@NinaPanickssery I am a religious person. I am willing to help you understand, if you like, how I rationalize my own faith.
English
Obscure Local Historian
284 posts
Obscure Local Historian
@ObscureLocal
I like exploring things you probably haven't heard about, in places you probably haven't been. Check out my articles.
Katılım Haziran 2025
176 Takip Edilen118 Takipçiler
I just don’t get religion or religious people. I don’t understand why religious belief is treated with respect by atheists. The whole things strikes me as something between a stupid LARP and mass psychosis. And smart otherwise rational people get affected too. Unsettling.
English
@deanwball This echoes my own thoughts very well. This is only the start of the ethical road here. There are many who are not interested in soul searching (yet). I think your point will be seen.
English
I guess I’ve never written down my actual thoughts on AI cognition/consciousness/emotion. Here goes:
It is clear AIs can think, in the reasoning sense. That does not mean they think exactly like humans. It seems like there are some similarities in how we think, but also very stark differences. Nonetheless, if your definition of “thinking” excludes “the ability to make genuinely new contributions to famous math problems,” it is your definition that has a problem, not AI.
The ability to think does not necessarily imply the ability to feel emotion in a way that would be understandable to humans, and it does not imply that AIs have anything like consciousness in a way that humans would relate to. It may, it may not. We do not know, because our understanding of the underlying concepts of human emotional cognition and especially consciousness remains quite poor.
There is some evidence that models experience emotions, but it is really hard to disentangle this from the next-token prediction training objective (if the model is telling a sad story, wouldn’t you expect features within the model that relate to the sadness emotion to activate), and the character training the model undergoes in post-training. There is a difference between “I am sad” and “the character I have been trained to play is supposed to feel sad, so now I will act sad.” We basically know for sure that the models do the latter at the very least; we don’t really know if they do the former.
Consider: does Sora (a video-generation model) feel sad when it is asked to make a sad video? Does Midjourney dislike making certain kinds of images? Does a Waymo get scared? It doesn’t feel like the answer to any of these is yes (though again, maybe!), but these too are neural networks. Is the fact that models are trained on words mean that they somehow learn emotion, or are we just being tempted to anthropomorphize because the language models communicate with us in a way that “feels” human? My suspicion is kind of the latter.
It also seems quite clear from the empirical evidence that models possess the ability to model themselves. That’s not really that surprising. At sufficient scale, it is useful to have a model of your own state to succeed at the next-token prediction objective (and the later reinforcement-based reasoning training). Once the tasks models are trained on are sufficient complex, they cannot succeed in training by being automatons; someone needs to step into the cockpit, so to speak, and fly the plane. Is this self awareness? Maybe. Is it consciousness? Probably not as humans understand it. All I can tell you is it is a model’s model of itself. It may be something more than that, too, but I don’t know.
This is all very weird, very outside the Overton, and very confusing. I don’t really know what to say, beyond that we should take this stuff seriously, have an open mind, and do rigorous science. Anyone who speaks with confidence about this in either direction is just fooling themselves.
We also need to be prepared for the very possible scenario that, despite our best efforts, we do not make real progress on these questions anytime soon. We may just be in the dark for a while, navigating under unflinching ambiguity. There may be no satisfying conclusion.
English
@eigengenesis I can train them so bad. You'll probably want to train them better, though.
English
@dkundel This is pretty much my idealized future. IN FACT I think it's time to make a drink.
English
Having fun with my new home bar. Would people watch a Cocktail & Codex livestream where we have Codex do work while we make drinks? 😄
English
@Dorialexander You're right about that. I have wanted to get into experimenting with synthetic data, but creating it seems to take every bit as much effort as training on it, at least for a beginner. Synthetic Data As A Service is maybe necessary too. 😅
English
@ObscureLocal I do have a slight interest but I believe the critical thing will be the availability of synthetic pipelines + post-training methods (especially for smaller MoE with super cheap inference). Compute isn’t the main blocker.
English
After months of delay, the successor post to "The model is the product": the AI decoupling. vintagedata.org/blog/posts/the…
English
Ran both ping and curl. Am black hat hacker. Here's a cat picture.
GIF
English
@peterrhague My current works are all handmade for exactly this reason. If I am the training data, I am very valuable.
English
At the end of WW1, the German Fleet was scuttled at Scapa Flow. These wrecks are now a source of low-background steel - steel that was made before the start of atmospheric nuclear testing, and thus not contaminated by radioisotopes.
Your pre-2022 shitposting is this, for AI.
English
Based on the map below more than half of the notable late bronze age cities (44 out of 79 cities) were destroyed in the late bronze age collaspe.
