Lots of people in the AI world who say "computers will be intelligent" define intelligence along the lines of "the ability to do what computers do"

People warn against anthropomorphising AI. The reverse is rampant: People (wrongly) thinking humans work like a computer

You could probably produce a good novel if you'd read every piece of (out of copyright) literature published in human history, like AI has. But it can't write one interesting paragraph. Is that remarkable?

The authors of AI 2027 misrepresent scientific research

Here:
oscarmdavies.substack.com/p/000017-on-ai2027

Sutskever doesn't give us any evidence machines will be intelligent.

"I have opinions ... but unfortunately, circumstances make it hard to discuss in detail"

Here:
oscarmdavies.substack.com/p/on-the-sutskever-and-dwarkesh-interview