hochstenbach

I fear the same thing will happen as what I see in software developement. The volume of proofs that can be generated by anyone will be so big that it will overwhelm the system. Currently I really don't have the time anymore to check if all the bright results in computing projects make any sense. And not only in software developement. In general, I see preprint servers getting overwhelmed by article submissions, AI generated articles. I don't recall a wave of CAD generation in the past.

There are other studies on this topic with similar results across LLM systems:

Y. Sui, M. Zhou, M. Zhou, S. Han, and D. Zhang, “Table Meets LLM: Can Large Language Models Understand Structured Table Data? A Benchmark and Empirical Study,” in Proceedings of the 17th ACM International Conference on Web Search and Data Mining, Merida Mexico: ACM, Mar. 2024, pp. 645–654. doi: 10.1145/3616855.3635752.

C. Pang, Y. Cao, C. Yang, and P. Luo, “Uncovering Limitations of Large Language Models in Information Seeking from Tables,” June 06, 2024, arXiv: arXiv:2406.04113. doi: 10.48550/arXiv.2406.04113.

One would expect that if such studies indeed indicate that AI has an effect on early-career workers in AI-exposed occupations, that this would be a global effect. I wonder if there are good comparable non-US studies available.

Humans are not allowed to do what AI firms want to do. That was one of the copyright office arguments: a student can't just walk into a library and say "I want a copy of all your books, because I need them for learning".

Humans are also very useful and transformative.

Quanta magazine has an article that explains in plain words what the researchers were trying to do : https://www.quantamagazine.org/chatbot-software-begins-to-fa...

Also we need to keep in mind that normal (legal) traffic for typical universities is in the million downloads per year.

Justin, I have to admit, I unfollowed you a while ago for this very reason. I'm reading a lot of the same blogs and news sites as you do and was a bit annoyed by the fact that your tweets didn't had any sources ("via ..."). You could provide me with a great service providing pointers to great news sources instead of just copying their headlines. This would've helped me in search for content. The second problem is by spreading out the stories over the day, they became a bit out of date it felt a bit pivoting of information. I'm sorry.