← Back to context Comment by vegnus 11 hours ago You're not reading it right. The bug was found using a vibecoded fuzzer. 4 comments vegnus Reply 12j3afAv 11 hours ago I wonder from where Claude stole this fuzzer. pjankiewicz 11 hours ago Or it used something called an "analogy" which is a valid way to solve new problems. criddell 9 hours ago The very first line of the bug report:> This is a bug found with our fuzzer: https://github.com/daedalus/fuzzer/ LoganDark 3 hours ago I think they're talking about the misconception that LLMs can only ever regurgitate their training data verbatim enough to constitute mass copyright violation. And that that's therefore "stealing"
12j3afAv 11 hours ago I wonder from where Claude stole this fuzzer. pjankiewicz 11 hours ago Or it used something called an "analogy" which is a valid way to solve new problems. criddell 9 hours ago The very first line of the bug report:> This is a bug found with our fuzzer: https://github.com/daedalus/fuzzer/ LoganDark 3 hours ago I think they're talking about the misconception that LLMs can only ever regurgitate their training data verbatim enough to constitute mass copyright violation. And that that's therefore "stealing"
pjankiewicz 11 hours ago Or it used something called an "analogy" which is a valid way to solve new problems.
criddell 9 hours ago The very first line of the bug report:> This is a bug found with our fuzzer: https://github.com/daedalus/fuzzer/ LoganDark 3 hours ago I think they're talking about the misconception that LLMs can only ever regurgitate their training data verbatim enough to constitute mass copyright violation. And that that's therefore "stealing"
LoganDark 3 hours ago I think they're talking about the misconception that LLMs can only ever regurgitate their training data verbatim enough to constitute mass copyright violation. And that that's therefore "stealing"
I wonder from where Claude stole this fuzzer.
Or it used something called an "analogy" which is a valid way to solve new problems.
The very first line of the bug report:
> This is a bug found with our fuzzer: https://github.com/daedalus/fuzzer/
I think they're talking about the misconception that LLMs can only ever regurgitate their training data verbatim enough to constitute mass copyright violation. And that that's therefore "stealing"