I'm building an applied AI lab and looking for a cofounder who is: - San Francisco-based - Deeply technical and empirical - Comfortable with non-consensus bets $5T market + already have real customers with high usage + first checks committed. (1/3)

More info about who I'm looking for: https://t.co/Io4JJV0laq Reposts are appreciated! ❤️

Current focuses: - Post-training for long-horizon teaching tasks - Scaling human-like temporal memory - Rearchitecting voice agents to feel present More info about current progress: https://t.co/zm9VRC5Pdc

These folks seriously cooked. If you want a deep dive on 12 factor agents in the wild, take a long look - roll your own event management - prompts are just functions that output JSON - if you roll your own memory you get complete control and can optimize incrementally

just published a detailed technical post about how we've been building *proactive* AI agents at Orin. https://t.co/07LDdwLoV8

Why AI based tutors are going to be such a big deal 1:1 tutoring = 2 sigma improvement in learning achievement Image from "The 2 Sigma Problem: The Search for Methods of Group Instruction as Effective as One-toOne Tutoring" by Benjamin S. Bloom

And messaging? Literally every site visitor is going to read your header and CTA - you should know which messaging works best and WHY. Even if you're not trying to drive inbound pipeline yet, just do this to learn about what resonates with your buyers. It's underrated.

But you're also wasting opportunity if you're doing nothing. Everybody that you talk to while pre-PMF will have checked out your site. It's a credibility check. And they're doing stuff on your site that give away what they care about - that's data you should have.

it's hard to shake the feeling that history is being made within four blocks of haight and steiner.