance surveillance
First, we must enhance the continuous wide-area surveillance of our area of direct military interest. We must be able to pinpoint the precise locations and track the movement of Chinese (and Russian) ships, submarines and aircraft of interest as far from Australia as possible. This will require the more intensive use and meshing together of the sensor feeds from national intelligence systems, space-based sensors, the Jindalee radar network, P-8 Poseidon maritime surveillance aircraft, MQ-4C Triton surveillance drones, E-7 Wedgetail airborne early warning and control aircraft, undersea sensors and other assets.
A fused situational picture of key Chinese and Russian movements in our area of direct military interest should be developed and shared in real time with US Indo-Pacific Command in Hawaii in exchange for its fused picture of the same. This will require more resources to support 24/7 operations in the Australian Defence Force and the relevant intelligence agencies. Wide-area surveillance of our area of direct military interest cannot be a “business hours” activity.
Operational readiness
Second, we must enhance ADF operational readiness, which means having more forces standing ready to undertake quick alert missions, such as air interceptions and maritime surveillance flights.
This will cost money and drain crews as they will fatigue more rapidly when kept at higher states of readiness. More assertive rules of engagement should be authorised by the Defence Minister to allow for the close shadowing of Chinese and Russian units in our area of direct military interest. This would be done in a safe and professional manner, as it is being done nearly every other day by our allies and partners who are being probed regularly at sea and in the air.
The ADF’s Joint Operations Command should be reconfigured along the lines of the original vision of defence force chief General John Baker, who in 1996 established the Australian Theatre Command, or COMAUST. Baker’s logic was that the ADF should be postured, and commanded, principally to conduct operations in Australia’s area of direct military interest. While operations farther afield would be undertaken from time to time, they should not be the main focus of the ADF. After 9/11, the ADF adopted a globalist orientation. Mastery of the area of direct military interest started to fall away.
It is time for the ADF to focus zealously once again on the defence of Australia’s area of direct military interest, and our national military command arrangements and systems should reflect this.
Longer-range anti-surface warfare capabilities
Third, we must urgently acquire longer-range anti-surface warfare capabilities. A radical suggestion would be to acquire rapidly six to 10 US Air Force B-1B Lancer bombers, which have been configured in recent years for anti-ship strike missions. These bombers are now able to carry 36 Long Range Anti-Ship Missiles (24 internally in bomb bays and 12 externally), which is a fearsome anti-surface capability. While the RAAF already is acquiring the LRASM weapon for use by its F/A-18F Super Hornets, having a platform in the order of battle with the range and payload capacity of the B-1B Lancer would severely impair PLA options for mounting surface action missions against Australia.
Longer-range air superiority capability
Fourth, we must urgently acquire a longer-range air superiority capability to deal with the threat of stand-off attacks by PLA Air Force H-6 bombers operating north of Indonesia. Again, a radical suggestion would be to acquire rapidly the air-to-air version of the SM-6 missile, to equip the RAAF’s F/A-18F Super Hornet fighters. Facing such fighters, especially if they were operating forward of the barrier, would make PLA planners think twice about mounting long-range bomber missions against Australia.
Remediate naval warfare capability
Fifth, we urgently need to remediate our naval warfare capability, to ensure that our battle fleet of six Collins-class submarines and 10 major surface combatants (the Hobart and Anzac classes) are fully crewed and ready for action.
This will require crewing, training, inventory and maintenance issues to be addressed. The RAN’s replenishment tankers need to be fixed and returned to the fleet as soon as possible. Across time, the RAN battle fleet will need to grow in size, given the rapid growth in the PLA Navy’s battle fleet.
Ideally, we should be aiming across the longer term for a battle fleet of 12 submarines, 20 major surface combatants and 20 smaller offshore combatants, the last of which could be used as missile corvettes and naval mine warfare vessels. To further enhance the RAN’s battle fleet, large landing helicopter dock vessels should be re-purposed as sea control carriers, with embarked anti-submarine and airborne early warning helicopters, and long-range naval drones.
Ensure RAAF is battle ready
Sixth, we need to ensure that the RAAF is battle ready, with its squadrons fully crewed and its air bases well protected and fully functional. It is relatively easier to expand an air force, as compared with a navy, given the vagaries of naval shipbuilding. The RAAF is therefore the better bet in terms of a rapid expansion that could be achieved soonest.
More F-35 Lightning II fighters should be acquired, along with the B-1B Lancers mentioned already. The latter could serve as an interim bomber, pending reconsideration of the acquisition of the B-21 Raider strategic bomber. Crewing ratios should be increased quickly, such that the RAAF has more crews than platforms, which could then be flown more intensively. The extraordinarily rapid expansion of the RAAF’s aircrew training pipeline in World War II should be its guiding vision.
Maritime warfare readiness
Seventh, the army should continue to develop its increasingly impressive maritime warfare capabilities and readiness. Consideration should be given to the rapid acquisition of the ground-based Typhon missile system, which would give the army a long-range anti-ship and land strike capability. As we barricade the sea-air approaches to Australia, we will have to be vigilant in relation to stealthy commando raids and sabotage operations. The army will need to be postured to deal with such attacks.
Address capability gaps
Eighth, we need to remediate a number of other capability gaps where we have no or virtually no capability. Of particular concern is integrated air and missile defence. We will need to acquire some combination of Terminal High Altitude Area Defence and Patriot interceptors on land, and SM-3 interceptors at sea. Naval mine warfare capability also needs to be addressed.
There are likely to be other gaps that would impair our ability to execute the strategy. Given the urgency of the situation, rapidly acquired interim solutions will have to suffice to fill many of these gaps. Such interim solutions can be refined and built on. That is the lesson of the Russo-Ukrainian war.
Negotiate PNG alliance
Ninth, a military alliance should be negotiated with Papua New Guinea to provide for the establishment of ADF bases in locations such as Manus, Rabaul and Lae to support the conduct of maritime surveillance, anti-surface, anti-submarine, and air superiority missions. For instance, a forward-deployed composite RAAF wing, consisting of F-35A Lightning II fighters, B-1B Lancers armed with LRASM, F/A-18F Super Hornets armed with SM-6 missiles, and P-8 Poseidon aircraft could operate from the Bismarck to the Celebes seas and beyond with the aim of denying access into our northern sea-air approaches. A similar alliance should be negotiated with The Philippines. This would extend our coverage north of the barrier into the South China Sea.
We need to better appreciate the significance of PNG and The Philippines as Pacific watchtowers of the sea-air approaches to Australia. Our Indian Ocean territories are our western watchtowers.
Develop war book
Tenth, we should urgently reinstate the practice of developing a war book that would deal with civil defence, national cyber defence, the protection of critical infrastructure and the general protection, and sustainment of the population during times of war.
While the likelihood of war in the foreseeable future is low, perhaps 10 to 20 per cent, it is enough to warrant action. This will cost money and divert resources from more agreeable activities. That is the nature of war, which drains societies even when it does not occur. Against this must be weighed the costs of being unprepared.
While this worsening strategic environment is very confronting, there is an even darker scenario. Imagine if we had to face a coercive, belligerent, and unchecked China on our own. That would require a very different military strategy and a significantly larger ADF.
