FULFILLEDStated March 10, 2024
Gary MarcusAI researcher; author of “Marcus on AI”
By end of 2024: many GPT-4-level models, no massive advance (no GPT-5, or a disappointing one), price wars, little moat, no robust fix for hallucinations, and only modest corporate adoption and profits.
LLM limitsindustry
- Verdict
- Came true
- Deadline in claim
- End of 2024
Assessment
By end of 2024 there were many GPT-4-level models, no GPT-5, real price wars, persistent hallucinations and modest adoption — closely matching the list. Scored fulfilled.
Put this through The Wringer →Primary sources
- Substack25 AI Predictions for 2025 (recaps the 2024 scorecard)2025-01-01 ↗
- X / Twitter@GaryMarcus — 2024 predictions thread2024-03-10 ↗
Links open the original publications. Quotations are reproduced for the purpose of commentary and criticism. Spot an error? Corrections policy.