Automated LC review — now right here in the GuildUpload documents → AI discrepancy analysis → correction guide.
NEW Start LC Review →
← Lounge

My thoughts on the resistance of mathematicians in the AI era

Machine translated

If you look closely, most of the many concerns regarding AI ultimately boil down to a struggle for livelihood.
As one example, mathematicians are expressing many concerns regarding recent proofs generated by AI.
Since these are obvious arguments, I won't quote them specifically,
but most of them essentially boil down to: "I have conducted immense research in this field, and I fear it might all become in vain because of AI."
They argue that a proof not performed by a human is meaningless,
but the funny thing is, in mathematics, there are many cases where progress is made by assuming a condition—such as "if an unproven theory that plays a crucial role in the mathematical community is true"—and exploring or expanding related knowledge based on that assumption.
For example, in the case of the Taniyama-Shimura conjecture, which is related to Fermat's Last Theorem, numerous studies already existed under the assumption that the theory was true before it was actually proven.
Applying this to AI, even if a human hasn't proven it (assuming the Shimura conjecture is true), many follow-up studies can still be conducted in relation to it. Of course, it is better when humans provide the proof and expand their own cognitive horizons in the process. However, what I am saying here is that while there are cases where both the process and the result are good, there is also meaning in having a good process or simply knowing the answer.
Of course, I understand what mathematicians are worried about. Rather than simply worrying about the reality of AI performing proofs, they fear that they will gradually lose the reason to study mathematics. However, this is not an AI problem, but a human problem. The advancement of AI is no longer a variable, but a constant.
In other words, rejecting AI is impossible; one must find existence and meaning within it. That will clearly be different from what existed before. AI is already penetrating various fields beyond mathematics and destroying our habits and existing knowledge. Yet, on the other hand, we are adapting to the AI era.
In my view, mathematicians in the AI era seem to be evaluating AI and its changes while maintaining non-AI attitudes and mindsets. Mathematicians are no exception, and they must strive to accept the changes in their academic disciplines brought about by the AI era.

Comments 0

No comments yet
0