3 ms·
A brilliant colleague Janelle Shane works specifically on how language models fail (obnoxiously often). This is her line of research to show how LLMs are overhy
by srvmshr 4y ago
A brilliant colleague Janelle Shane works specifically on how language models fail (obnoxiously often). This is her line of research to show how LLMs are overhyped / given more credit than should be. I think her fun little experiments will give a better answer than I ever can :).
She's on substack & Twitter
https://janellecshane.substack.com/p/okay-gpt-3-candy-hearts https://janellecshane.substack.com/p/okay-gpt-3-candy-hearts
- dekhn 4y ago
- rootw0rm 4y ago+1 for SHRUBBERY'S PRECIOUS if anything, this proves that candy hearts should be larger, and less disgusting
- jameshart 4y agoI don't understand how frequency of failure is a relevant concern here. Humans are frequently monumental idiots. The human language and inference model produces nonsensical, dangerous, and stupid results with alarming frequency. (not to say I believe language models are sentient, but... you need something more than 'sometimes they spout made up BS' to refute the claim, because it turns out that is not particularly indicative of sentience.)
- gnramires 4y agoTo be fair, I didn't know what a exactly a 'candy heart message' was. I probably would have failed this particular Turing test.
- astrange 4y agoI think it would be fair to give you Google access or at least an encyclopedia.