Very interesting article in New Scientist page 5 issue 3613 about the effects of AI and LLM use on #mathsresearch . The tenet of the article is a that AI tooling will alter a lot of the way progress is made and mathematicians might be damned to sift though AI output. What I am not seeing is examples of things in maths today that are not in the models used, i.e. not in their training. Anyone have examples of mathematical things that the current models are not even aware of? #mathematics #llm #ai

Unus Nemo reshared this.

in reply to adingbatponder 👾

I would guess that #maths departments will have to make their own in-house #AI tools air-gapped and trained by their own people and data and confidentiality contracts galore. This will require a great deal of money. This will need #alumni to fork out. And the race and competition will be fierce. This is of major importance for our universities and maths. Wait. This applies to all research. So university research might just become a matter of infrastructure funding. Unis will have to team up big.

Unus Nemo reshared this.

⇧