https://borretti.me/article/the-other-road interesting perspective on AI -- the transhuman goal of "become as gods" contrasted with the "build an AGI god" goal. I had always thought of AI as part of the project of understanding what a human mind is, and potentially enhancing it -- but that depends entirely on how we develop AI and what we sue it for, and what we seem to be doing is more like building very non-human-like architectures for automating cognitive labor, rather than turning a computational lens on ourselves. If AI "works" but we never get around to understanding it, and end up in a sort of passive WALL-E future, that's not really the transhumanist dream.
I'm not at all sure that scaling LLMs is more risky than other architectural paradigms -- it might be less so, actually.
But definitely, there are a lot of things that "AI" could be. From the point of view of wanting to achieve utopian outcomes, there's no good reason to scale as fast as possible.
"Many frontier lab employees, especially at Anthropic, are out of touch and this will impact their forecasting and/or descriptions of current AI events. I say this without blaming individuals, but it’s a common agreement among my friends not at OpenAI/Anthropic (Ant especially) that people at the labs operate with a religious energy. It’s very common to go through very out of touch interactions with them. I do not blame most of the individuals who get distorted views being part of these companies, but the interactions are wild and spill over into a lot of wack discussions in the AI media ecosystem. Living in this environment that normalizes such out of touch behavior will inevitably distort any human’s understanding of technical progress."
At this point, I view the phrase "out of touch" as a red flag. It simply means saying something that people disapprove of, without apparently being aware that you come across badly. But is it actually good to have your head on a swivel, hyper-aware of everyone who might disapprove? Being "out of touch" is what happens when you focus on your own work and your own peers, which is usually a good thing.
this is not saying that the AI lab insider perspective is necessarily correct. they may very well have groupthink & bias.
but this guy is literally arguing from "other people don't like you, therefore you're bad".