Tweets

2024-03-06T03:01:39+00:00 | 🔗

@eshear Should have done it as CEO :(

2024-03-06T02:43:19+00:00 | 🔗

. @ilyasut purposefully dug through his archives to find Elon's acknowledgement and cosigned the recent OpenAI letter. https://t.co/1PR8Yh5nFr https://t.co/qZfBJmCriz

2024-03-06T02:42:34+00:00 | 🔗

@wagieeacc @karpathy It's also fascinating that Ilya is a signer.

2024-03-06T00:50:17+00:00 | 🔗

@BenBlaiszik @MaximZiatdinov Definitely! I’m very excited about being able to get the latest expert consensus quickly and getting caught up/to the front fast. To synthesize and cross-reference. If we are able to interpolate across all scientific lit. would be no mean feat.

2024-03-05T22:50:51+00:00 | 🔗

Almost this. We are at AGI, it just turns out median is never good enough :). https://t.co/d9lNlmY0XP

2024-03-05T19:57:27+00:00 | 🔗

@BenBlaiszik @MaximZiatdinov That’s the key, “do”ing, I think

2024-03-05T01:38:17+00:00 | 🔗

Wow yeah. If that isn’t enough they should offer Dean of Harvard Law School. Because Dean of HLS is clearly a more desirable position the Associate Justice of the Supreme Court, and we definitely surely certainly have no evidence otherwise. Who would say no to Dean of HLS? https://t.co/TtM3J9Vync

2024-03-04T21:30:40+00:00 | 🔗

@mckaywrigley You misspelled “here” as “near”.

2024-03-03T21:50:21+00:00 | 🔗

@TimothyDuignan I think so, when you have enough data for ML models to apply humans would have recognized the pattern already. Where I think ML is useful is in surfacing and systematizing data, and distributing insights in some way (very broad category for what this includes)

2024-03-03T20:59:47+00:00 | 🔗

RT @pmarca: Let's assume, for discussion, that AI in 2024 is like atomic technology in 1943, that AI should therefore be handled like the M…