I wouldn’t argue with someone who said reasoning models are a substantial advance

Oh, I would.

I’ve seen people say stuff like “you can’t disagree the models have rapidly advanced” and I’m just like yes I can, here: no they didn’t. If you’re claiming they advanced in any way please show me a metric by which you’re judging it. Are they cheaper? Are they more efficient? Are they able to actually do anything? I want data, I want a chart, I want a proper experiment where the model didn’t have access to the test data when it was being trained and I want that published in a reputable venue. If the advances are so substantial you should be able to give me like five papers that contain this stuff. Absent that I cannot help but think that the claim here is “it vibes better”.

If they’re an AGI believer then the bar is even higher, since in their dictionary an advancement would mean the models getting closer to AGI, at which point I’d be fucked to see the metric by which they describe the distance of their current favourite model to AGI. They can’t even properly define the latter in computer-scientific terms, only vibes.

I advocate for a strict approach, like physicist dismissing any claim containing “quantum” but no maths, I will immediately dismiss any AI claims if you can’t describe the metric you used to evaluate the model and isolate the changes between the old and new version to evaluate their efficacy. You know, the bog-standard shit you always put in any CS systems Experimental section.

It’s kind of a shame to have to downgrade Gary to “not wrong, but kind of a dick” here. Especially because his sneer game as shown at the end there is actually not half bad.

Gary Marcus has been a solid source of sneer material and debunking of LLM hype, but yeah, you’re right. Gary Marcus has been taking victory laps over a bar set so so low by promptfarmers and promptfondlers. Also, side note, his negativity towards LLM hype shouldn’t be misinterpreted as general skepticism towards all AI… in particular Gary Marcus is pretty optimistic about neurosymbolic hybrid approaches, it’s just his predictions and hypothesizing are pretty reasonable and grounded relative to the sheer insanity of LLM hypsters.

Also, new possible source of sneers in the near future: Gary Marcus has made a lesswrong account and started directly engaging with them: https://www.lesswrong.com/posts/Q2PdrjowtXkYQ5whW/the-best-simple-argument-for-pausing-ai

Predicting in advance: Gary Marcus will be dragged down by lesswrong, not lesswrong dragged up towards sanity. He’ll start to use lesswrong lingo and terminology and using P(some event) based on numbers pulled out of his ass. Maybe he’ll even start to be “charitable” to meet their norms and avoid down votes (I hope not, his snark and contempt are both enjoyable and deserved, but I’m not optimistic based on how the skeptics and critics within lesswrong itself learn to temper and moderate their criticism within the site). Lesswrong will moderately upvote his posts when he is sufficiently deferential to their norms and window of acceptable ideas, but won’t actually learn much from him.

Very practical no notes

This ties back into the recurring question of drawing boundaries around “AI” as a concept. Too many people just blithely accept that it’s just a specific set of machine learning techniques applied to sufficiently large sets of data. This in spite of the fact that we’re several AI “cycles” deep where every 30 years or so (whenever it stops being “retro”) some new algorithm or mechanism is definitely going to usher in Terminator II: Judgement Day.

This narrow frame focused on LLMs still allows for some discussion of the problems we’re seeing (energy use, training data sourcing, etc) but it cuts off a lot of the wider conversations about the social, political, and economic causes and impacts of outsourcing the business of being human to a computer.

Highlights from the comments: @wjpmitchell3 writes,

Actual psychology researcher: the problem with IQ is A) We don’t really know what it’s measuring, B.) We don’t really know how it’s useful, C.) We don’t really know how context-specific it is, D.) When people make arguments about IQ, it’s often couched around prejudiced ulterior motives. No one actually cares about IQ; they care about what it’s a proxy measure of and we don’t have good evidence yet to say “This is a reliable and broadly-encompassing representation of intelligence.” or whatever else, so if you are trying to use IQ differences to say that there are race differences in intelligence, you have no grounds. The best you can say is there are race differences in this proxy measure that we’re still trying to understand. It’s dangerous to use an unreliable and possibly inaccurate representation of a phenomena to make policy changes or inform decisions around race. The evidence threshold has to be extremely high because we’re entering sensitive ethical spaces, which is something that rationalist don’t do well in because their utilitarian calculus has difficulty capturing the intangibles.

@arnoldkotlyarevsky383 says,

Nothing wrong with being self educated but she comes across as being not as far along as you would want someone to be in their self-education before being given a platform.

@User123456767 observes,

You can kind of tell she grew up as a Calvinist because she still seems to think she’s part of the elect she’s just replaced an actual big G God with some sort of AI God.

@jaredsarnie3712 begins,

I feel like so much of what she says boils down to finding bizarre hypothetical situations where child sexual abuse is morally acceptable.

And from @Fruuuuuuuuuck:

Doomscroll gooner arc

God I remember having to cite RFC at other vendors when I worked in support and it was never not a pain in the ass to try and find the right line that described the appropriate feature. And then when I was done I knew I sounded like this even as I hit send anyway.

Comic Book Guy energy

So, Zelenskyy is goth?

If anybody doesn’t click, Cremieux and the NYT are trying to jump start a birther type conspiracy for Zohran Mamdani. NYT respects Crem’s privacy and doesn’t mention he’s a raging eugenicist trying to smear a poc candidate. He’s just an academic and an opponent of affirmative action.

Ye it was a real “oh fuck I recognise this nick, this cannot mean anything good” moment

This meeting could have been a text document of plausible sounding jibberish nobody needs to read.

Don’t make me tap the sign:

no gods, no kings

including possible effects on the protocols from issues like such as AI fuzzing attempts, to social engineering by AI’s,

“You know those massive problems we already had going back decades? Well what if the same problems happened in the future but with the letters ‘A’ and ‘I’ prepended? Scary!”

through to how we deal with and approach and facilitate Avatars and Agents.

Gosh darn it don’t tell me LLM hype is going to ruin the existing definition of “agent” already well established in web standards.

Alright that’s it: anime streaming needs to return to fansubbing

Fansubs are openly doing it for the love of the anime, so chances are they’d avoid AI slop like the plague (though the CHUDs would be okay with ChatGPT subs if it meant avoiding The Woke^tm)

(note: this link contains a skintight anime bosom so don’t open it in front of your boss unless your boss is chill)

Good thing I’m a fucking NEET, then