Post by Amrith Krishna

AGI Timelines Are All Over the Map What the disagreement says about business incentives, technical hurdles, and the meaning of “general intelligence.” - 2 years - 5 to 20 years - Decades Those are real predictions from some of the most influential voices in AI. Sam Altman talks about a couple of years. - Yann LeCun says decades. - Geoffrey Hinton now says 5–20 years. - Shane Legg gives even odds for 2028. The spread is staggering. These are people working at the very frontier of the field, yet their mental models for AGI’s arrival barely overlap. In my latest piece, I group them into four camps: - The Imminentists: Scaling Optimists who see AGI as a continuation of current trends. - The Threshold Watchers: Probabilistic pragmatists who speak in likelihoods, not dates. - The Revised Optimists: Experience-driven realists whose timelines have shifted after recent surprises. - The Foundationalists: Architectural skeptics who believe key capabilities are still missing. Each camp’s forecast reflects a different combination of professional incentives, technical focus, and personal experience. The disagreement is a window into the standards war over what "general intelligence" even means. Which clock would you set? #AGI #GPT5 #GPT #AI #AIML #GenAI #GenerativeAI #LLM #LLMs