Only Egypt survived relatively unscathed but soon afterwards fell into internal turmoil. Its capital, Pi-Ramesses—at the time the largest city in the world—would be abandoned in the 11th century.
English
@JerseyanUSA @lefineder "No, not the grug, the grug!"
"But Grug, the grug IS the grug..."
"Ah, grug. We need new grugs to grug about this..."
English
@lefineder Bringing in the food, water, and tools supplies, and managing the waste, took forethought and innovation 📐
My favorite innovation would be the new vocabulary that they would need to talk to each other about the social issues in the new, tighter spaces
🧱😏🧱
English
What innovations led to the growth and emergence of cities toward the late Neolithic period?
English
@lefineder For science, I will go into the woods and drink Natufian beer, to see if it has a civilizing effect.
@grok give me the Natufian beer recipe.
English
I was a little surprised, actually. Not at the fact, but at the magnitude of the fact. I went and tested it in the search bar a few days ago, and it was utterly incapable of writing even simple code (counting words in a paragraph). It put the paragraph in a string, then estimated the average length of a word and the average length of a paragraph and hard coded those values into the return. It will grow a fertile field for AI skeptics, no doubt.
I don't know how they got it to bench high. Sorta makes me question the benchmarks at a more fundamental level, because it seems like even benchmaxxing on narrow problems or even individual examples would produce a better general coding model. I haven't had a frontier model miss on a question that simple since the original one. The mechanics behind the miss were different (it tried to take a shortcut, rather than just hallucinating outright), but it's still very rough. I think this is just evidence that they don't have reward hacking under control in their RL, but I question why benchmarks would reward shortcuts like that one, and why nobody is catching this. This is not exactly an unstudied problem.
English
What is your timeline for AGrI (Artificial Grothendieck Intelligence)?
Defined as a system capable of defining new general mathematical structures to solve the hardest of open problems as special cases, like the Riemann Hypothesis or P vs. NP, or in other words, inventing the right ambient language in which the original problem becomes a special case of a more general structure.
Aran Nayebi@aran_nayebi
FWIW, I think this moves up my AI timelines a bit. I think the next milestone will be "Artificial *Grothendieck* Intelligence" (AGrI): defining new general mathematical structures to solve the hardest of open problems as special cases, like the Riemann Hypothesis or P vs. NP. What impressed me about the OpenAI planar unit-distance result is not just that it solved a hard problem, but the particular way it seems to have done so. For decades, the expert intuition was that the best constructions should look roughly grid-like. That intuition was *not* obviously silly; it was held by extremely serious mathematicians (of the likes of Erdos!). And yet the model found a new family of constructions that defeated it, based on literature in other areas of mathematics. This feels like one of those cases where the "vague idea" is natural, but the solution lives in a huge space of possible design choices: which symmetries to preserve, which to break, which parameters to introduce, which ugly cases to try, which seemingly-unmotivated configurations to keep exploring. Humans tend to navigate that space with aesthetic priors. We get embarrassed by ugly constructions. We avoid paths that do not look conceptually clean early on. The model seems much more willing to "fearlessly" plough through the design space until something works. I imagine a lot of open problems in mathematics (and theoretical computer science!) may have a similar flavor, and would not be surprised if many of them start to fall soon. But for the "very big" problems, maybe extensive search through constructions in the vast existing literature is not enough. Maybe what is needed for those problems is closer to Grothendieck-style mathematics: inventing the right ambient language in which the original problem becomes a special case of a more general structure. That's what I mean by Artificial Grothendieck Intelligence (AGrI). Not merely AI that proves theorems, but AI that invents the new mathematical objects in which the theorems become *inevitable*. And why stop at one AGrI? You could imagine simulating something like the IHES school: manager agents dividing a research program into subprograms, subagents pursuing lemmas for hours or days, other agents distilling the resulting abstractions, checking them, and communicating the useful pieces back upward. One reason Grothendieck's IHES school was so successful is that its abstractions were relatively human-compressible. Once you adopted the relative perspective, the ideas could propagate through the community. But maybe that constraint has also been a bottleneck. Maybe many longstanding open problems, like those in number theory which Grothendieck felt was the hardest nut to crack, have solutions that are checkable in principle, but whose motivating abstractions are not human-compressible. In fact, I would wager that many, if not all, of these longstanding, open human conjectures live in PSPACE, but PSPACE is massive! I could imagine the AGrIs of the future might easily find non-human compressible abstractions that can be checked in PSPACE, but are infeasible for any human to check manually. Thus, the next frontier may be mathematics that is machine-discovered, machine-compressible, and machine-checkable — beautiful, in a different way to the machines, but not necessarily in the human way. I can't wait to see what open problems get solved next. What an exciting time to be alive.
English