4 ms·
Recently an article came out where someone said that the company I work for is a big user of WebAssembly, but the reality is that we don't use it. After findin
by salynchnew 3y ago
Recently an article came out where someone said that the company I work for is a big user of WebAssembly, but the reality is that we don't use it.
After finding the contributed article (on a well-known news site, not Wired though), it looks like a tech founder might've been using ChatGPT to write an article about the uses for WASM. The arguments were generally sound, but I don't think that anyone did the work to manually check any of the facts they presented in it.
- notabee 3y agoThis is kind of like the advent of spellcheck, where a whole class of errors started to appear regularly in almost every article because publishers stopped paying for the human labor to manually review for things like homonym or word ordering errors. Except much worse, because it could allow spurious or even harmful facts to accrue and spread instead of just grammatical mistakes.
- visarga 3y ago> Except much worse, because it could allow spurious or even harmful facts to accrue It already did, even in the "purely human" era. I think LLM text will gradually become more trustworthy than a random website by consistency filtering the training set.
- kumarvvr 3y agoUnfortunately, it is more than likely that the training inputs to upcoming LLMs will be partly from older LLM outputs.