3 ms·
> ChatGPT, asked to translate to a high schooler: "Basically, this study found that young people with depression sometimes struggle to break automatic habits, e
by JumpCrisscross 1y ago
> ChatGPT, asked to translate to a high schooler: "Basically, this study found that young people with depression sometimes struggle to break automatic habits, especially when they’re trying to avoid something bad and there isn’t a prize or reward for doing it."
The translation is just as much of a word salad as the original, just with simpler vocabulary. Worse, it mangles the key point.
Prepotent responses aren't "automatic habits," but overriding responses (e.g. pain) [1]. The "sometimes" qualifier is unsubstantiated when describing "association". And the struggle isn't amplified ("especially") when avoiding something bad absent reward, the first part of the sentence is conditional upon the absence of a reward. (It's nonsense to say pool drownings are especially common in pools.)
[1] https://dictionary.apa.org/prepotent-response https://dictionary.apa.org/prepotent-response
- cheesecompiler 1y agowow, you managed to make it more convoluted than the original.
- JumpCrisscross 1y ago> you managed to make it more convoluted than the original Not relevant. I'm not trying to break down the original text, that was done adequately by the top comment. I'm showing why the LLM summary is nonsense.
- floatrock 1y agoAnd engineers don't really deal with frictionless spherical cows, but high schoolers might. All models are wrong, but some are useful.
- JumpCrisscross 1y agoNon sequitur. This LLM’s summary is wrong and useless. It would be like talking about frictionless spheres and then applying the wrong equations.