This isn’t about super-powered surveillance (well, it is that too), but about the fact that AI can now watch the world, remember hours of what sees at a time, and “reason” about what it sees in detail, with few mistakes. Using visual, not just text, data greatly expands AI power

👍

Daniel Dennett has died. 😢 dailynous.com/2024/04/19/dan…

Virtually nobody is pricing in what's coming in AI. I wrote an essay series on the AGI strategic picture: from the trendlines in deep learning and counting the OOMs, to the international situation and The Project. SITUATIONAL AWARENESS: The Decade Ahead

Good thread.

Looks like we got self-driving cars after all.

For those who want to follow more people in AI Policy/Governance work I have a well-curated Twitter list here: x.com/i/lists/143945…

For those who want to follow more people working on the cutting-edge of AI tech I have a well-curated Twitter list here: x.com/i/lists/173139…

Meanwhile in France: in office buildings, it is illegal to switch on the air conditioning if the interior temperature is less than 26 °C or 78.8 °F. (Décret n° 2007-363)

The raw chain of thought from DeepSeek is fascinating, really reads like a human thinking out loud. Charming and strange.

The timelines they are a-changin'

Today I was unsure about whether to do something so I asked tylercowen for advice, and he started it off with: “Keep in mind that most of what we do is pointless, in a way.”

"How, exactly, could AI take over by 2027?" Introducing AI 2027: a deeply-researched scenario forecast I wrote alongside Scott Alexander, Eli Lifland, and Thomas Larsen

Agreed