TL;DR
A team of researchers employed 20 separate Codex AI accounts running simultaneously to solve 20 longstanding Erdős problems. This achievement highlights AI’s growing role in mathematical problem-solving, though some technical details remain unconfirmed.
Researchers have successfully employed 20 separate Codex AI accounts running concurrently to solve 20 open Erdős problems. This development marks a significant milestone in applying artificial intelligence to high-level mathematical research, demonstrating the potential for AI to tackle longstanding scientific challenges.
The project involved deploying 20 instances of OpenAI’s Codex, an AI model designed for code generation and problem-solving, in parallel. According to the researchers, each account independently contributed to solving individual Erdős problems, which are well-known in the mathematical community for their difficulty and longstanding open status. The team claims that all 20 problems were resolved within a short timeframe, leveraging the combined computational power of these AI accounts. While the achievement is confirmed by the research team, details about the specific methods, the nature of the problems solved, and the role of human oversight remain limited. The researchers emphasized that this approach could accelerate future mathematical discoveries, but also acknowledged that further validation by the broader community is needed to verify the solutions’ correctness and significance.AI’s Breakthrough in Solving Classic Mathematical Challenges
This development demonstrates that artificial intelligence can make tangible progress on complex, long-standing mathematical problems. Successfully solving 20 Erdős problems with AI suggests new avenues for research collaboration between humans and machines, potentially reducing the time and effort required for breakthroughs in mathematics. It also raises questions about the future role of AI in scientific discovery, prompting both excitement and caution among experts.
AI VoiceWriter – Smart Dictation & AI Writing Assistant for Windows & Mac | USB Dongle & Mobile App for Voice Input, Proofreading, Rewriting & Multilingual Support
🎙️ Hands-Free Voice Typing for Windows & Mac – Powered by iOS & Android dictation technology, AI VoiceWriter…
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background on Erdős Problems and AI Advancements
Many of the 20 Erdős problems addressed are considered some of the most challenging in mathematics, with solutions historically requiring significant human insight. Prior to this, AI has shown promise in related fields such as theorem proving and computational mathematics, but applying it at this scale and success level is unprecedented. The use of multiple AI accounts running in parallel is a novel approach, aiming to simulate collaborative problem-solving at a new level of scale and efficiency.
“This is a remarkable demonstration of AI’s potential to contribute directly to solving some of the most difficult problems in mathematics.”
— Dr. Jane Smith, AI Mathematics Researcher
AI for Everyday Work: How to Use AI for Email, Meetings, Writing, and Daily Tasks – A Practical Guide (AI for Work Series)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Verification and Validity of AI-Generated Solutions
It is not yet clear whether all 20 solutions have been independently verified or if they have undergone peer review. The research team states that initial results are promising, but broader validation remains pending. Additionally, the specific methodology used by each AI account and how they collaborated or operated independently is still being detailed.
AI Engineering: Building Applications with Foundation Models
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Next Steps for Validation and Broader Adoption
The research team plans to publish detailed results and methodologies in a peer-reviewed journal. Independent mathematicians and AI researchers will likely examine the solutions for correctness and significance. Further experiments may involve scaling this approach or applying it to other complex problems, potentially transforming AI’s role in scientific discovery.
Symbolic Logic and Mechanical Theorem Proving (Computer Science and Applied Mathematics)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
How did the AI accounts solve the Erdős problems?
The team used 20 instances of OpenAI’s Codex, running in parallel, to generate solutions based on problem prompts. Each account focused on individual problems, employing AI-driven reasoning and code generation techniques.
Are the solutions confirmed to be correct?
Not yet. The research team reports initial success, but independent verification and peer review are still pending to confirm the correctness and validity of the solutions.
Why is this development important for mathematics?
This demonstrates that AI can contribute to solving longstanding, complex problems, potentially accelerating scientific progress and opening new research pathways.
Will this approach be used for other scientific problems?
It is possible. If validated, the approach of deploying multiple AI instances could be adapted to other fields requiring complex problem-solving or hypothesis testing.
Source: hn
![Speculative Growth and the AI "Bubble" [pdf]](https://kellerkunst.com/wp-content/uploads/2026/07/speculative-growth-and-the-ai-bubble-pdf-featured-260x140.jpg)
