I think most technical people who are not AI researchers have a tendency to underestimate what frontier AI can do.

In my organization, the people who use AI the most are not the most productive nor the ones who generate value for the company. Some of the most productive employees, in fact, have a strict "no AI" stance. I understand how AI is helping me, and I wonder how much better these guys could get with some AI help, but I'm convinced a lot of people overestimate AI's capabilities to help them be more productive.

Have you stopped overestimating it yet?

It’s not about what the AI can do, it’s about your ability as a human to keep up with the details so you can make effective decisions. You the human are ultimately responsible.

Because where dangers lie is when you lose track of that one tiny important thing that turns out to really really matter.

The AI won’t get fired, you will.

Have you stopped beating your wife?

Given that AI CEOs are in the habit of frequently lying about what frontier AI can do, it's hard to get a good grasp either way.

We can estimate from our hands-on experience and go from there, but very few people even have acess to the latest models. Some of the stuff I've read makes it sound like they're technically impressive but not all that useful in practical scenarios.