Agree that a real source would be better but I think a Claude-source is better than nothing at all, especially in a case where the point is just to get the rough magnitudes and the details don’t need to be exact. I’m not sure if editing my comment to just the first paragraph would be an improvement.
We’ve all seen the reports on how LLMs get good results on standardized exams and tests of general knowledge, so on this kind of question I’d presume it to be basically reliable (again, as long as we’re talking just rough estimates and magnitudes, and take care not to ask leading questions). If my comment had instead included my own Fermi calculation or a random Wikipedia link, I’m guessing that that would have been more accepted, even though I doubt it would have been more reliable.
Agree that a real source would be better but I think a Claude-source is better than nothing at all, especially in a case where the point is just to get the rough magnitudes and the details don’t need to be exact. I’m not sure if editing my comment to just the first paragraph would be an improvement.
We’ve all seen the reports on how LLMs get good results on standardized exams and tests of general knowledge, so on this kind of question I’d presume it to be basically reliable (again, as long as we’re talking just rough estimates and magnitudes, and take care not to ask leading questions). If my comment had instead included my own Fermi calculation or a random Wikipedia link, I’m guessing that that would have been more accepted, even though I doubt it would have been more reliable.