Falsified prediction: we predict worst about what dominates our frame
jd_pressman · x · 2026-10-16
The author revisits an old falsified prediction about image generation: he wasn't scalepilled enough and overestimated the lifespan of and interest in text-to-image generators, since 4o LLMs have been encroaching on image gen and text control got very good. His meta-lesson: you're most likely to make bad predictions about things that dominate your frame of reference or that you have personal investment in — though that's rarely actionable with the information you have.
Related event: Researcher admits text-to-image prediction proved wrong(2 posts)→
More from AGI Musings
- Kroger cuts 3,000 corporate jobs (37%) for AI; humanoid robots next in the 2030s — csuwildcat · 2026-10-16
- Ben Thompson: generating human desire will be the post-AI economy — generativist · 2026-10-16
- Interpretability won't save us: critics call superintelligence project reckless regardless — zetalyrae · 2026-10-16
- John Nash's 1998 Hierarchical Introspective Logics paper recirculates in AI circles — fkasummer · 2026-10-16
- Ex-OpenAI interp lead: most neurons in trillion-param models are now legible — CatAstro_Piyush · 2026-10-16
- Researcher disavows successionism: watch the button-presser who wants to end the universe — jd_pressman · 2026-10-16