PinnedLLMs DO reason: old logic + new data = new inferencesThe thought that LLMs can’t create new inferences is completely bonkers.Nov 3, 20243Nov 3, 20243
PinnedThe AGI / LLM debate is SILLY: the naysayers insist on FrankensteinLLM-based AGI can deliver everything you can possibly measure pertaining to intelligence.Dec 26, 20242Dec 26, 20242
PinnedThe future began on August 19, 2021But we knew this was coming since the 1950s.Dec 11, 20244Dec 11, 20244
PinnedDo higher modes of intelligence even exist? AGI? Super Intelligence?Like even hypothetically?Dec 4, 202410Dec 4, 202410
PinnedHere’s why people believe doom and gloom science BUT NOT the green tech guysWe can ABSOLUTELY solve climate change in around 15–20 years (barring WW3) and it’ll be cheaper than running the fossil fuel plants.Apr 9, 202324Apr 9, 202324
GPT-4o: In a league of Its own for *accurate* long-context performanceWe should all be quoting – and checking up on ‘NoLiMa’ metric scores.Apr 10Apr 10
Penrose’s consciousness-by-non-computation is interesting, but has 3 major shortcomingsIMO.Apr 106Apr 106
We should all have shifting political allegiancesRight NOW, for me? 6/10 Trump / Elon. 4/10 Dems. But it changes every month! How about you?Apr 95Apr 95
Why consciousness as ‘an illusion’ contains a hidden assumption (+ is devoid of mechanism)It’s simple but multi-angled.Apr 98Apr 98
AGI & LLMs: If you don’t understand these 3 facts, you’ll miss the entire thing.Each is simple. Together it’s revealing.Apr 53Apr 53
