Dr. John Torous

But there is a world where, without, again, getting on the clinical side, we can use AI to really improve access to good quality information, personalized information, skills-based things that don't cross into treating a mental illness.

And there's no reason not to deliver on that potential if we can do it safely.

I think the greatest risk is, again...

losing the trust of patients and clinicians by over-hyping the space, by making it so dangerous that regulators have to go far beyond what Holly's talking about.

She's just saying, I want to know what's good and label it.

But if it becomes so dangerous that it becomes a political issue that people are afraid of it, that's where you're going to have a different issue.

So I think that we have to be

mindful in what we say it can do.

And if we're going to make medical claims, back those up.

Because if we do push it that way, we will break things, we will harm people, and we will miss the opportunity to have this work as we all want it.