Delip Rao: in 2018, claiming you foresaw today's AI was sci-fi imagination

deliprao · x · 2026-09-07

Responding to Miles Brundage, Delip Rao lays out his full case: transformers arrived in 2017, decoder LLMs in 2018 didn't automatically scale, instruction alignment came in 2020, and function calling plus ReAct only in 2022—only after that last invention could anyone reliably extrapolate to today's AI world.

So anyone making such claims in 2018 was either making them up or indulging in sci-fi imagination, and retroactive vindication is a dishonest reading of history. The most complete version of his side of the debate.

Related event: Brundage and Delip Rao Clash Over AI Misuse Predictions and the Merits of Scaling(5 posts)→

Original post →

More from AGI Musings

AGI Musings channel →