3 ms·

>You'd usually be better off training on trade publications, scientific journals, or fandom than Reddit. Bold claim. There are plenty of engineers and doctors

>You'd usually be better off training on trade publications, scientific journals, or fandom than Reddit. Bold claim. There are plenty of engineers and doctors on reddit answering nuanced questions that arent talked about in any journals. That is what is missing. The absurdly specific stuff that reddit gets. Sure you can answer with nonsense that sounds realistic, but you could also answer with the exact text that solved the problem in the past. Its the difference between asking the question "How do I save money at the grocery store?" and getting generic answers like "Make a budget" and getting specific answers like "Buy fresh rather than processed".

dmonitor 3y ago
The replies of engineers and doctors are significantly outweighed by the responses of bored teenagers.
- hospitalJail 3y ago
  Yeah, but I imagine bored teenagers are not patrolling the embedded subreddit.
haldujai 3y ago
> There are plenty of engineers and doctors on reddit answering nuanced questions that arent talked about in any journals. There are also plenty of people answering nuanced questions with complete BS. It's not that bold to claim that as when taken as a whole the signal-noise is worse on Reddit than in a publication.
marcosdumay 3y ago
The GP ignored books for some reason. But you get very different kinds of knowledge from each of those. They are not really comparable. You won't find any useful explanation about how to solder surface electronic components in a PCB in a book, and you won't find a deep enough explanation about how to dimension an uncoupling capacitor for an amplifier in Reddit.
michaelt 3y ago
Well, it depends on what you're asking the LLM to generate. If I ask an LLM to write me an essay in the style of a caveman, the LLM simulates a redditor simulating a fictional caveman (who speaks english but with short words and bad grammar). After all, that's what was in the training data. If I ask the same LLM to write me an essay in the style of a Harvard professor - who's to say the same thing doesn't happen? So I can see how, when training a model, you might want it to know what's really written by a Harvard professor, and what merely professes to be.
importantbrian 3y ago
> There are plenty of engineers and doctors on reddit answering nuanced questions that arent talked about in any journals. There are lots of really good subreddits that have expert feedback, but I think the issue is that going through all that and separating the wheat from the chaff is going to be a much more involved process than feeding in those other sources of expert opinion.
- gingerrr 3y ago
  Ironic that reddit self-sabotaged the one popular place to converse with verified experts in their field so hard (AMAs) as that was probably the most trustworthy input source on the site considering participant vetting

dmonitor 3y ago

The replies of engineers and doctors are significantly outweighed by the responses of bored teenagers.

hospitalJail 3y ago
Yeah, but I imagine bored teenagers are not patrolling the embedded subreddit.

haldujai 3y ago

> There are plenty of engineers and doctors on reddit answering nuanced questions that arent talked about in any journals. There are also plenty of people answering nuanced questions with complete BS. It's not that bold to claim that as when taken as a whole the signal-noise is worse on Reddit than in a publication.

marcosdumay 3y ago

The GP ignored books for some reason. But you get very different kinds of knowledge from each of those. They are not really comparable. You won't find any useful explanation about how to solder surface electronic components in a PCB in a book, and you won't find a deep enough explanation about how to dimension an uncoupling capacitor for an amplifier in Reddit.

michaelt 3y ago

Well, it depends on what you're asking the LLM to generate. If I ask an LLM to write me an essay in the style of a caveman, the LLM simulates a redditor simulating a fictional caveman (who speaks english but with short words and bad grammar). After all, that's what was in the training data. If I ask the same LLM to write me an essay in the style of a Harvard professor - who's to say the same thing doesn't happen? So I can see how, when training a model, you might want it to know what's really written by a Harvard professor, and what merely professes to be.

importantbrian 3y ago

> There are plenty of engineers and doctors on reddit answering nuanced questions that arent talked about in any journals. There are lots of really good subreddits that have expert feedback, but I think the issue is that going through all that and separating the wheat from the chaff is going to be a much more involved process than feeding in those other sources of expert opinion.

gingerrr 3y ago
Ironic that reddit self-sabotaged the one popular place to converse with verified experts in their field so hard (AMAs) as that was probably the most trustworthy input source on the site considering participant vetting