I'm Alan Alda and this is clear and vivid conversations about connecting and communicating. Let's say I wouldn't do this, but let's say I wanted to create a whole influence campaign

in my basement using my personal computer.

I can sit there and I can say to the chatbot... in my personal computer.

I can sit there and I can say to the chatbot,

create a person, let's pick a stereotype, we'll pick a woman,

she's middle class, she is pro safety, she's pro-sighty

pro-sifty she's pro-children and she doesn't like you know this group she

doesn't like Indians or something like that okay

then you say take that person and feed it back in and say write the program to generate 500 other people very different from her.

The fact that the computer can allow me to then make

replicas, men and women different voices and so forth, that's scary. You would end up

with a whole influence campaign of fake people and the social media people would say,

oh my God, there's a whole new trend.

That's Eric Schmidt.

For many years he had at Google where he had a bird's eye view of the dramatic evolution of artificial intelligence.

My conversation with Eric is the second in our three-part series on AI.

And while he's alarmed by the many dangers of AI,

especially his ability to create fake people

in this election year, he's also enthusiastic

about the huge opportunities he sees for AI benefiting medicine,

education, and the tackling of global problems like climate change.

He's now actively involved in international efforts

...