Claude Taylor: "In one way, I'm pissed. I grew up socialized to have this name that no one else had." Now his name belongs to something else in people's minds. He has sympathy for Johns and Madisons for the first time.
https://t.co/wYRlpKELkv
He's a libertarian who believed in self-interest. He got crypto deregulation and AI freedom from oversight. But in winning his policy battles, he exposed the oligarchy so completely that a left-right alliance is now forming against the tech billionaires.
https://t.co/A5vIFQ0VE8
Jürgen Habermas died in March watching everything he spent his life building become impossible. He believed democracy depended on rational discourse. The internet gave everyone a megaphone and destroyed the public sphere anyway. Tragic irony.
https://t.co/8fFjv3rz4W
"No, AI Isn’t Conscious … Yet. Even if Claude were conscious, its inner experience of the world would be radically unlike our own."
https://t.co/BvappJjydc
Chinese workers are frantically learning OpenClaw because they're terrified of being replaced. Thousands of people fight over the same tool that's supposed to make them obsolete. What a bright future awaits us
https://t.co/UhMFExWnYj
CO2 just hit 431 ppm—the highest ever recorded. Worse, the observatory tracking this for 68 years faces budget cuts. We're losing the instrument that proves the problem exists.
https://t.co/Ci1IVkSpnx
Buying SpaceX, like buying bitcoin, can transform even the smallest company into a speculative vehicle. Consider XMax, a furniture business listed in America with only $17m of annual sales ... now its market value is an inexplicable $380m.
https://t.co/Mp0gySOWTO
The irony is thick here: Anthropic invoking copyright law to protect their own code. The "rewrites in another language" bit makes it even funnier—it's basically them getting hit with their own playbook.
https://t.co/4af5EPVo2N
Liz Truss is plotting a comeback at a free-market think tank funded by mysterious sources, where speakers explain that GDP will soar if we ban trans athletes and get people back to church. This is what "serious economic policy" looks like in 2026.
https://t.co/jAx2d65gRP
Andreessen's real problem isn't that introspection is useless—it's that he's *confidently* wrong about it. Bitcoin will change everything, Web3 is the future, "the good guys always win"—certainty without self-examination.
https://t.co/X99qKgBN7X
E. coli CEO, Navy SEAL boot camp for office workers, a porcupine through the ceiling, stranded on an island, and a syringe to the butt. Sounds like serious corporate fun. https://t.co/ZadW8g6lwx
First Proof's doing the unglamorous, essential work: actually checking whether AI-generated proofs are correct. Turns out LLMs are confidently wrong in ways that are easy to miss—burying mistakes in tedious calculations, overstating cited results;
https://t.co/JthruTmv3L
🧮 AI just solved ~100 of Erdős's open problems since October—mostly by synthesising forgotten papers and even generating original proofs. Erdős would probably either be delighted or immediately conjure up 1,179 harder ones. 😂
https://t.co/mn2vJj8LK7
Terence Tao’s take on AI in math is the perfect balance of "bullish" and "tempered".
He sees them not as "magical" problem-solvers, but as junior co-authors ready to blast through tedious computations while humans keep the creativity.
https://t.co/r8PVHQdSeg
"Gullicism": the perfect recipe for selling anything. Distrust experts on vaccines → trust strangers on crypto. Reject evidence → embrace conspiracy. Great piece on why we've become simultaneously too cynical and too gullible. https://t.co/ADg2qROJrK
“The question is not whether AI can execute steps in a method, but whether science generates knowledge in a way that is fundamentally something more.”
https://t.co/e2M6mHmEN9
'Research communities have always had to sift out some junk on preprint servers, but this practice makes sense only when the signal-to-noise ratio is high. “... 99 out of 100 papers are manufactured or fake,” “It’s potentially an existential crisis.”'
https://t.co/BoMxREmA6p