3 ms·

At least lay the blame correctly: humans get this wrong and they write it all over the internet and we then train our models on incorrect information. An aston

At least lay the blame correctly: humans get this wrong and they write it all over the internet and we then train our models on incorrect information. An astonishing amount of what we each believe is plain wrong, especially when we have little stories about it.

hackernewds 3y ago
which means GPT will never be smarter than humans? same conclusion
- axutio 3y ago
  No, but instead means that GPT can't be smarter than the people who are writing the published content on the web (generally, the very smartest humans if you're considering high quality published works alone). If you could do a perfect job of training a GPT on the entirety of published academic literature, the total of what that GPT could spit out would be limited by the knowledge contained in academic literature. At the same time, you'd have created a tool that is cheap and does a good job of synthesizing knowledge/answering questions across all disciplines. The model will never replace the scientists who are working at the very bounds of their fields, but it doesn't have to in order to be extremely useful, even useful enough to replace a majority of knowledge workers. Just because GPTs can't be smarter than the smartest humans doesn't mean they can't be smarter than most humans.
  - rocqua 3y ago
    Whilst that is probably true, it is not necessarily true. Nothing fundamentally prevents GPT from synthesizing new and novel insights. If a novel insight is trivial to notice when combining knowledge from 4 disciplines, it could be neigh impossible for humans to find, and obvious to GPT. Or perhaps the insight isn't trivial but follows analagous reasoning from some other obscure result. Or perhaps a million other things. I don't mean to claim GPT will do this. I just mean to point out it can't be fully excluded GPT is able to.
krupan 3y ago
This is still a net negative for AI. If all it does is regurgitate bad info at a higher speed than humans do, it's not a very useful tool is it?

hackernewds 3y ago

which means GPT will never be smarter than humans? same conclusion

axutio 3y ago
No, but instead means that GPT can't be smarter than the people who are writing the published content on the web (generally, the very smartest humans if you're considering high quality published works alone). If you could do a perfect job of training a GPT on the entirety of published academic literature, the total of what that GPT could spit out would be limited by the knowledge contained in academic literature. At the same time, you'd have created a tool that is cheap and does a good job of synthesizing knowledge/answering questions across all disciplines. The model will never replace the scientists who are working at the very bounds of their fields, but it doesn't have to in order to be extremely useful, even useful enough to replace a majority of knowledge workers. Just because GPTs can't be smarter than the smartest humans doesn't mean they can't be smarter than most humans.
- rocqua 3y ago
  Whilst that is probably true, it is not necessarily true. Nothing fundamentally prevents GPT from synthesizing new and novel insights. If a novel insight is trivial to notice when combining knowledge from 4 disciplines, it could be neigh impossible for humans to find, and obvious to GPT. Or perhaps the insight isn't trivial but follows analagous reasoning from some other obscure result. Or perhaps a million other things. I don't mean to claim GPT will do this. I just mean to point out it can't be fully excluded GPT is able to.

krupan 3y ago

This is still a net negative for AI. If all it does is regurgitate bad info at a higher speed than humans do, it's not a very useful tool is it?