Researcher grades his own 16 AI predictions nine months later: what held up and what didn't

avt_im · x · 2026-09-13

Nine months after publishing sixteen AI predictions, Alexander Terenin reviews each one in a detailed blog post, self-grading them as correct, partially correct, wrong, or too early to tell. His headline bets: AI would transform software engineering almost overnight, shifting the job toward communication and managing coding agents rather than writing code. The post also analyzes whether his misses came from over- or under-estimating progress.

Related event: Researcher Reviews 16 AI Predictions Nine Months Later(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →