Clojurians Log v2
ai-skeptics 2026-08-21
https://lisper.cc/mechanical-sympathy Is this a reasonable argument against LLMs ?
Interesting.Cant we think of AI as a machine as well?
That is a good question. Perhaps not. If we were really sympathetic to machines, we wouldn't make it artificially intelligent ? 😉
They may not be thought of the same way as unintelligent machines. it's a good question to ponder though.
Will they understandNo, they won't.the limits of the human mind, be sympathetic and not spew slop?
@edward.hughes1911 Good article; thanks for sharing! It aligns perfectly with the issues I've seen over the years and the position I take as an engineering leader when this comes up, but Kent Beck makes the case more thoroughly and eloquently than I've been able to. This is a great piece to cite when making this argument with a metrics-obsessed C-level.
https://www.experimental-history.com/p/i-like-em-thick Nice article about the "deep" aspect of art like things, with AI skepticism comments at the end
Great article. I needed to be hands free and saw the audio button at the top. Maybe I'm just visiting the wrong sites, but I fully expected a TTS read of it. Imagine my surprise when the author read his own work to me, mistakes, digressions, and all. Fantastic Edit: I'd definitely recommend listening as you read along
😯 fantastic effort, and it shows the commitment of the author to the idea he's talking about—his post itself is thick
Thanks Arne. It was a good read. The psy book looks quite interesting but, as an IC, I realized this stuff often makes me want more from my employers. 😅 OTOH it looks like the kind of book that might help me know myself better and set better boundaries, expectations, etc. for myself. The talk by Richard Campbell was a really good intro! I might want to follow him a bit more; he spoke about a podcast that might be interesting.
Little penalty isnt the framing the actual study uses. It's a 20% decline
The authors of the study linked by https://clojurians.slack.com/archives/C0B7Q4XRM1U/p1787207782136449?thread_ts=1787183577.659249&cid=C0B7Q4XRM1U state in the conclusion on page 23 almost verbatim what the article says > By contrast, AI students who spend as much time on homework as non-AI students achieve similar exam scores and higher homework scores.