Yeah, it's tricky, but I'm still doing stuff. Just not publishing here at the moment. Hoping to get back into it soon but we'll see. How's life with the little one?
The 26GW worth of capacity announcements from OpenAI alone within just weeks really underscores how the bottleneck has shifted from compute availability to power infrastructure. What's particularly compelling about the Broadcom partnership is that OpenAI is designing their own chips, which signals a longer term strategy beyond just purchasing GPUs from Nvidia or AMD. The fact that AI is already making novel hypotheses about cancer cellular behavior and independently rediscovering black hole physics shows we're past the point of just pattern matching. The exponential curve you highlighted isn't just about more compute, it's about compute being used to discover fundamentals that unlock even faster progress in science and enginering.
Indeed. Sam Altman speculated that perhaps at 10GW of compute an AI could cure cancer (but maybe it's 100GW).
We're at the point I think where one can argue that these systems do in some ways exhibit genuine intelligence, and aren't just super sophisticated patern recognition programs. Interestingly, I think it also shows that unlike with biology, intelligence and sentience appear to be separate phenomenon. Or at least that they can be.
Power availability definitely will be the bottleneck. Myself and a colleague are trying to get multiple 100+GW energy hubs built, based on nuclear (fission & fusion). Bit of a long shot, but we're making progress smoothing the path politically in a couple places, and starting to reach out to companies.
Great roundup! Not too long and very digestible which is great in such a fast moving, complex field.
Thanks Shoni! How's things going on the writing front for you? I always find the limiting factor to be time.
Yeah, it's tricky, but I'm still doing stuff. Just not publishing here at the moment. Hoping to get back into it soon but we'll see. How's life with the little one?
Good! Busy. Very busy lol. But good. Very much worth all the business and trouble.
The 26GW worth of capacity announcements from OpenAI alone within just weeks really underscores how the bottleneck has shifted from compute availability to power infrastructure. What's particularly compelling about the Broadcom partnership is that OpenAI is designing their own chips, which signals a longer term strategy beyond just purchasing GPUs from Nvidia or AMD. The fact that AI is already making novel hypotheses about cancer cellular behavior and independently rediscovering black hole physics shows we're past the point of just pattern matching. The exponential curve you highlighted isn't just about more compute, it's about compute being used to discover fundamentals that unlock even faster progress in science and enginering.
Indeed. Sam Altman speculated that perhaps at 10GW of compute an AI could cure cancer (but maybe it's 100GW).
We're at the point I think where one can argue that these systems do in some ways exhibit genuine intelligence, and aren't just super sophisticated patern recognition programs. Interestingly, I think it also shows that unlike with biology, intelligence and sentience appear to be separate phenomenon. Or at least that they can be.
Power availability definitely will be the bottleneck. Myself and a colleague are trying to get multiple 100+GW energy hubs built, based on nuclear (fission & fusion). Bit of a long shot, but we're making progress smoothing the path politically in a couple places, and starting to reach out to companies.
Energy distribution priorities are already changing fast. 5 years from now it'll be a totally different world.