our response?

Solow Paradox

>>16731601 (OP)
>>16731603
more like obvious AI shill company lies

ChatGpt still makes mistakes with simple Python code, for C++ it's basically unusable.

>>16731601 (OP)
>unsourced chart
go back migger

>>16731704
go back jeet.
https://arxiv.org/abs/2507.09089

>>16731601 (OP)
Thanks doc

>>16731601 (OP)
Thank you doctor

>>16731601 (OP)
Why does the sample size get cut in half between the two groups?

>>16731601 (OP)
My response is that however much time a dev thinks AI will save, multiply that number by negative two, and you get the actual savings.

>>16731717
>were going to force random devs to use ai and see if they can work faster than normal
and im sure theres absolutely no learning curve whatsoever when doing so which the study took into account right?

>>16732034
read the paper, jeet.

>>16731606
Same for the local models. If set up correctly you can get them working nicely, but it still takes work to coax out anything that doesn't require correction or setup.

>copy textbook question into google
>Leave out a crucial part of the question
>google AI fills it in as if I had copied it
AI is literally just copying Chegg. Pic related