Threads

No threads match these filters.

1. Cognitive tests at scale, for pennies

   Published October 25, 2026 · research Judgment modelsCognitive scienceAI & machine learning active

   Jev, from TypeSafe AI, answers a typed question about a text with a probability in about a fifth of a second for about two thousandths of a cent, and when it is confident it gives the same answer each time. That makes psychology-style experiments practical at a scale that normally needs hundreds of volunteers: rating thousands of words, or running a battery of classic bias tests in minutes. What turns up matters to AI builders, Jev users and brain researchers: some human quirks appear, some do not, and on first impressions it does the opposite of people.

   Results →Results →Guide →

2. Growing up with AI

   Published October 22, 2026 · essay Education & learningChildren & teensAI companionsAI & society active

   An essay of advice for kids, teens and future generations growing up with powerful AI.

3. Frontier labs as national laboratories

   Published October 22, 2026 · essay Frontier AI labsNational laboratoriesAI policyAI safety active

   An essay on whether national security, secure methods and international competition will push the frontier AI labs toward national-laboratory status, and what past labs suggest about the shape that could take.

   Tracker →Newsletter →

4. AI-native biology, beyond the announcements

   Published October 22, 2026 · research AI-native biotechDrug discoveryProtein structure & functionAI & machine learning active

   A research series on the companies rebuilding biological discovery around AI: what they are betting on, how their approaches differ, and what changes as the work meets reality.

   Table →Newsletter →Table →

5. Scale Tool

   Published October 9, 2026 · tool Education & learning active

   A React and TypeScript explorer in which one editable decimal value drives source-backed comparisons across nine distinct units. It keeps estimates, models, theory, and unknown quantities explicitly separate from measurements.

6. Jev Turing

   Published October 8, 2026 · experiment Judgment modelsCognitive scienceAI & machine learning active

   An independent experiment that uses a Jev Boolean judgment as the only logic primitive for gates, an SR latch, and Rule 110 cell updates, with recorded runs and a ground-truth checker. It is not affiliated with TypeSafe AI or Vercel.

7. AI Agent Journey Journal

   Published August 19, 2026 · journal AI & machine learningAI & society active

   A static, evidence-backed journal about changes from conversational AI assistants toward persistent local agent systems. It records directional changes in architecture, operation, governance, and understanding rather than routine updates.

8. Passwords in the Age of AI

   Published August 17, 2026 · guide Account security & scamsAI & society active

   A four-page, printable guide for friends and family on keeping accounts safe now that AI makes familiar scams easier: the five ways accounts get opened, password managers and passkeys, a security-key setup, recovery, and a worksheet for the accounts everything else depends on.

   Guide →