saying AI will ruin humanity’s existence or bring “disempowerment” of the species is a completely awful view that has no way of happening just simply due to the fact that its not profitable.

The economic incentives to churn out the next powerful beast as quickly as possible are obvious.

Making it safe costs extra, so that's gonna be a neglected concern for the same reason.

We also notice the resulting AIs are being studied after they are released, with sometimes surprising emergent capabilities.

So you would be right if we would approach the topic with a rational overhead view, but we don't.