pull down to refresh

Interesting! Especially the one buried deep down:
This basically is the reality check on #1237966
I read this as that currently, GenAI is impacting nerds and media more than expected, and nearly everything else less.
I'm surprised legal is so small. It actually makes me question a bit the methodology of this paper, or its data sources. I'd think for legal it'd at least be used as a souped up search engine.
reply
It actually makes me question a bit the methodology of this paper, or its data sources.
from report:
Just as the OpenAI metric has limitations, so too does Anthropic’s usage data. Figure 22 shows the occupational shares of all “conversations” with Claude (the AI chatbot), and illustrates the occupation groups that are over- and under- represented in this usage, compared to their exposure ranking and employment share.
reply
I think the google numbers would reflect this better?
reply
Can’t say I agree with the nerd bit! 😂
I also thought this chart was super interesting, it gives a good idea of where AI’s doing a solid job, and more importantly, where it could be doing way better.
reply
33 sats \ 1 reply \ @optimism 5h
it gives a good idea of where AI’s doing a solid job
I'm not convinced that this is a measure of AI doing a solid job versus where AI giants' FOMO marketing is doing a solid job, yet.
I should really go code some javascript and also be a 1337 vibecoder, because in system programming it definitely doesn't work.
reply
Looks like I didn’t see the full picture. What you’re saying makes sense. My thinking was, if it's being used more than expected, it must be a good product. But like you said, that might not be true.
reply