There is very little evidence or theoretical support for a hard limit on AI capability. In fact, overwhelmingly, most of the evidence and thinking supports the notion that over time, the capabilities of AI will greatly over-match human beings in specific contexts. The same even goes for general intelligence.

How benign or malicious an entity is, is orthogonal to how potentially dangerous it is. Humans are demonstrably very dangerous. What we're creating has the potential to be orders of magnitude more dangerous.

So yes, we desperately need alignment! And even if we achieve it, how do we know the AIs won't convince Homo sapiens to simply stop breeding and fade away in alignment with our values?

Hard to know the future.