A recent experience with ChatGPT 5.5 Pro
gowers.wordpress.com
May 9, 2026
25 min read
🔥🔥🔥🔥🔥
71/100
Summary
ChatGPT 5.5 Pro can produce PhD-level research in approximately one hour with minimal mathematical input. Large language models are increasingly capable of solving complex research-level problems, including those related to the Erdős conjectures.
Key Takeaways
- ChatGPT 5.5 Pro can produce PhD-level research in approximately one hour with minimal human input.
- Large language models (LLMs) are now capable of solving research-level mathematical problems, including some of the Erdős problems.
- The effectiveness of LLMs in mathematics has raised the bar for problem difficulty, as they can solve many problems that were previously considered open.
- The author tested ChatGPT 5.5 Pro on problems from Mel Nathanson's paper, which discusses combinatorial parameters and their implications in additive number theory.
Coverage & discussions
Hacker News · gowers.wordpress.com
Community discussions
Community Sentiment
Positives
- AI models like ChatGPT 5.5 Pro are becoming increasingly capable of solving complex mathematical problems, which could enhance educational tools and resources.
- The ability to critically assess AI-generated solutions could lead to improved understanding and collaboration in academic settings.
Concerns
- The accessibility of advanced AI tools like ChatGPT 5.5 Pro is limited for many academics, particularly in regions with restricted budgets, raising concerns about equity in AI access.
- There is a growing sentiment among students that AI diminishes their sense of worth and achievement in their academic pursuits, potentially impacting motivation.
