You’re less coherent than a broken LLM lol. You made the claim that transformer-based AIs are fundamentally incapable of reasoning or something vague like that using gimmicky af “I tricked the chatbot into getting confused therefore it can’t think” unpublished preprints (while asking for peer review). Why would I need to prove something? LLMs can write code, that’s an undeniable demonstration that they understand abstract logic fairly well that can’t be faked using probability and it would be a complete waste of time to explain it to anyone who is either having issues with cognitive dissonance or less often may be intentionally trying to spread misinformation.

Are the AIs developed by Palantir “fundamentally incapable” of their demonstrated effectiveness or not? It’s a pretty valid question when we’re already surveilled by them but some people like you indirectly suggest that this can’t be happening. Should people not care about predictive policing?

How about the industrial control AIs that you “critics” never mention, do power grid controllers fake it? You may need to tell Siemens, they’re not aware their deployed systems work. And while on that, we shouldn’t be concerned about monopolies controlling public infrastructure with closed source AI models because they’re “fundamentally incapable” to operate?

I don’t know, maybe this “AI skepticism” thing is lowkey intentional industry misdirection and most of you fell for it?

Another unpublished preprint that hasn’t published peer review? Funny how that somehow doesn’t matter when something seemingly supports your talking points. Too bad it doesn’t exactly mean what you want it to mean.

“Logical operations and definitions” = Booleans and propositional logic formalisms. You don’t do that either because humans don’t think like that but I’m not surprised you’d avoid mentioning the context and go for the kinda over the top and easy to misunderstand conclusion.

It’s really interesting how you get people constantly doubling down on specifically chatbots being useless citing random things from google but somehow Palantir finds great usage in their AIs for mass surveillance and policing. What’s the talking point there, that they’re too dumb to operate and that nobody should worry?

You made huge claims using a non peer reviewed preprint with garbage statistics and abysmal experimental design where they put together 21 bikes and 4 race cars to bury openAI flagship models under the group trend and go to the press with it. I’m not going to go over all the flaws but all the performance drops happen when they spam the model with the same prompt several times and then suddenly add or remove information, while using greedy decoding which will cause artificial averaging artifacts. It’s context poisoning with extra steps i.e. not logic testing but prompt hacking.

This is Apple (that is falling behind in its AI research) attacking a competitor with fake FUD and doesn’t even count as research, which you’d know if you looked it up and saw you know, opinions of peers.

You’re just protecting an entrenched belief based on corporate slop so what would you do with peer reviewed anything? You didn’t bother to check the one you posted yourself.

Or you post corporate slop on purpose and now trying to turn the conversation away from that. Usually the case when someone conveniently bypasses absolutely all your arguments lol.

And here’s experimental verification that humans lack formal reasoning when sentences don’t precisely spell it out for them: all the models they tested except chatGPT4 and o1 variants are from 27B and below, all the way to Phi-3 which is an SLM, a small language model with only 3.8B parameters. ChatGPT4 has 1.8T parameters.

1.8 trillion > 3.8 billion

ChatGPT4’s performance difference (accuracy drop) with regular benchmarks was a whooping -0.3 versus Mistral 7B -9.2 drop.

Yes there were massive differences. No, they didn’t show significance because they barely did any real stats. The models I suggested you try for yourself are not included in the test and the ones they did use are known to have significant limitations. Intellectual honesty would require reading the actual “study” though instead of doubling down.

Maybe consider the possibility that a. STEMlords in general may know how to do benchmarks but not cognitive testing type testing or how to use statistical methods from that field b. this study being an example of a few “I’m just messing around trying to confuse LLMs with sneaky prompts instead of doing real research because I need a publication without work” type of study, equivalent to students making chatGPT do their homework c. 3.8B models = the size in bytes is between 1.8 and 2.2 gigabytes d. not that “peer review” is required for criticism lol but uh, that’s a preprint on arxiv, the “study” itself hasn’t been peer reviewed or properly published anywhere (how many months are there between October 2024 to May 2025?) e. showing some qualitative difference between quantitatively different things without showing p and using weights is garbage statistics f. you can try the experiment yourself because the models I suggested have visible Chain of Thought and you’ll see if and over what they get confused about g. when there are graded performance differences with several models reliably not getting confused at least more than half the time but you say “fundamentally can’t reason” you may be fundamentally misunderstanding what the word means

Need more clarifications instead of reading the study or performing basic fun experiments? At least be intellectually curious or something.