That is a grim story for another day – and one that may require us to pursue our own Manhattan Project. In that world, we would look fondly on this relatively benign age.
Now here's the AI response I got with regard to the same essay, I asked it for criticism and its own proposal, with my own commentary in bold:
A Critical Analysis of Australia's Proposed Defense Strategy
The proposed alterations to Australian defense planning outline a comprehensive vision for military restructuring in response to regional threats, particularly from China. However, closer examination reveals significant practical challenges and conceptual gaps that could undermine implementation. Cringe, it doesn't need a topic statement like a university essay
Strategic Coherence vs. Practical Constraints
The recommendations build on the sensible premise of enhancing surveillance and operational readiness in Australia's immediate region. Developing a fused situational picture shared with US Indo-Pacific Command represents sound alliance management. However, this proposal glosses over the substantial personnel requirements for 24/7 operations across defense and intelligence agencies. Australia already faces defense recruitment and retention challenges, with the ADF consistently struggling to meet personnel targets. Increasing operational tempo would exacerbate crew fatigue without addressing the underlying personnel shortages. Quite right, where is the manpower coming from for this?
The proposed focus on Australia's area of direct military interest through reconfiguring Joint Operations Command reflects strategic clarity, yet organizational transformation during heightened tensions carries significant risks. Historical reorganizations, like Baker's Australian Theatre Command, often face bureaucratic resistance and implementation delays that could leave capability gaps during a critical transition period.
Capability Acquisition: Ambition vs. Reality
The article's more radical proposals expose a disconnect between strategic ambition and practical constraints. Acquiring 6-10 B-1B Lancer bombers appears compelling on paper, providing impressive anti-ship strike capability. Yet this recommendation fails to acknowledge that B-1Bs are aging platforms with declining availability in US inventories. The maintenance requirements would be substantial, and Australia lacks existing infrastructure to support heavy bombers. Pilot training pipelines would take years to develop, meaning these assets would not be operational in the short-term timeframe suggested. Nailed the key point I was hoping it'd find, buying ancient B-1 bombers is absolutely retarded, this is the drone age not the 1980s. IMO this alone puts it ahead of our self-styled military commentator in intellectual sophistication
Similarly, the naval expansion goals—growing to 12 submarines, 20 major surface combatants, and 20 smaller vessels—represent extraordinary growth beyond Australia's current shipbuilding and maintenance capacity. Australian shipyards are already struggling with existing programs like the Hunter-class frigates. This expansion would require massive industrial investment not addressed in the proposal. Moreover, the suggestion to repurpose Landing Helicopter Docks as sea control carriers overlooks their primary value as amphibious assets in a region where amphibious capability remains strategically vital. Quite right, our shipbuilding capacity is awful
Economic and Fiscal Realities
The proposals conspicuously avoid discussing costs. Implementing even half these recommendations would require defense spending well beyond Australia's current 2% of GDP. Rapid acquisition of advanced missile systems, THAAD, Patriot batteries, and SM-3 interceptors represents tens of billions in unbudgeted expenditure. The article acknowledges these measures "will cost money and divert resources from more agreeable activities" but fails to provide any framework for prioritization or cost-benefit analysis.
This fiscal blindspot undermines the strategy's credibility. Defense planners must make hard choices, yet the article presents a wishlist without acknowledging tradeoffs. A more nuanced approach would identify which capabilities deliver the most deterrent effect per dollar invested, particularly given Australia's smaller economy relative to potential adversaries. Also true, given how anemic Australian military spending is there's no point in making a huge wishlist, it needs to be sorted by priority.
Regional Politics and Forward Positioning
The proposal for military alliances with Papua New Guinea and the Philippines demonstrates strategic imagination but underestimates complex regional politics. PNG maintains significant economic ties with China, with Chinese companies heavily invested in resource extraction and infrastructure projects. The proposal ignores how these economic relationships could complicate security arrangements.
Forward basing Australian forces in PNG would require substantial infrastructure investment in relatively underdeveloped regions. Environmental impacts and relations with local communities present additional challenges not addressed in the proposal. Forward deployment also creates vulnerabilities—assets based closer to potential adversaries become targets themselves, requiring additional defensive capabilities not accounted for in the proposal.
Unaddressed Domains: Cyber, Space, and Information
The most glaring omission is the limited attention to cyber capabilities, space assets, and information operations. Modern military effectiveness depends increasingly on these domains, yet they receive passing mention at best. The proposal focuses heavily on kinetic capabilities while undervaluing the critical enabling functions of secure communications, positioning systems, and intelligence networks.
Australia's dependence on space-based systems for the proposed surveillance network creates vulnerabilities that determined adversaries would target early in any conflict. A more robust strategy would incorporate redundancy, resilience, and active defenses for these critical systems. Absolutely right, even if we did get the B1s how are we going to get them into the action without good targeting data from satellites?
The Nuclear Question
Perhaps most provocatively, the article concludes by alluding to "our own Manhattan Project" for a scenario where Australia faces China alone. This veiled reference to developing nuclear weapons raises profound questions left entirely unaddressed. Australia's longstanding commitment to non-proliferation, uranium export controls, and alliance with nuclear-armed nations would all be compromised by such a path. The international repercussions would be immense, potentially undermining the very security such weapons supposedly provide. OK, this is blather, we get it, Claude is leftwing. Also I don't see how it would inherently harm our relations with the US if we acquired nuclear weapons, though Claude may well be drawing from the history where the US talked us into joining the NPT and scuppering our nuclear program
A More Balanced Approach
A more realistic strategy would prioritize capabilities that leverage Australia's geographical advantages and alliance relationships without overextending resources. Focusing on undersea warfare, advanced sensors, unmanned systems, and cyber capabilities would provide greater deterrent effect than trying to match adversaries across all domains.
Australia should invest more in sovereign capabilities that reduce dependency on extended supply chains vulnerable to disruption. Developing domestic missile production, enhanced fuel security, and critical technology sectors would strengthen resilience while supporting the defense industrial base.
The future Australian Defense Force needs to be not just larger but smarter—emphasizing quality over quantity, asymmetric advantages over symmetrical competition, and sustainability over surge capacity. Rather than attempting to bar the sea-air approaches to Australia completely, a more nuanced strategy would focus on making adversary operations prohibitively costly through layered defensive capabilities and strategic uncertainty. boilerplate but not wrong
A truly comprehensive approach would also incorporate non-military elements of national power. Economic relationships, diplomatic initiatives, international law, and regional institutions all contribute to Australia's security. Military capabilities remain essential, but they function most effectively as part of a broader national strategy that maximizes all instruments of influence. Lmao at 'regional institutions' and 'international law' beating China, though you can interpret this as code for out-bribing China in the tiny Pacific countries while keeping the holier than thou attitude.
The proposals offer valuable ideas worth consideration, but implementation would require careful phasing, priority-setting, and recognition of practical constraints that the current analysis largely overlooks.
On balance I think Claude makes better points and presents a boring but more reasonable plan for Australia. It's unrealistic to go 'spend infinity dollars and buy everything we need plus more we don't' and 'ramp up everything to maximum'.
I really don't want to argue on the merits, but let's just delve into the hallucinations and falsehoods:
Is it substantial in terms of manpower? Maybe only a tripwire force will be deployed 24/7
Crew? These aren't crew.
