Seems a good time to (re-)post my *2017* opinion piece on super intelligence and AGI - still relevant, still valid.

www.otago.ac.nz/otagobulleti...

Thank you! Signed and re-posted. Very disappointing to see the President of the European Commission repeating such ridiculous claims uncritically.

Yes, and also LLMs change so conclusions drawn on version X won't always continue to hold with version X+1 ...

(minor: I think it's arguably too negative to say that they hardly share - there is some sharing, although it is limited, and, perhaps also crucially, presumably quite selective ...)

I think it's problematic to view the AI research community as having a single agenda. It's diverse, and there are many views.

Also worth reminding those who may not be aware that LLMs are a sub-field of Deep Learning, which is a sub-field of Neural networks, which is a sub-sub-field of AI.

The thing that gets me is that of all communities, I’d certainly expect the AI research community to be very aware of the limitations of LLMs!

So nine women, working together, can complete one pregnancy in a month? 🤷

Described accessibly (Hopefully) in a Conversation article.

A leading Ai conference ought to know better!