Gigaom AI Minute – May 15

Transcript

I often get those auto-dial calls, it's a person's voice saying, "Hi, can you hear me okay?" only to realize it's not really a person it's just a machine. And it got me thinking, we're always going to want to know when we're dealing with a human versus when we're dealing with a machine.

As the technology gets better and it's harder to tell on its own who you're dealing with, do you think it's possible that we'll have an agreement that mechanical devices--such as the assistants we have now--won't have human voices, that they'll have digitized voices that are clearly mechanical? So that somebody can easily know that they are dealing with a machine and not a person.

I think this could actually be the case, that something that sounds too human might leave the user feeling that they've somehow been duped, and I don't think that won't go over well. So we might very easily get to a world where although we can emulate human voices perfectly, we choose not to.