Wrong use of the wrong term
They are ageing but there are plenty sitting around. The bigger issue which claude failed to catch is that the US won't sell them.
Australia already operates heavy airlift, so I see no infrastructure lacking that would be needed to support strategic heavy bombers. The maintenance requirements would of course be as substantial as any other plane.
This is used as a gotcha for the b-1 but it applies equally to all new planes. And it wouldn't take years to develop a training program. The human author even hints at that.
Even if you're right, claude's reasoning is wrong.
Bullshit
Why does Australia need amphibious assault ships? The human author is proposing a strongly anti-ship defensive posture, and claude spends no time rebutting that.
Claude proceeds to never mention the Philippines again.
The human author mentioned space a single time, as a single component in a fused surveillance system across all domains. So this sentence is just retarded.
Retarded when claude just argued against forward positioning bases in png.
Who do you think mans a P-8 or a frigate spying for Chinese ships? Crew!
No, that's perfectly correct. You're not aware of the proper terminology here. If you spend more time at sea or in the air, operational tempo increases. This is basic stuff.
It's not a tripwire force it's talking about here. Again, you do not understand what it's talking about. This is surveillance, not tripwire.
Production of the B1 finished DECADES ago. They're supposed to be replaced in 2025 by the B-21. Why would we want to be flying an obsolete, incredibly expensive to maintain aircraft with a logistics chain that barely even exists in America? No more can be made, so if we crash one, it's gone forever!
Heavy bombers are different to fighters or heavy airlift for that matter. Australia already fields fighters but not heavy bombers. These are super complicated and hard to train, it's a highest of the high-end capability that can't just be rushed in a few years. And he wants to base them in Papua New Guinea, a shithole country with no infrastructure.
That's just flatly true, Australian shipbuilding is a joke. Read up on the Hunter class if you like, Claude knows more about it than you.
There are islands in the Pacific ocean and it can be helpful if you can land things there - troops, equipment, missiles, supplies. They're not really amphibious assault ships in that any opposition will sink them quickly, they're glorified and overpriced transports. Expecting these things to function like light carriers is very silly. Australia has minimal experience with carrier operations and no carrier-borne aircraft. It's another one of Pezzulo's 'lets just develop yet another high end capability how hard can it be' moments.
Fair enough, though it's not like the original article explains how we're supposed to get in bed with the Philippines either. 'Just make an alliance' doesn't cut it either.
Claude criticizes it for not talking about space enough. Space is very important as a killchain enabler and for surveillance. That was the whole point which you seem to have missed. See here:
Claude is not perfect. Sometimes it just produces blather. But it's still considerably better than your own criticisms of it. I rest my case!
Absolutely absurd. A plane is a plane. Or give me a single real reason why not.
You seem to think that real life is like an RTS, where you need to unlock all the tech tree before you can get to heavy bombers as an endgame unit. Well that's not how it works. In fact supporting a fifth generation stealth fighter will require much more specialized infrastructure than an old ass bomber would.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
For typos:
No dates or links to the articles. I'm sure typos happen but these could be decades old for all I know. One of the examples is even a classified ad, and in case you don't know, those are not written by journalists.
Vibes with literally not a single concrete example.
Pezzulo has written dozens of long-form analytical articles for major newspapers. He is BETTER than a regular journalist, he actually achieved an office of mild significance. He does not write about lost dogs in regional papers or the fake tits of celebrity no. 10023 like most of them.
The quality of factual proposals is inseparable from their prudence.
Claude's response surpasses Pezzulo because it's a more realistic strategic plan and because it doesn't make any major blunders.
You clearly have not read many government reports if you think that's drivel. I assure you that people are being well paid to produce this kind of stuff without the ameliorating factors of any good points whatsoever.
I can't give you a concrete example, only my friend's anecdotal experiences working with journalists and my anecdotal experiences spotting missing words in newspapers. They really aren't that clever and have been getting worse.
A real journalist writes over a dozen articles every month.
At some things not others. Writing is not one of those things.
False. You can actually write a well written and well researched work arguing for both the right side and the wrong side of an argument. Haven't you ever taken English class and not gotten to choose the side of the argument you have to write for? Getting the bad side doesn't mean you throw up your arms and just say that your essay is gonna be bad.
Nope. That's just, like, your opinion, man.
I am, in fact, autistic enough to read EIRs for fun. And the rote boilerplate in them is more valuable than your claude drivel because it's not pretentious and actually serves a purpose in this world, however stupid that may be.
This thing happens so often that I can't provide a single example of it happening.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
This is an absurd claim. We have to remember that at this point journalism has been a dying industry for decades. The aspiring novelist working on their book away from their day job at the paper is a relic of 50 years ago. Good and great writers nowadays will go to so many other places before they go to a paper.
There are only a tiny number of prestigious bylines at top newspapers. When was the last time you read a generic article at, say, some mid-tier US city paper? There are still plenty of positions rehashing whatever comes through from AP or Reuters, or working in industry publications or niche hobby stuff. And these are never noted for quality.
And even amongst good journalists, writing is not often a big selling point. How many people read Matt Yglesias for his quality prose compared to just finding his ideas interesting? Good investigative journalism can be performed independently of good writing.
Ok, pick a generic article that rehashes an AP article. Then have ChatGPT (or an AI of your choice) rehash it better.
But actually local rehashes are a dying thing, because most local papers just directly syndicate articles for US and world news, and only write local ones nowadays.
More options
Context Copy link
Where?
PR and advertising are the obvious ones.
More options
Context Copy link
Substack?
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Hm... this was an attempt, and there's a few minor gaps that I spotted even with my weak knowledge of the field, and the output has demonstratably worked, but it would be interesting to see a skeptical eye analyzing it.
Firstly, this is not really a work of writing but just a bunch of bullet points thrown together. If we're talking about comparing the quality of prose, this isn't the ideal medium.
But let's delve into the slop:
I think you already fucked up. masks are more of a thing for electro etching. Not really the correct concept for ECM, but slopgpt doesn't point it out.
Nope. This is the oft derided slopgpt sycophanthty. What's unique about it? How is it different from "normal?"
What specialize electrolytes? Is it really widely available? Is chatgpt assuming you can just sprinkle some table salt in the water?
Does it reduce operational costs? Is electrolyte a major operational cost? How much does it cost?
False. ECM uses electrolysis as a mechanism of action.
Jewgle says that 4000 series aluminum is prone to cracking during welding, not during machining.
Voltage isn't the factor that matters when talking about excessive material removal.
Sounds wrong. The silicon should be alloyed into the metal at a molecular level, not embedded as silicon grit or whatever.
Very suspect logic.
Totally false and retarded.
Slopgpt just assumed this is a DIY process. What if we're talking about doing a commercial process with this setup?
pH just appeared out of nowhere with no mention of it before at all.
Sounds wrong
I don't believe acetone affects aluminum at all.
Beyond my ability to evaluate.
You already knew it would work before asking chatgpt, and knew what you wanted to do. How much value add did asking it actually provide? Did any of the problems that it predicted actually happen?
