"God-like" is an unfortunate choice of phrase here, especially for the headline. Things like ChatGPT and other large language models are not bringing us significantly closer to anything superhuman, and the term just makes the concerns seem silly and science fictiony.

I'd rather people worry about the dangers that are uncontroversially approaching, or already here. Embedding bias and injustice in supposedly objective programs, sucking up and processing vast amounts of personal data and combining that with facial recognition to do massive surveillance, and so on.

It's scary and exciting to worry about "God-like" superbrains, but it also distracts us from things that are already happening (and that powerful human organizations would rather we ignored).