someone asked me yesterday: who is the one figure that would be able to start an AI eval organization and be globally seen as neutral and trusted both technically and ideologically and financially. the only name that popped into my head was @ID_AA_Carmack (not that he has any interest in doing so!).

on the idea of evaluators: think it's important that we have a distributed ecosystem of indepedent evaluators. the more eyes and people with distributed skill sets the better. it would be a good idea to fund several efforts on this.

I find myself disagreeing with Terence Tao (a very scary phrase to say!) and think the core question posed is whether the "misalignment" is between the mathematical community and humanity. In other words - are the Millenium Prize problems meant to incentivize advancements for humanity or foster the field of mathematics and mathematicians? As a counter example, if there was a prize for creating a drug that cured a rare strain of cancer, we would not care if it was AI that did it. We would be happy it has been solved. On the other hand, if long running agents figured out the puzzle in "Kryptos", the cryptographic statue in Langley, there would be something a tiny bit sad about the human artistry removed.

there's a lot of product magic in agents like muse or instinct in making auth into websites work. however, there still seems to be some ways to go given the various login systems you run into (passkeys, OTP via text or email, 2fa, "auth on other device").

More on @hilbertspaess - by all accounts he seems sincere since we have many here who vouch for him. I don’t agree however that him or just a subset of the people who work on the pipeline of tokens we use everyday speak for everyone else in the industry.

People often think these posts come from researchers who don’t truly believe them. I disagree, that’s not been my experience with Ant or OAI or any other researcher at or near the frontier. I find them to be intellectually honest in terms of believing what they say. However I don’t think any one company or community or discipline or school of thought should represent an entire industry that is so obviously critical. We need new and diverse sources of thought and epistemology here including in the terms we use. as an example: what do the creators of the open models think ( now many here and abroad)? What about the cyber ecosystem ( since security is a priority )? What about chip manufacturers or fabs? Much of the frustration of this discourse is how centralized it is.

Been really loving from @Meta and when I got early access it actually wound up helping me buy a bunch of stuff I would have struggled with otherwise. it has many thoughtful architectural choices around remote browsing ( really interesting UX modality ) and security that I think push the state of the art forward.

seeing the Astra + Blender work makes me think open source tools will have a strong advantage in the AGI era given how much A) online content exists around them and the models learn from that. B) they tend to be more programmer friendly / hackable than closed tools.

I tried to teach myself 3d modeling with Blender and Maya many years ago and quickly gave up with the amount of effort involved. Stunning to see the GPT-Astra created Blender scenes purely from images.

talking to a few ai safety / alignment organizations and find broadly they are not sure what to make of the data center protests - also it’s not something that has shown up in any timelines or predictions. I see a split between A) “we probably need data centers in US than in China” B) “less data centers is a proxy for a slow down in capabilities and that could be a good thing “

one real issue I have with various agents now is "where is this code going to run" - does it need access to my local filesystem - can I fire it off from my phone and if so does it run on the cloud or needs my Mac mini fired up / synced? The current UX metaphors around this seem to need work.

everyone should go read @dwarkesh_sp’s post - it does a great job of laying out the timeline and what we know ( and don’t know) I do have two issues with it A) the use of anthropomorphic language. These are not civilizations nor do they have desires just like a CPU thread or a bunch of programs don’t. this doesn’t mean we downplay the importance of this moment for cyber - but using human parallels for what I believe is code is dangerous territory. B) IMO it gets the impact of open source models wrong and is unreasonably dismissive for reasons that are not clear. As @ClementDelangue highlights below HF was blocked from using closed models to analyze what was happening and had to turn to open weight models to help them make sense of it. This is a very key moment for how we think about intelligence and cyber and every bit of extra understanding and clarity helps.

This @axios piece on fracking and the national discourse around it as a parallel to data centers was fascinating. The current consensus answer seems to be : address local concerns while making the case for economic benefits. Have been sending a lot of people to work by @jasminewsun @AndyMasley and others on datacenters this last few days.