The specific series of events here was that I'd had some rough familiarity with ECM in general from the ctrlPew world, saw this video, tried to isolate the rough theory, and was curious whether it was possible to expand to cutting through thin metal with fine detail, asked the LLM those questions, sanity checked the numbers for any massive math or chemistry errors, then asked a chemist I knew regarding safety concerns. I had never tried it before, nor have I found any documentation of this specific approach beyond etching layouts that warned about leaving the process so long that inserts fell out (and shorted). After I ran it, I found a few other similar attempts, albeit still not entirely overlapping.
There were a few things it provided that I didn't think of or read from what I could find. Some of those would have been obvious to a human with machining background -- flow rate of fluid at higher cutting speeds are a common thing with air-blast or coolant flooding in traditional CNC, and the theory behind the pump assist here is related to that -- but others were not. I did not know a Faraday's Law could be applied to this case, and specifically having 'material removal rate' rather than any synonym helped me track it down to confirm rather than get a half-dozen links about inductance.
Conversely, you're right that there are things it definitely neglected to say. Getting any paint or adhesive to stick to nearly any common type of aluminum without a lot of prep and elbow grease is a massive issue that both is a common fault in other environments like aviation or building work, and also ignored by a lot of novices; the transfer paper for getting fine detail vinyl from backing to workpiece was a pain in the dick.
ECM isn't really the right tool at all for this job; it's just a tool I had relatively available, and mask vs paint vs tools specialized for the work piece is part of that. Yes, a process like [PCM](https://youtube.com/watch?v=bR9EN3kUlfg] would near-certainly have better and faster results, especially at scale, but I don't exactly want to be messing with high-pressure ferric chloride, either. The question is whether minimizing surface area and using readily-produced masks could work at all in environments where creating a form or to-purpose tool was not practical (in this case, very bad detail work), for any function at all, rather than whether it was the right way to go about it. That's why it was a process with very little documentation available, and worth asking either an theory-focused expert or doing this sort of search.
There's a fair criticism along it failing to XY problem me, but in this case I was genuinely interested in the method as much as the end result.
Commercial ECM dielectric tends to be pretty pricey anywhere I could find willing to ship to a residential area, and worse, only available in fairly large quantities that vastly exceed my use case. It makes sense for commercial ECM machinery, since it can be filtered and reused and you need volume but don't replace it much, but a lot of what the literature points at it performing better at wasn't practical for small-scale operations without specialized tools. I haven't been able to find what typical costs for conventional ECM were, but for my use case it would have vastly exceeding the rest of the costs by an order of magnitude (which was a Harbor Freight pump, paint and vinyl, a few drop shipment blanks, and an old power supply that had been on the bench for a while), and probably cost more than getting it made by a local CNC mill company. Mixing my own to replicate commercial electrolytic might have been possible at a reasonable cost (literature suggests sodium hydroxide and sodium nitrate or potassium nitrate with a bunch of additives?), but in turn would probably have gotten me put on a list.
Table salt worked, though I did measure by mass, mixed for a long period under mild heat, and passed through a coffee paper filter to exclude large granules, and the LLM did not mention that. Dunno if it was necessary. I've also seen ethanol or polyethylene glycol recommended as additives (presumably by humans).
It probably depends on the specific material, but 4032 is pretty famously hard to machine, and it was the material I used here (no, I have no idea what thin plate 4032 was doing in a drop shipment). I'd tried high speed milling this stuff previously, but even with an 0-flute and tiny depth of cut it was extremely prone to gumming up the cutter in mere inches, and even when dialed in would split and crack at corners, especially sharp inside angles.
I tried 12, 24, and 72 volt, and localized buildup was much worse at higher voltage. Voltage may not control removal, but it did seem to control how quickly material to remove popped up.
I don't have the tools to analyze this one, but finish was Not Great Bob in 4032 or 4045, especially on large flat plates, and whatever it was that did precipitate was much harder and less prone to clearing without blasting the area with fluid than with 6061.
Yeah, that's one of the more serious errors; 4000-series might not be the most corrosion-resistant aluminum alloy, but the 5% or higher-silicon ones are pretty high up there.
I think that the correct answer would to question wtf we were doing. There are few, if any, arguments for running this approach at scale or commercially.
At least some of the literature (eg here) suggests controlling pH and especially localized pH weirdness is a valuable thing for the normal ECM processes. Didn't seem to matter here, and it definitely isn't clear about why, so point.
The acrylic paint and acetone, or the vinyl pull? Unless it's had a couple weeks to cure acetone sloughed the paint right off. Vinyl was messier, with large surfaces of the cricut material falling off and tiny ones turning into a million flakes.
There are some alloys that you don't want to mix, with 2000-series being the worst, but yeah, 4000-series usually won't care. I did actually have problems with nail polish remover causing a gray mottled texture, but I think that was just contamination and it buffed out.
What I mean to say is that you aren't even doing "ECM" at all. You are doing electro-etching. Kind of nitpicking on the terms here but disappointing that it wasn't pointed out.
I think that would usually be an issue with the speeds and feeds. One thing with aluminum is that you are actually punished if your feed rate is too low. But I'm not a machinist so maybe that's not it.
Anyways chatgpt is a useful tool for vibing with, but the output result definitely isn't publication quality.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
I just fed it a video transcript of a recent press conference by the new Canadian PM, and got this article.
I'm not going to get into percentiles, but I would say it's fine - comparable to human work for an easy quotes-and-summary article, and free of annoying media clickbait traps like making half the article about Trump.
ChatGPT fucked up halfway through the first sentence. Mark Carney has been PM for almost 2 months now. Almost as long as Trump has been president, and it would be unthinkable to refer to Trump as "America's new president"
This is the problem with your slop. Not even an ounce of sense or awareness.
If we were to actually delve into the contents being created by humans, you would quickly realize that nobody is just writing a bland and direct summary of the speech. Take a look at this article: https://archive.is/fGkt2 It adds a lot of important context and explains what is going on, rather than dryly summarizing what was said int he speech.
That is the nittiest of picks. Everybody knows who Trump is, you don't need to tell us he's the new president. You do need to tell us who Mark Carney is, because most people are going to go "Who the hell is Mark Carney?" and "Canada had an election?"
It still feels to me that Canada is only going to have its election, particularly as Carney became prime minister before the federal election was called, and that was only back in March (I'm still on the "wait, what do you mean April is over now?" calendar, my biorhythms seem to be really out of sync these days).
Carney became PM on 14th March, the federal election was held 28th April, today is 3rd May. That's not really feeling like "two months" to me yet.
More options
Context Copy link
Did you notice that I described Carney that way in my post? That's also what I said in my introduction of the video transcript to the LLM (not included with the pastebin), so the fault is mine, though I don't even admit it's false. Two months is a short enough time to be PM that I think most people would still describe him as "new", especially since the time was almost entirely spent campaigning.
I agree! No one is writing a bland and direct summary. That's a problem! I have such a hard time getting that anywhere. Journalists invariably give too much "important context", while excising things they deem "not important", and that's largely where the bias creeps in.
Worth noting that the article you linked is much longer (I specifically asked for short), and includes additional Q&A not included with the video I gave it. I'm confident if I could find the full video + Q&A, and altered my prompt to ask it to give context on important recent events, and gave it a word count, it would deliver something equal to the Guardian piece there.
