But then there are also extremely positive things. So yeah, it should be thought of, in my opinion, as one has to be mindful of the risks and and it will be increasingly problematic to know what's true and what's not. And I know one positive example is, if you know Montreal AI, Vincent Boucher, he developed AI agents, and then taught them to be doctors.
And then you could, and then he did the measurement against real doctors. And it turned out that that real doctors had like 30 to 40% success in putting the diagnosis on a patient at the first kind of occasion when they come to see the doctor. But the AI doctors had a success rate of 70%.
And that's that's because you can have like one AI becoming one doctor. And then you have another one becoming another sort of a doctor. And then you could have them say, yeah, like hundreds and thousands of things, and then they connect them together.
And then they all learn what everybody else know. So you kind of get 1000 doctors in one. So it's by that being said, I don't know, I wouldn't, if I would feel especially comfortable seeing an AI doctor.
I would feel a bit would feel cold and not not nice. That's kind of I'm, I'm wary about AI, I use it a lot. I love the tool.
But one should also think, like you said, probably about consequences. And, and there are risks. And I mean, almost anyone can get an AI, you can train it yourself if you want.
And of course, the shady people will also get and they will boost out really bad things and try to convince people and they will have a super smart helping them in that. That's, yeah, pros and cons. Big ones in both directions.
