How intelligent will AI become?

Whenever I talk somewhere about how good AI already is and how good it will become, and I did this many years ago, it rains down comments that all emphasize what AI can’t do today and that it makes mistakes and that humans are still much more intelligent. The AI researchers themselves say that it’s a stochastic parrot. ChatGPT doesn’t even understand what it’s writing. That’s true to a certain extent. But: such comments regularly miss the point. Because: We are still AT THE BEGINNING! And: The foreseeable future development will be exponential. It is misleading and dangerous to look at today’s shortcomings and make decisions based on them.