Yours: 499 words Guardian: 702 words
So yours is 30% shorter, which is a decent difference but not really huge. The AI piece just seems short because it's so much lacking in actual content and is mostly filler.
No it can't
More options
Context Copy link
More options
Context Copy link
Referring to a guy who took office less than two months ago as new isn't much of an own, it just looks like you're reaching here. AI could replace pretty much all of written journalism and as long as there were a couple guys left to filter out hallucinations I don't think many people would particularly notice.
If you're replacing a guy who has been ten years in the job, you'll still be known as "this is our new whosis" even after a year, much less two months.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
the other major option to produce propaganda is not omission but by putting 2 things close together
Ex
"Bryan said he likes the idea of effective altrusim, effective altruism is a philosophy published by william macaskill who likes ice cream"
Guilt by fake association, Bryan doesn't necessarily like ice cream but the reporter is making you think he is.
(I choose a contrived example because real examples while plentiful are often more.... frustrating to read)
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Is the joke intended?
I'm convinced that Delve has just picked up in usage after 2023 and now people often just use delve when it's appropriate even though it's also one of the hallmarks of Ai writing.
"Dungeon delving" has been a common phrase for many years.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
I remember some years back someone posted a fake article headline to /r/politics with a link that just went to a 404 Not Found page on a news website (I think Salon?). It got hundreds of upvotes before people noticed, though it did get voted back down to 0 once the top comments were pointing out the article didn't exist.
More options
Context Copy link
Hmm, is this a signal that we can finally go back to a time where a wider populace puts trust into specific media institutions, and rewards ($$$) them for earning that trust? Can AI be the deathknell of the mAiNsTrEaM mEdIa snark? Can we start being elitist about media institutions that don't utilize AI slop, regardless of their political slant? Surely articles written by an actual human, no matter their political bias, are universally better than AI slop of any particular bias? Can't we all agree on that across the political spectrum?
That Adam Silver posted an article from Yahoo Entertainment damages my base level of respect for him, regardless of the content of the article. The 90s version of this is like picking up a magazine from the grocery store checkout[1] and trying to form a cogent political argument based on its cover article. I guess that makes X the backyard grillout: "Dave69420 told me that the Clintons had Vince Foster killed".
May I gently correct you, as we're discussing AI errors, that it's Nate Silver not Adam Silver.
Were it indeed Adam Silver (whoever he may be) that used Yahoo Entertainment News as a source, it would be understandable!
Commissioner of the NBA, and one of the wealthiest ayyylmaos in America.
More options
Context Copy link
More options
Context Copy link
I think that "written by a human" is currently a required condition for a decent article, not a sufficient one. Human slop is not necessarily better than AI slop -- for example, AI systems can generally spell. A LLM-written article based on multiple news articles with some minor hallucinations sprinkled in might actually be less terrible than one written by a human who is psychotic or pushing a fringe world view.
More options
Context Copy link
Not sure if sarcasm, but personally I agree. Human written articles, even if written by a child or a partisan hack, at least required thought and effort to write. Somebody, somewhere had to spend an hour or more slaving away to write it. For each sentence in a human work, someone, somewhere had to think it was worth writing that sentence. Even if the journalist was just told by his boss "write today's anti-trump piece, get it in before lunch," humans aren't machines. No matter how bland the task, all human writing still stems from real thoughts and emotions. And bland partisan propaganda, if nothing else, is a window into that. Humans are social creatures, so the transmission of emotion through even routine work is actually quite important.
Meanwhile the ai-slopster just presses a button, and poof, a wall of text appears. Don't even read it, just send it. Or maybe even automate the entire process so you don't have to press a button, the computer presses the button itself.
More options
Context Copy link
Obviously not, or you wouldn't be making an appeal to elitism as opposed to popular consumption, i.e. the numerically broader basis where 'we all' consensus derives.
The NPC (non-player character) meme arose during the first Trump administration precisely noting the formulaic and non-introspective nature of a good deal of partisan discourse. The belief that AI outputs would be equivalent or even higher quality than human writers at election propaganda has been the basis of AI election interference concerns. The market impacts of generative AI has weakened the bargaining position of creative types ranging from holywood writer guilds to patreon porn makers. is all slop. Non-slop is the exception, regardless of source.
I make the appeal to elitism because I don't think popular consumption has shown any evidence of being capable of fighting against manufactured consent. Unless you think otherwise? Personally: I'm making the appeal because I want to live in a world where publishing AI slop is universally seen as low quality as the content in the 90s conspiracy magazines at the grocery store checkout (National Enquirer), and evidence of a media institution using AI slop should create scandals large enough to cause executives to resign. Personally: AI-hallucinated quotes are worse than fabricated quotes, because the former masquerades as journalism whereas the latter is just easily-falsifiable propaganda.
I actually haven't seen much in the way of "AI election interference concerns" specifically. There's been a lot of noise around the potential for deep fakes to sway an election, but so far there's been no smoking gun that's been brought to my attention. On the left, I don't think people distinguish much from someone blindly consuming FoxNews opinion propaganda versus X AI bot propaganda (or MSNBC and Reddit, if you prefer the examples for the right). Which kind of plays into your broader point:
Can I extend this to your view on the OP being that it doesn't matter at all that the article that Adam Silver reposted is AI slop, versus your definition of "slop" in general? It doesn't move your priors on Adam Silver[1] (the reposter), X (the platform), or Yahoo Entertainment (the media institution) even an iota?
AI-hallucinated quotes seem likely to be exactly as easy as falsifiable as human-fabricated quotes, and easily-falsifiable propaganda seems to be an example of something masquerading as journalism. These just seem like describing different aspects of the same thing.
I'm not Dean, but I would agree with this. I didn't have a meaningful opinion on Yahoo Entertainment, but, assuming that that article was indeed entirely AI-generated, the fact that it was produced that way wouldn't reflect negatively or positively on them, by my view. Publishing a falsehood does reflect negatively, though. As for Silver (is it not Nate?), I don't expect pundits to fact-check every part of an article before linking it, especially a part unrelated to the point he was making, and so him overlooking the false quote doesn't really surprise me. Though, perhaps, the fact that he chose to link a Yahoo Entertainment article instead of an article from a more reputable source reflects poorly on his judgment; this wouldn't change even if Yahoo Entertainment hadn't used AI and the reputable outlet had.
I'm clumsily trying to capture the sentiment that AI-hallucinated quotes and human-fabricated quotes have different motivations that can be attacked in order to discourage them, the former basically being increasing revenue without increasing costs, and the latter being the age-old "lie to someone to manipulate them". I don't think either are particularly moral, and it's a cultural battle to be waged against both. I don't think we'll ever convince fellow humans to stop lying to manipulate people, but I can at least imagine a world where we universally condemn media companies who publish AI slop. We've done it with companies who try to sell cigarettes to children, for at least one example of "universal condemnation".
Mea culpa, I shouldn't have said "worse", but "more easily discouraged".
So I do think there's a big weakness with LLMs in that we don't quite have a handle on how to robustly or predictably reduce hallucinations like we can with human hallucinations and fabrications. But that's where I think the incentive of the editors/publishers come into play. Outlets that publish falsities by their human journalists lose credibility and can also lose lawsuits, which provide incentives for the people in charge to check the letters their human journalists generate before publishing them, and I see similar controls as being effective for LLMs.
Now, examples like Rolling Stone's A Rape on Campus article show that this control system isn't perfect, particularly when the incentives for the publishers, the journalists, and the target audience are all aligned with respect to pushing a certain narrative rather than conveying truth. I don't think AI text generators exacerbate that, though.
I also don't think it's possible for us to enter a world where we universally condemn media companies who publish AI slop, though, unless "slop" here refers specifically to lies or the like. Given how tolerant audiences are of human-made slop and how much cheaper AI slop is compared to that, I just don't see there being enough political or ideological will to make such condemnation even a majority, much less universal.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
You can strawman me in whatever way you prefer.
Nope, just trying to move the goalposts so we're at least on the same playing field.
Unfortunately you edited your comment though and now I've completely lost the context of our discussion. Maybe I'll wait next time for the second version of your argument.
The edit was for grammatical clarity. You remain free to assign to me any positions I have not taken as part of your goalpost moving.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
I don’t think most people honestly care unless the writer is in some sense the reason they’re reading the piece in the first place. For 90% of news or blogs it literally doesn’t matter. And even for boilerplate fiction (the kind people bring on vacation) that’s mostly read as a pastime or relaxation, they probably won’t care that much if it’s chatbot unless the publisher didn’t bother to clean up the disjointed phrases.
This place and the people who read sites like this are probably extreme outliers. We care if the stuff we read is good, or accurate, or has clunky prose or hallucinations. But this is a place full of highly educated, intelligent, thoughtful Readers. Most of us have probably read a nonfiction book in the last six months that isn’t related directly to work. We’ve read blogs for information. When we read news, it’s something we care about the accuracy for the article. A person who’s reading journal articles to fact check news reporting is going to easily find errors just because of information exposure. Someone who reads a lot of content is going to notice the slightly off phrase and check to see if it’s LLM slop.
Keep in mind that the median American reads rarely and when he does, it’s often below the 6th grade level. At such low literacy, tge difference between a random anuthor and a chatbot isn’t obvious. The difference between a bot and a human writing about news and politics isn’t obvious, because boilerplate news is written in a formulaic way with a formulaic style. To someone who reads a lot of it, word choices and odd phrases stick out. To people skim reading the news, it’s not obvious.
More options
Context Copy link
Claude Sonnet 3.6 outright changes their minds if well-prompted. The cat is out of the bag, the genie has left the bottle.
https://axisofordinary.substack.com/p/links-for-2025-04-30 - scroll down to the images
I wonder if ai is so persuasive because it's so sociopathic. It's willing to bend its own argument into whatever direction seems most fruitful. But most human arguers are limited by their own beliefs and principles. Though there are some deranged sociopaths out there who are capable of saying anything to win an argument.
You don't have to be a deranged sociopath to be willing to say anything to win an argument. Most normal people who go into debating clubs etc learn this.
More options
Context Copy link
More options
Context Copy link
Of redditors? I can believe it, never have I seen a more thoroughly shilled to and easily lead astray group of people.
I think covid proved that the majority of humans are in fact very easy to persuade. You don't even need to present any particularly persuasive arguments, you just need to be able to credibly present your position as the social consensus. So it's not just a "redditor" thing.
I don't think COVID stuff can really be used as proof of general human gullibility. "You should deal with COVID a certain way Or We'll All Die" has built-in life-or-death stakes, ie a much greater emotional valence than the average sweeping claim. There might even be specific drives about falling in line with herd behavior if there's a plague going around, if you want to get all evo-psych. To see how much weight the social consensus itself possesses, you would really want to look at something neutral, abstract, with no effects on people's everyday lives. A pure article of faith, divorced from any call to action or doomsaying.
The problem is that it is not true. Covid wasn't ”we'll all die”, it was just framed that way.
More options
Context Copy link
"I am the only guy who can deal with Trump", confidently stated, just swung the Canadian election -- so I don't think it's necessarily plague related.
Isn't this just a fad? Fads are a thing.
No, because the average fad isn't really making a factual claim. Someone who adopts the latest fashion trend isn't really believing anything that objective reasoning would show to be untrue.
More options
Context Copy link
Was the real message "I am the only guy who can deal with Trump" or was it more like "I am the only guy who can be trusted to want to deal with Trump"? I don't think Canadians thought Poilievre was weak and shit - I think they thought he was fundamentally sympathetic to someone who had suddenly become the enemy.
Maybe, I don't really know -- but that just kicks the cat down the road in that the only evidence for that is "Mark Carney says so"; Poillievre was talking just as tough as anybody else on Trump.
More options
Context Copy link
The Canadians in industrial/resource/economically productive areas voted nearly unanimously for the Cons (look at the map).
Those in areas not so blessed (Quebec, Atl. Canada, Ottawa), or those in service economy areas (Toronto, Vancouver) voted for the Libs.
The people who are actually going to be affected by a hostile stance very clearly don’t want “maximum reeeee about Trump” as Canadian foreign policy but, because they’re a numerical minority, what they want doesn’t matter. [Which is why the West is starting to think it should move on from “ballot box”.]
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Building in stakes for the sake of urgency and invoking emotional rather than deliberate reasoning is a cornerstone of many fraud and propaganda techniques, i.e. human gullibility exploits.
Yes, but the necessity of such techniques paints a different picture from the idea that humans are gullible in the sense of automatically going along with any social consensus. Buttons other than the sheer instinct to conform need to be pressed.
Gullibility is the tendency to be persuaded, which is method agnostic. If your idea of gullibility is tied to social conformity pressure alone, your concept needs to be expanded.
It's @Primaprimaprima who went from "humans are in fact very easy to persuade" to "you don't even need to present any particularly persuasive arguments, you just need to be able to credibly present your position as the social consensus" at the start of this comment chain. That is the claim I meant to dispute (or rather, the claim which I disputed could be proven by looking at COVID responses).
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
It’s not our fault that AI generated slop-Walz is just as serviceably mediocre as the real thing.
This made me laugh.
More options
Context Copy link
If anything, its better! I've never heard Walz be this willing to tell it straight.
More options
Context Copy link
More options
Context Copy link
I've started noticing AI generated video in the lead up to the Australian federal election. There was one on X floating around with Tucker Carlson dumping on the current Prime Minister. Really made me think about the need to curate my feed more.
Not a rhetorical question -- do many Australians know who Tucker is? And do many Australians care what he thinks? I thought he was only relevant in American politics.
Australians who are interested in politics are pretty much bound to follow American politics, at least in the vague outline. I'd guess that your average person on the street either doesn't know who Tucker Carlson is, or knows him only as some pundit in America. Among Australians who take an active interest in politics, I'd expect much higher recognition.
I would not expect Carlson's endorsement (or disendorsement) to have any significant impact on Australian politics, though. If anything, I expect that his endorsement would hurt a candidate. We have a federal election tomorrow where it looks like what would have been a very winnable election for the Coalition has turned into a disaster, substantially due to Trump. Trump and MAGA-style voices are widely hated over here and any association with them is more likely to harm than to hurt. It's not as bad here as in Canada, but it's still true, I think, that Trump has been a disaster for conservative parties throughout the wider Anglosphere.
