@boazbaraktcs Such great advice! In fact all students should stop reading books and just let the chat bots do the thinking. I'm sure they will turn out to become really smart and well rounded people /s
@miniapeur Depends on your goals. If you think it's fun, sure.
To participate in formalization projects? Some basic lean knowledge helps a bit, but the last thing you need to be is a lean expert. GPT 5.5 can write the actual code.
And don't listen to anyone tell you otherwise.
@drjoshcsimmons But but we are all going to live in abundance and have an army of free robots serving us, as promised by the AI land... right? RIGHT???
@boazbaraktcs It's amazing how far detached from reality tenured profs can be... Did it ever occur to you that maybe a slight additional concern of mathematicians will be how to feed their families and pay their mortgage, rather than no longer "discovering mathematical truths"?
@itaisher@littmath@chaitinsgoose@OpenAI Exactly. And don't forget that lead scientist Bubeck is constantly assure us that "zero human input was needed" and everything was done completely "autonomous" by their model. It's almost as if they're scared of still having any trace of humans in the loop...
@AcerFur Well the (obvious) goal of OpenAI and every other AI lab is to make human expertise completely irrelevant. @SebastienBubeck keeps telling us how "fully autonomous" and "0 human input" the latest result is. With that I'm mind, this viewpoint makes perfect sense!
@AcerFur Becoming an AI tech bro and joining the force that is making your math colleagues irrelevant (and soon to be unemployed) seems to be working out well for you! π
@LucaAmb Nobody is going to pay you for teaching if the number of math students collapses which it surely will. Anyone not seeing that coming is naive beyond hope
@SebastienBubeck@octonion It's always good to be reminded that the core mission of the AI labs is to *enable* humans to achieve more/better/fascinating things instead of just replacing them...
/s
@davidbessis@wtgowers Well said. It's bizarre that there are several prominent mathematicians who are unwittingly(?) shaping the AI-lab narrative that "math = theorem proving", hence AI has "solved math" and we no longer need human mathematicians.
@LucaAmb It's going to be funny when the AI investors realize that Erdos problems are so important that some of them had already been solved for decades without anyone noticing and the more prominent ones, like the current one, have 0 practical relevance...