Three Years of AI Progress: 50-Msg GPT-4 Cap to 1M Context, 39% to 96% GPQA
alliekmiller · x · 2026-10-18
Allie Miller juxtaposes an old ChatGPT Plus screenshot (50 messages per 4 hours) with today's frontier capabilities: GPT-4's 8K/32K context vs Astra's 1M, and GPQA Diamond scores jumping from 39% to 96% — a reminder of how far AI has come in three years and seven months.
More from Models
- Model repeatedly claims to be ChatGPT once OpenRouter default system prompt removed — scaling01 · 2026-10-18
- New model repeatedly identifies itself as ChatGPT in hands-on test — scaling01 · 2026-10-18
- GLM 5.3 finds arbitrary-write vulnerability in Apollo 11 guidance code — moyix · 2026-10-18
- Google AI flags its own generated character as a celebrity mid-project — noisywan · 2026-10-18
- Grok's "complex age dynamics" phrasing for historical abuse draws flak — eigenrobot · 2026-10-18
- Dev revisits Opus 4.6 project with Opus 5.5, finds it spotting old bugs unprompted — rms80 · 2026-10-18