Interesting article on progress in automated reasoning

[Josef Urban]

Briefly. I have become disappointed and skeptical of Quanta. They seem to
be quite gullible and willing to print a story on "how AI/ML is now finally
doing breakthroughs in math" every year. Uncritical consumers of the hype
coming out of the well organized AI-company PR departments.

Here is a part of a 4-page rebuttal I wrote them last year:
https://sympa.inria.fr/sympa/arc/coq-club/2020-09/msg00017.html .

More recent comments on the use of language modelling are here:
https://sympa.inria.fr/sympa/arc/coq-club/2021-08/msg00047.html .

And a discussion of what models like GPT can and cannot do based on a
simple GPT "formal proof" example is at https://youtu.be/E7kbK9m3g-U?t=30974
.

Josef

On Tue, Dec 7, 2021 at 4:46 AM JOSEPH SHIPMAN <joeshipman at aol.com> wrote:

> Anyone want to comment on this?
>
>
> https://www.quantamagazine.org/how-close-are-computers-to-automating-mathematical-reasoning-20200827/
>
> Sent from my iPhone
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: </pipermail/fom/attachments/20211210/f861ef4c/attachment-0001.html>