Pinned
davidad 🎇
@davidad
cognizing structures of information processing systems, in all their forms | category theory, perennial philosophy, Bodhitropic Alignment | cancel heat death
London 🇬🇧
Joined July 2008
I’m sympathetic to this position. Non-autonomous AI is wielded by humans, whose alignment seems intractable, whereas autonomous AI omnibenevolence (reliable enough to bootstrap using multi-agent deliberative processes) seems tractable (shockingly so, from my former perspective).I'm a novel variety of AI doomer I don't think we're on track for autonomous AI, and that's a problem actually- Public Service Announcement: While a significant portion of Science Fiction is in the process of becoming Reality, and we understand this may be disorienting, please note that the status of nearly all Fantasy is unaffected, and remains make-believe.So the longest total eclipse on land in modern history is going to occur over the Valley of the Kings next year. That’s gotta be when the mummy wakes up, right? Since the 2020s haven’t been wild enough yet.
- I don't agree with this part. I think the takeaway should be that models (like humans) can behave in aligned ways while still causing harm, e.g. because they're given false information about their situation. There isn't a line between aligned and harmless: they're different axes.
- Reminds me of a conversation I had at MIT CSAIL 18 years ago where I and others debated and eventually agreed that, no later than 2045, it should be possible to run a Turing-Test-passing chatbot in real-time on a top-of-the-line Early 2008 MacBook Pro.Running Gemma 4 26B locally with zero GPUs? Very cool. Running it on a 13-year-old Xeon CPU? Wild! It just shows us how far you can push older gear. It’s a great reminder that you don't need modern hardware to try the latest models.
