Monday, October 12, 2026

Mathpocalypse now.


https://www.nytimes.com/2026/10/08/science/mathematicians-respond-openai-release.html

‘Breathtaking,’ ‘Devastating’: Mathematics Reels After New OpenAI Release
Hundreds of new A.I.-generated findings moved the frontiers of higher math in a single day — dispelling any doubt that the field is forever changed.

Physics, chemistry and the rest of science will be next to be left in the dust. This could potentially change the entire way work is done at LLNL and LANL and change what kind of workforce they will even need.

Rather than slow, human-like incremental logic, computer scientist Scott Aaronson and other researchers noted that the AI generated "alien math" using highly recursive, bizarre code structures.

The good old days are gone

3 comments:

Anonymous said...

You need to filter this nonsense, Scooby.

Anonymous said...

This is coming. How we manage it is the question. I don't see a difference between AI and the supercomputing. Those answers are just as complex and incomprehensible. The key is how they can be understood and translated into something we can understand. This gets to establishing basic principles. Explainability is one reflection of this. AI has deep problems here. It is a huge challenge. The other way is verification (and validation). Math results need to be verified and transformed into something that we can understand. This gets to the famous Hamming quote, computing is for insight not numbers. Extra true with AI. Humans need to think, undertand and create.

Anonymous said...

"You need to filter this nonsense"

In what way is this nonsense? I take you are not involved in any kind of mathematical science. The recent results in AI on mathematics is intense subject in academic circles, national labs and industry. The field of math has had the most upheaval in just the last year. If you know anyone in the field you would know this is huge topic. The NNSA labs have had long history in mathematics so this is certainly relevant. AI is already causing huge issues in academic publications, data analysis, code writing, and review process and so on. In case you have not noticed the DOE and NNSA now have huge involvement in AI, not to mention the economy.

I take that you long since retired from labs but for the rest of us, this is now a daily topic we have to deal with. This blog is dedicated to LLNL and LANL so discussion of how AI is changing the way science is done is certainly not nonsense.

LLNS Contract discussion

SUGGEST NEW TOPICS HERE

Submit candidates for new topics here only. Stay on topic with National Labs' related issues. All submissions are screened first for ...