Former OpenAI Founder Builds Thought-to-Text via Bitter Lesson Scaling

alex_verem · x · 2026-08-06

Former OpenAI founder Naomi Bashkansky is currently developing thought-to-text technology to achieve telepathy. The core approach relies on the AI sector's 'bitter lesson': throwing massive amounts of compute and scaling up data collection so models can predict text directly from brain signals, rather than relying on hand-crafted algorithms.

Original post →

More from Research

Research channel →