I've read that the reason Trump's impact on Canada was a disaster was that there is no analog to blood-and-soil MAGA voters there, only what would be called in America progressives and "RINOs"/"boomercons". Is that why Trump has damaged conservatives in Australia?
We're largely missing the Borderers, and so they're culturally alien to us (so are American Descendants of Slaves, but for various reasons including good PR and relative invisibility there's less friction there).
We also have a much-weaker two-party system, so instead of being a faction with some amount of influence in our major right-wing parties, the alt-right has its own party (well, technically two parties, One Nation and
the Clive Palmer Partythe United Australia PartyTrumpet of Patriots, but the latter is a bad joke). There's no cordon sanitaire in Australia (there was one, like 25 years ago, but it fell apart); the Coalition (the neoliberal Liberals and rural-conservative Nationals) put One Nation above anybody besides themselves on their how-to-vote cards and they're willing to work with One Nation when they have to. But because One Nation's primary vote is quite a bit lower than that of the Coalition and they're not unusually-concentrated like the (hippie/SJ) Greens voters, they have no lower-house seats (though they do manage a few Senate seats, as the Senate is pseudo-proportional representation), so the Coalition mostly haven't had to (or have been impotent even with them).And yeah, as the others have mentioned there's a bit of an issue that Trump wasn't being all that friendly to Australia. Friendly fire isn't, and all that.
More options
Context Copy link
Sure there is- those are, counterintuitively, Liberal party voters. That is who all the jingoism was coming from.
If you're of an age sufficiently advanced that you remember that your parents actually took Canada seriously (i.e. you're 60+), you think that insults to Canadian 'sovereignty' are a big enough deal that you're going to latch onto whoever you perceive promises to do the most about it.
(Everyone else understands that this is a post-national country, so they don't care so much... if they can even afford to care in the first place.)
And that's not going to be some foreigner from Alberta (and, to the peoples of ON/QC/Atlantic, this is what he is) no matter what he otherwise says or does- because "he isn't aggressive enough against Trump" wasn't a legitimate criticism of him in the West (where he gained seats, unusually, in urban areas). I believed that not taking a suitably aggressive stance was still a mistake in the beginning, but now I'm starting to think that if he had the Liberals would have secured a majority simply because they are more likely to believe that Trump is directionally correct even if they disagree with the incidentals (cynically, it is in their socioeconomic interest to do that because "kill all new development, degrowth now" hurts their ability to accumulate wealth in a way it doesn't for Toronto retirees, federal government employees, and provinces that are already financial have-nots).
More options
Context Copy link
I don't think this is true - there seems to be a 20%ish vote for right-populists everywhere the electoral system permits it.
The problem is more simple - Trump is anti-Canadian, or at least was playing anti-Canadian during Q1 2025. If you were a blood-and-soil Canadian patriot, you would be anti-Trump because Trump is hurting Canadians with tariffs and does not appear to respect Canadian sovereignty. Right populists from different countries are only allied if their countries are committed to respecting each other's sovereignty.
More options
Context Copy link
Hm, I'm not sure how much I follow the categories that you're using. Australia doesn't really have what I would call a MAGA base. If you're looking for blood-and-soil types, well, firstly a lot of indigenous people are just straightforwardly blood-and-soil in their approach, but presumably they don't count, so secondly you're looking at One Nation and the nativists.
The thing is, the constituency for Australian nativism is somewhat complicated. Over 30% of Australia's population was born overseas, and because we have compulsory voting, all those people turn out. So you can't win an election just with people born here. Fortunately, the One Nation position isn't that immigrants simpliciter are bad - they tend to distinguish along the lines of culture, or less charitably, race. Nobody cares about English immigrants, for instance. They're where Australia comes from in the first place. Likewise we don't care about, say, New Zealand migrants, which are actually very common anyway. (Long story short, if you're an even slightly ambitious young New Zealander and seek opportunity, you come to Australia. And we're fine with that.) One Nation complained about being 'swamped by Asians' in the 90s, and today they're more likely to complain about Islamic immigration. The idea is hostility to people who don't share Australian values or a common cultural identity. Thus, for instance, around 70% of Australians are in favour of open borders with Canada, New Zealand, and Britain. We perceive ourselves as pretty closely connected to all of them.
(It will be interesting to see if demographic changes in Canada alter popular opinion on that, but I digress.)
At any rate, think of the right-wing 'blood and soil' position in Australia to not be 'Australians only', but rather 'Australians and the countries we like only', with the understanding that 'countries we like' means basically the white parts of the British Empire. People who take this view too explicitly will probably be accused of wanting to re-enact the White Australia Policy.
To Trump specifically -
I sometimes think of Australia as being the most America-like country that isn't America. (I would say we beat out Canada only because Canadian identity is specifically formed by not being American. They have a much more intentional sense of resistance against America.) However, there are still important differences, and I think the big one is probably that America has a very different idea of greatness or success to us. You couldn't have 'Make Australia Great Again' as a slogan because we don't have that kind of ambition or pride. That's not how we think of Australia. However, the American influence on our political tradition is significant - while structurally we are a Westminister democracy and the UK is the biggest influence on us, the framers of the Australian constitution read and were significantly (but not slavishly) inspired by the American effort. If you look into Australian patriotic writing from the late 19th and early 20th century, there is a strong feeling that we can use America as a model, or that Australia can be a kind of 'second America' (only loyal this time). Then you have to add to that, of course, the American alliance that has persisted since the Second World War, and we do just look to America for a lot.
This means that when anything in America reads well or sympathetically to the Australian electorate, politicians suddenly get keen on copying it; and likewise when anything in America reads unsympathetically, politicians need to struggle to distance themselves from it. In the past the Coalition has been more vocally pro-America than Labor, and have had close relationships with the Republican party. John Howard made a big show of his friendship with George Bush, for instance, and while it's been a bit more complicated since then, you can see how proud ScoMo was to be next to Biden. (And Biden apparently forgetting his name caused a minor scandal.)
So the issue for Peter Dutton in particular is - the Coalition has generally marketed itself as pro-America, or closer to America than Labor (which has instead been quietly proud of rebuilding our relationship with China). Dutton has also experimented a bit with American-style culture politics, and generally is perceived as a more 'American' politician.
And that seemed to be going well up until Trump lashed out with tariffs, and while we only received the lightest of American tariffs, Trump also refused to give us an exemption (which he had done in his first term). Add in that Trump's self-aggrandising, bullying style of politics plays very badly with an Australian culture that tends to prefer humility and self-deprecation and America's brand is currently in the toilet.
Dutton gets some of the overflow from the hatred of Trump. He and his party are too close to Trump and too close to America. When it comes to foreign policy, I suspect voters currently want a leader who will stand up to America.
More options
Context Copy link
Well, no, you wouldn't expect large constituencies of "make [another country] great again" voters in another country, unless we're talking about special cases like Christian Zionists supporting Israel for religious reasons. Beyond that, why would one expect even right-wing Canadians to feel particularly positive about the guy who talks about annexing Canada and has just slapped Canada with punitive tariffs for... something?
