GPT-5.6 used a prompt to close a 30-year gap in convex optimization

old.reddit.com

July 18, 2026

1 min read

🔥🔥🔥🔥🔥

69/100

Summary

GPT-5.6 utilized a prompt to bridge a 30-year gap in convex optimization research. This advancement builds on OpenAI's recent proof announcement related to convex optimization.

Read original article

View story →

Coverage & discussions

Hacker News · old.reddit.com

GPT-5.6 used a prompt to close a 30-year gap in convex optimization ↗

Community discussions

Hacker News discussion ↗

Historical count: 5 · 5 saved in sample · Historical sample

Analysis ready

Community Sentiment

Mixed

Positives

Crazy how intelligence is cheap, efficient and commonplace now — this could redefine our roles and values in society.
This apparently required a 10-page prompt, showing that effective prompts can unlock substantial capabilities in AI.
The optimization of convex bounded Lipschitz functions is exactly what most modern AI models are based on, highlighting a significant intersection between math and AI.

Concerns

Some commenters are deeply pessimistic about the future, suggesting banning LLMs to preserve human relevance, which feels dystopian and counterproductive.
There's a concern that LLMs lack spatial reasoning, which is fundamental for technology and industry, indicating a significant gap in current AI capabilities.
The infrastructure and training required to achieve these results are massive, challenging the notion that AI is truly 'cheap' and 'efficient'.