That thing where somebody breathlessly talks about a “great” AI feature being added to an app they use, then show that said feature immediately breaks and doesn’t work properly, but will almost certainly be fixed soon, despite the flaw in question being baked into how LLMs work.
🤨
I keep seeing this happen and people never seem to be aware of it when they do it.
Follow
@baldur Do you think these are people who understand the basics of how LLMs work but are caught up in the hype/magical thinking around LLMs right now, or are these just people who fundamentally don't understand what an LLM does?
For example, when I see people asking an LLM if its previous answer is accurate, to me that shows that they don't understand what an LLM does at all.
