Users Shun Reasoning Models

Updated: 2025.08.28 1M ago 2 sources
Sam Altman says only 7% of ChatGPT Plus subscribers used the new o1/o3/o4 reasoning models. Despite benchmark gains, most users favor lower‑latency, cheaper defaults over chain‑of‑thought features. — Adoption lag reshapes safety, monetization, and regulation because frontier capabilities may remain niche unless integrated into fast, default experiences.

Sources

Mass Intelligence
Ethan Mollick 2025.08.28 82% relevant
The article cites OpenAI’s finding that only ~7% of Plus users selected o3 (a reasoning model) before GPT‑5 and reports that, after GPT‑5’s auto‑router launch, Reasoner use rose to 24% among payers and ~7% among free users—directly extending the adoption gap noted in the existing idea.
Links for 2025-08-11
Alexander Kruel 2025.08.11 100% relevant
“Only 7% of ChatGPT Plus subscription users were using the o1/3/4 reasoning models” — Altman on X, linked in the roundup.
← Back to All Ideas