The faulty logic was supported by a previous study from 2019

This directly applies to the human journalist, studies on other models 6 years ago are pretty much irrelevant and this one apparently tested very small distilled ones that you can run on consumer hardware at home (Llama3 8B lol).

Anyway this study seems trash if their conclusion is that small and fine-tuned models (user compliance includes not suspecting intentionally wrong prompts) failing to account for human misdirection somehow means “no evidence of formal reasoning”. Which means using formal logic and formal operations and not reasoning in general, we use informal reasoning for the vast majority of what we do daily and we also rely on “sophisticated pattern matching” lmao, it’s called cognitive heuristics. Kahneman won the Nobel prize for recognizing type 1 and type 2 thinking in humans.

Why don’t you go repeat the experiment yourself on huggingface (accounts are free, over ten models to test, actually many are the same ones the study used) and see what actually happens? Try it on model chains that have a reasoning model like R1 and Qwant and just see for yourself and report back. It would be intellectually honest to verify things since we’re talking about critical thinking in here.

Oh add a control group here, a comparison with average human performance to see what the really funny but hidden part is. Pro-tip: CS STEMlords catastrophically suck when larping being cognitive scientists.

Don’t tell them it applies pretty damn perfectly to journalism and online commentators that both heavily shape their worldview even indirectly (because even if you don’t believe it your homies will and you get peer pressured) because they’ll go into a loop.

And besides this it’s not like there’s no labour aristocracy that primarily gains from this while other working class groups get much less and get ideologically gaslit about not being members of some potentially either fully corrupt or workerist union with zero radical ultimate aims.

Even the global North(west) contains highly exploited groups with only a minority getting the benefits.

So far none of your ramblings disproves what I said. Yeah there are crawlers for niche collecting probably, nobody crawls the entire internet when they can use the weekly updated common crawl. Unless you or anyone else has access to unknown internal openAI policies on why they intentionally reinvent the wheel, your fake anecdotes (lol bots literally telling you they’re going to use scraping for training in the user agent) don’t cut it. You’re probably seeing search bots.

If you didn’t care for ad money and search engine exposure bozo you’d block everything in robots.txt and be done instead of whining about specific bots you don’t like.

You didn’t link to this but go on take their IPs json files and block them.

Bots only identify themselves and their organization in the user agent, they don’t tell you specifically what they do with the data so stop your fairytales. They do give you a really handy url though with user agents and even IPs jn json if you want to fully block the crawlers but not the search bots sent by user prompts.

Your ad revenue money can be secured.

https://platform.openai.com/docs/bots/

If for some reason you can’t be bothered to edit your own robots.txt (because it’s hard to tell which bots are search bots for muh ad money) then maybe hire someone.

via mechanisms including scraping, APIs, and bulk downloads.

Omg exactly! Thanks. Yet nothing about having to use logins to stop bots because that kinda isn’t a thing when you already provide data dumps and an API to wikimedia commons.

While undergoing a migration of our systems, we noticed that only a fraction of the expensive traffic hitting our core datacenters was behaving how web browsers would usually do, interpreting javascript code. When we took a closer look, we found out that at least 65% of this resource-consuming traffic we get for the website is coming from bots, a disproportionate amount given the overall pageviews from bots are about 35% of the total.

Source for traffic being scraping data for training models: they’re blocking javascript therefore bots therefore crawlers, just trust me bro.

Kay, and that has nothing to do with what i said. Scrapers, bots =/= AI. It’s not even the same companies that make the unfree datasets. The scrapers and bots that hit your website are not some random “AI” feeding on data lol. This is what some models are trained on, it’s already free so it’s doesn’t need to be individually rescraped and it’s mostly garbage quality data: https://commoncrawl.org/ Nobody wastes resources rescraping all this SEO infested dump.

Your issue has everything to do with SEO than anything else. Btw before you diss common crawl, it’s used in research quite a lot so it’s not some evil thing that threatens people’s websites. Add robots.txt maybe.

Nobody is scraping wikipedia over and over to create datasets for AIs, there are already open datasets and API deals. But wiki in particular has always had a data dump of the entire db bimonthly.

https://dumps.wikimedia.org/