Engineering failure: many projects are abandoned, not improved

davidmanheim · x · 2026-08-31

Countering the view that every failure leads to better systems, the author notes this is survivor bias. Engineering history is full of projects that were abandoned after trials, not just the ones that succeeded.

Related event: Failed Iterations Don't Guarantee Success in AI Alignment(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →