Source item

fun to do a last minute “state of ai startups” talk (w/ @pranavreddy) this am at #neurips for old friend @swyx at his live paper-reading club couple quick takeaways below on what’s working, where we disagree with the dominant narratives, and some of our latent priors in investing👇 https://t.co/Fe8KbMBLgU AI years are like dog years: what’s changed https://t.co/vkDofoOJRZ Model market is increasingly competitive https://t.co/o47K4uAj9s Open source models are good, less than a year behind, here to stay https://t.co/gO3etcLr6p Intelligence is cheap, but not too cheap to meter https://t.co/CsPW5COKcg Test time is new dimension for scaling (in some domains) that better aligns compute spend to user value than pre-training upfront capex https://t.co/Ezi2lWluZp Doesn’t look quite like a bubble vs 2021, especially when you take out the foundation model rounds (red) https://t.co/CyQeAZTCxz Some companies are working! Users voting with their feet and dollars https://t.co/uAKKwAulEh AI’s “Catch-22” is clearly false https://t.co/kzHmi4iiAS What end markets are attractive for startups is changing https://t.co/pxuUtapAVA Supposed advantages of incumbents to capture the value of AI are weaker than they appear https://t.co/nHu2UNsgeg Post-2024, the market is much more structurally supportive of startups Change favors the small, fast and creative /end https://t.co/QDiLHk0lGl