Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.
This is not debate club. Unless it’s amusing debate.
For actually-good tech, you want our NotAwfulTech community
TIL some rats have started a literal monastery to try to defeat the robot god with good ole religion (well, Zen buddhism)
here's a mildly critical view that apparently still believes the approach has legs
https://www.lesswrong.com/posts/ENCNHyNEgvz9oo9rr/briefly-on-maple-and-the-broader-community
I note in passing that there seems to be a mild upsurge in religious-friendly posts on LW lately.
That opening has a strong 'omg what the fuck happened there, and why are you still friends with these people if it is that bad' vibes.
Lots of ex Maplers I've talked to are variously angry, some (newly) traumatized, confused, etc., but the vibe has generally been "gosh it's fucking complicated
Lot of people told me they were mad, but I got the vibe it was complicated. lol what...
The product of monasteries is saints, at least in small quantities.
Wrong, it is beer.
I haven't seen anti-safetyist arguments that actually address the technical claims made by Eliezer etc.
I agree with him there. But hard to make arguments against something which doesn't exist. ;)
it will take some very unusual kind of virtue and skill
Love how we went from 'you need to learn rationality, and be aware of your biasses' to you need to have virtue. And ignore the screaming in the backgrounds, that is just the academics who studied ethics doing their normal thing again.
Anyway the rest gets pretty dark pretty quickly, and I just see red flags (for people who believe in data this devolves very quickly in just going 'people are prob better off due to this, the new trauma doesn't count because of preexisting conditions, consent and trauma always happening'). And this was the article they wrote not wanting to harm the project.
Wait one more remark:
Furthermore I think that probably with the exception of the one actual AI researcher there, people at Maple basically don't understand what AI is
Hahahaha, perfect.
It also reminds me of the interview with Metz where they got mad Metz used religious terms. (you know, last week).
well, Zen buddhism
Yeah, this is the Valley after all. Some have used Buddhism as a building block for constructing “metarationality”.
I have a degree of appreciation for Chapman because he was willing to more-or-less call out Yuddite rationalism as a failure and start to gently guide people away from it. But I also came to the conclusion that his whole project has never fully escaped the self-aggrandizement/self-importance inherent to the rats. That ultimately leads to the performative humility and "radical acceptance" that make so many attempts at appropriating non-Western religions to US culture ring completely hollow.
Broadly, the whole TPOT/post-rationality/meta-rationality thing still stinks like a bunch of people who thought advanced degrees and/or advanced technical skills would earn them a lot more compensation and social status than they actually ended up with, and are still dead-set on getting all that by hook or by crook.
I hadn't heard of MAPLE before, is it tied to lesswrong? From the focus on AI it's at least adjacent to it... so I'll add that to the list of cults lesswrong is responsible for. So all in all, we've got the Zizians, Leverage Research, and now Maple for proper cults, and stuff like Dragon Army and Michael Vassar's groupies for "high demand" groups. It really is a cult incubator.
Michael Hiltzik in LATimes: "Say farewell to the AI bubble, and get ready for the crash"
Fun quote:
The rest of [AI 2027], mapping a course to late 2027 when an AI agent “finally understands its own cognition,” is so loopily over the top that I wondered whether it wasn’t meant as a parody of excessive AI hype. I asked its creators if that was so, but haven’t received a reply.
And because it's the LA Times, there's a chatbot slop section at the bottom to provide false balance.
I'm enjoying the mood today. We're all looking for what the next Big Dumb Thing will be that we'll be dunking on next year, like we're browsing the dessert menu at a fancy restaurant.
On top of that, there's clear signs that we've grown quite an audience from dunking on AI. Ed Zitron reached 70k subscribers just a couple weeks ago, and Pivot to AI is at nearly 9k on YouTube.
If and when the next Big Dumb Thing comes along, chances are we're gonna have a headstart against the hucksters.
In other news, bodhidave reported a case of Google AI and ChatGPT making identical citation fuck-ups:
Even if they aren't actively relying on each other here I would assume that we're reaching a stage where all of the competing LLMs are using basically the entire Internet as their training data, and while there is going to be some difference based on the reinforcement learning process there's still going to be a lot of convergence there.
Plus, there's the hefty amount of AI slop that's been shat onto the Internet over the years, plus active attempts to sabotage LLM datasets through tarpits like Iocaine and Nepenthes, and media-poisoning tools like Glaze and Nightshade.
So, if and when model collapse fully sets in, its gonna hit all of them at once. Given that freshly trained LLMs are gonna be effectively stillborn, if ChatGPT et al. collapse, it'll likely kill LLMs as a tech for at least the next ten years.
The true meaning of "pooping back and forth forever"
sick reference. I don’t even know how I knew this.
heh, I'm too basic to make a deep cut like that, I just went through a Cards Against Humanity phase in grad school!
Ed Zitron's planning to hold AI boosters to account:
Well if the bubble pops he will have to pivot to people who pivot. (That is what is going to suck when to bubble pops, so many people are going to lose their jobs, and I fear a lot of people holding the bags are not the ones who really should be punished the mosts (really hope not a lot of pension funds bought in). The stock market was a mistake).
I imagine it'll be a pretty lucrative pivot - the public's ravenous to see AI bros and hypesters get humiliated, and Zitron can provide that in spades.
Plus, he'll have a major headstart on whatever bubble the hucksters attempt to inflate next.
the hucksters attempt to inflate next.
Quantum, it has already started: https://www.schneier.com/blog/archives/2025/07/cheating-on-quantum-computing-benchmarks.html
Y'know, I was predicting at least a few years without a tech bubble, but I guess I was dead wrong on that. Part of me suspects the hucksters are gonna fail to inflate a quantum bubble this time around, though.
Quantum computing is still too far out from having even a niche industrial application, let alone something you can sell to middle managers the world over. Anybody who day-traded could get into Bitcoin; millions of people can type questions at a chatbot. Hucksters can and will reinvent themselves as quantum-computing consultants on LinkedIn, but is the raw material for the grift really there? I'm doubtful.
Quantum computing isn't just hard, it's hadamard
Hucksters can and will reinvent themselves as quantum-computing consultants on LinkedIn, but is the raw material for the grift really there? I’m doubtful.
By my guess, no. AI earned its investor/VC dollars by providing bosses and CEOs alike a cudgel to use against labour, either by deskilling workers, degrading their work conditions, or killing their jobs outright.
Quantum doesn't really have that - the only Big Claim™ I know it has going for it is its supposed ability to break pre-existing encryption clean in half, but that's near-certainly gonna be useless for hypebuilding.
New Atlantic article regarding AI, titled "AI Is a Mass-Delusion Event". Its primarily about the author's feelings of confusion and anxiety about the general clusterfuck that is the bubble.
better, or equivalent to, a mass defecation event?