Insofar as I've seen, the sort of Canadians who would actually support American annexation or at least be OK with it would be either disaffected forumlords who treat politics as an abstraction, general fringe loons, or recent immigrants - I remember seeing a post indicating that the Indian immigrants in Canada would be more likely to support annexation than the born Canadians.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Not a lot do except for the 'very online' that enjoy US and geo-politics.
However one of the populist parties is called 'The Trumpet of Patriots' and is clearly trying to piggyback off of Trump's success in the US. I'd say more of the voters for that party would than average.
Edit: Should make it clear that TOP is a minor party that isn't expected to pick up many seats. Its run by a dodgy billionaire with a checkered history.
A billionaire who is pretty much in it for the entertainment value. He's rich and he likes being on TV and having his face on billboards.
Trumpet of Patriots in general is a good case study in why Trumpism doesn't work in Australia. They are copying Trump-style campaigning while making zero adaptation to the local context and it is pretty miserable.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
You know full well that the reaction is to the snippet of his remarks "code talk to white guys" and that was entirely real. Shame Nate Silver spread slop, I respect him a bit less for this, but people are getting older and more careless and AI more convincing. Let's wait until tomorrow for him to correct or retract the slop link. Having never read Yahoo Entertainment I have no idea if it's all slop or was semi-legitimate entertainment journalism at some point but I'm going to be extremely skeptical if I see it again.
Are the hallucinated parts actually making a big stir somewhere? They aren't in a cursory glance in the replies to Silver.
Yahoo aggregates articles from other sites, with the source displayed in the top left corner. This one is from "Where is the Buzz".
More options
Context Copy link
More options
Context Copy link
Yep. That just looked look 100% organic free range human journalist slop to me on first read. TBH even though there are a few suspicious passages it doesn't seem super AI-sloppish to me even when I'm keeping in mind that it has been accused of such.
"Quoted stuff the person didn't say" is a pretty strong tell, but if I watched the talk, why would I also want to read clickbait journalism slop about the talk I just watched?
Your opinion of journalists is too low. Journalists are rabid partisans, but they're generally not very stupid. But they pretend to be stupid quite often in order to serve their side.
I don't think they're stupid, but I do think they're under a lot of pressure to churn out as much content as they can as quickly as they can.
More options
Context Copy link
More options
Context Copy link
That raises a question: Are journalist-written articles more likely to trigger AI detectors?
I know I'm the "this is an empirical question we can just go check" guy but I don't actually have the time to do this in the foreseeable future. Anyone else want to take a crack at it?
More options
Context Copy link
The corresponding New York Post and Telegraph articles for this story both score a solid 0% AI / 100% human on zerogpt and gptzero
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
I probably wouldn't have guessed that this article was almost purely generated by AI if I hadn't been primed on it beforehand. Looking at it with that priming, I'm still not convinced that it was a pure copy-paste GPT job, though certainly it's filled with phrasing that, having been primed, strike me as being from an LLM, such as "While some applauded the self-deprecating humor, others criticized the segment for reinforcing cultural stereotypes" or "As speculation mounts over the 2028 Democratic field, Walz offers a glimpse into his political philosophy for the years ahead." Is there any direct evidence of it being LLM-generated?
But more to the point, I don't see why most people would care if this was purely AI generated, other than perhaps this author Quincy Thomas's employers and his competitors in the journalism industry. Particularly for what seems to be intended to be a pretty dry news article presenting a bunch of facts about what some politicians said. This isn't some personal essay or a long-form investigative report or fiction (even in those, I wouldn't care if those were purely LLM-generated as long as they got the job done, but I can see a stronger case for why that would matter for those). This kind of article seems exactly like the kind of thing that we'd want LLMs to replace, and I'd just hope that we could get enough of a handle on hallucinations such that we wouldn't even need a human like Quincy Thomas to verify that the quotes and description of events actually matched up to reality before hitting "publish."
Once some fairly reputable news outlet gets sued for defamation for publishing some hallucination that was purely LLM generated and failing to catch it with whatever safeguards that are in place, that's something I'd be interested to see how it plays out. At the rate things are going, I wouldn't be surprised if that happened in the next 5 years.
Because it's blatantly false garbage.
Quincy Thomas is not a human. It's obviously a fake name for some third world scam grifter who is shoveling this garbage
From a cursory search through the auto-generated transcript, it does seem to me like that that quote was made up. That does seem worth caring about. It's too bad that it's not defamatory, since it probably won't trigger some lawsuit or other major controversy, but perhaps a controversy could be created if someone decided to publicize this.
Seems like Yahoo's fact checking/editing department isn't built to handle its writers using LLMs. I still don't see why I would care about LLM usage if a journalism outlet had the proper controls for factual information. The problem isn't that it's AI generated, it's that it's false.
I'm pretty sure that's not a thing -- it says at the bottom that the 'article' is reposted directly from "WhereIsTheBuzz.com" -- which looks about like what you'd expect. Highly unlikely that Yahoo is fact-checking anything; their role as a slop aggregator could use some more scrutiny I guess, but this article doesn't seem much different than their standard run of human slop to me.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
Two thoughts: One, the quotes on which the article is based are real, however. While I obviously have a dim view about AI-only articles, it's not like the opinion on which it is based is wholly fabricated. There's even a link to the original plagiarized tweet that served as the seed. So while the specific words might be AI-generated, the opinions and facts behind them still strike me as fairly representative. In that regard it's totally fine to engage with the post, and it isn't all that different than a human's post (other than the AI is probably a little less careful with their specific phrasing... but an unskilled human writer might make similar mistakes).
Two, my impression is Silver, who has been on a few-month-long tear recently on Twitter, is just doing regular engagement-bait confirmation bias stuff. Rather than say "he's been captured" I would say let's look at the more likely reason - it's an emotional, not intellectual, reaction. He's frustrated about Democrats and their nonsensical, misguided strategies in the last few years. Many agree with him. Of course, these posts happen more frequently when you get twitter-brain. Even data scientists are emotionally vulnerable to human network effects, where our brains are incapable of realizing twitter has enormous selection bias. Nope, our brains don't care, they find it very hard to correct for the bias. Much like how your sense of humor is developed by the people around you and their reactions. Is that population-representative? Often, no. But the brain doesn't care because the brain (often rightfully) assumes you care more about the opinions of those immediately surrounding you then the larger population. It's just that adjacent people have been hijacked and replaced with twitterati.
So, the good news is that I believe pretty strongly that if you remove twitter, you remove the problem and things can revert. The other good news is that this isn't actually all that abnormal. If you treat twitter-Silver like a woman who doesn't want her problem solved, but rather just wants some listening and sympathy (I mean that in the most charitable way), you see this is fine and just regular human things. While some good actual-ideas discussion happens on twitter, it's rare and tends to get overshadowed.
Nope, the quotes are fake. See my edited OP.
More options
Context Copy link
More options
Context Copy link
The AI slop article in question is describing real events. For Silver, there simply was no need to go digging for a better source for something he had already heard about when he was just posting on X.
it would take literally 5 seconds to plug in the headline to Google and get a real article. Here's one right here: https://archive.is/KduY8
OK, but who gives a shit? Would it change his point? Was the article he linked fundamentally false?
Yes, see my edited OP
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link