Artificial Intelligence in Ophthalmology

Yeah, that's, that's fascinating, and it's, it's, it's really interesting how you got involved with this.

You know, I think a lot of people view AI with a lot of trepidation and fear, especially when ChatGPT first came out, and it was kinda taking over.

But you're right.

It's been coming for a long time.

If we think back, you know, all the way back to when I was a medical student in the nineties when the EKGs had the automatic readout, I mean, that's kind of artificial intelligence, right? Very basic.

And I think that we're starting to see that more and more in clinical practice.

It's certainly coming to ophthalmology.

What do you think the pros and cons are of AI with clinical practice in general? Anything we should be aware of?

Yeah.

I mean, it, it's-- th- this is sort of where the rubber sort of hits the road, right? Like, the moment we start thinking about how it's gonna touch people's lives in clinical practice, that's where all the concerns that we have really come into play.

I do think that it's very important to do rigorous validation of technology before you start relying on it.

I think a lot of folks have shown that if you use these tools really effectively, they can make you very efficient and a better clinician.

So that's the, that's the, uh, state that we should all, as a profession, be trying to push our field towards.

What I don't want to see happen is where humans are basically surrendering their autonomy and their decision-making to these models.

They should be using it in partnership and to help them become better clinicians and take better care of, of patients.

But what we don't want is sort of a regression to the automation, right? Like, we don't want basically all of human performance to be surrendered to these AI models, and that's where we haven't quite figured out what that looks like.

You know, we worry that if we do that, then in some ways it might make us slower.

It might actually not make us faster and more efficient.

But, you know, are there ways that we can insert AI into the clinical workflow where we might be able to have, um, you know, our cake and eat it too?

Yeah.

I mean, I completely agree.

I think, you know, AI has the ability to make us more efficient.

I hadn't actually thought about how it could make us less efficient.

But the goal, of course, is to be more efficient and deliver better care, but that better care relies on AI being correct.

I think we both agree that we're not quite there yet, but, you know, what happens? Is, is it gonna make us obsolete? I personally don't think so, but what do you think? And specifically for ophthalmology, as we're ophthalmologists, how do you think it can help us in certain aspects of our practices?