💯My sneaking suspicion is that a RenTech type organization is the lagging indicator of real AGI (I.e. consistently outperforming the market, not artificial benchmarks). Any lab that accomplishes AGI will likely be similarly secretive and closed to outside capital. Any

Oh god are we really doing this? Jeff Dean trained an n-gram model on the entire internet in 2007. Jelinek coined the term "language model" in the '70s. It's called "Claude" because Claude Shannon was estimating the entropy rate of the English language in 1951!

New blog post: the state of AI safety in four fake graphs.

Novel environments, no precedents or plagiarism possible. Humans 100%, AI <1%.

This is a bit misleading The only way to know if humans with no ARC-AGI-3 training can solve it, is to ask uncontacted tribes to solve it

Indeed, it was *because* I was not from the aerospace industry that SpaceX made such radical breakthroughs. Same for Tesla. Those in the industry would have if they could have.

Examining the quality of the methodology & participants--& depth of the questions-- this study is the most ambitious subjective-probability forecasting exercise I've ever seen. Agree or disagree with the conclusions, this is a state-of-the-art benchmark to beat

Old books are better because they sprung from a culture where reading, not watching, was the predominant mode of leisure. That not only meant authors were well-read and thus better writers, but that their readership was well-read too, so authors could expect more of them

“Management questionable…” - Don Valentine about Jobs presumably. Apple was a lot more nonconsensus at its origins than it appears in hindsight.

Must read on the history of manufacturing quality control/ supply chain strategy, Demming’s predecessor (Sarasohn), Aki Morita, and much much more…

“I want you to make a list of every app any customer would ever want to use. And then the two of us will prioritize that list. And then I’m going to write you a blank check, and you are going to build the largest development team in the history of the world, to build as many apps

It took humanity ~ 50 years to go from time-sharing compute to….time-sharing compute.

Wonderful. Is it outdated of me to be nostalgic for when kids wanted to grow up to be astronauts instead of influencers?

I’ve long held benahorowitz.eth explained it best: “When a company goes astray, you talk to employees and they say, ‘We have no strategy. We don’t know where we’re going.’ The strategy is the story. They’re not different. The strategy is the story you tell. It’s the why. If you can’t

Apparently Johnny Kim is not an outlier, astronauts as a group really are awe inspiring.

“You can see the computer age everywhere but in the productivity statistics.” - Robert Solow (1987)

A camel is a horse designed by a committee. That’s why you often need an outsider to tell you that what you have is a camel, not a horse.

The future of social science is multiverse analysis, preferably with automated AI verification systems.

Dear Tim Cook - make the Apple neural engine low level APIs fully accessible to developers. Unleash the local and decentralized AI developer ecosystems on your amazing silicon. Return to the days of the Apple II and help rewrite computing history. Combined with sitting out the