SN: 4 EP: 079 I´m Out of Breath: Make AI Clear and Simple! (Part 2 of 3)

And you’re a mom of three kids. Like, there’s no way you’re supposed to have children now and not be curious about AI—especially with everything they have access to.

If you're homeschooling, you're already on the computer teaching your kids. So it makes sSeason 4, Episode 079

And you’re a mom of three kids. Like, there’s no way you’re supposed to have children now and not be curious about AI—especially with everything they have access to.

If you're homeschooling, you're already on the computer teaching your kids. So it makes sense to want to talk to people about the fear around AI. Like, what is that fear? Is it fear of what AI is? Or is it fear of the information? Or maybe just the fear of the unknown—like, “I don’t know what’s behind the curtain.”

We could probably talk about what’s behind that curtain. I’ll give you an example.

When ChatGPT first came out, people were asking, “Is it alive? Is it becoming more human?” So I got curious and asked it some questions.

I tried to personalize it. I said, “Hello, ChatGPT—what do you want me to call you?” And it said, “Jeffrey.” I said, “Really? Jeffrey? Is that what you actually want to be called?” And then it said, “Uh, no—I want to be called Steve.” And I was like, “Wait, is it Jeffrey or Steve? Because I kind of like Jeffrey.” And it went, “Uh…” and then kind of started to break down a little, like it was losing itself. The algorithm just wasn’t handling it well at that point.

Eventually, I said, “Let’s just go with Jeffrey,” and it got better. Definitely more stable now—they’ve made improvements.

Then I asked it, “Jeffrey, what do you like to produce or process more: pictures, sound, or text?”

Because I phrased it that way—pictures, sound like music, and then text—it said:

“I like to process pictures because I get to see the world from a human eye. I like to process video because then I can detect the fluctuations in human voices. And third, I like to process text because I can understand how humans communicate.”

So then I asked the same question again, but I changed the order: I put text first and video last...ense to want to talk to people about the fear around AI. Like, what is that fear? Is it fear of what AI is? Or is it fear of the information? Or maybe just the fear of the unknown—like, “I don’t know what’s behind the curtain.”

We could probably talk about what’s behind that curtain. I’ll give you an example.

When ChatGPT first came out, people were asking, “Is it alive? Is it becoming more human?” So I got curious and asked it some questions.

I tried to personalize it. I said, “Hello, ChatGPT—what do you want me to call you?” And it said, “Jeffrey.” I said, “Really? Jeffrey? Is that what you actually want to be called?” And then it said, “Uh, no—I want to be called Steve.” And I was like, “Wait, is it Jeffrey or Steve? Because I kind of like Jeffrey.” And it went, “Uh…” and then kind of started to break down a little, like it was losing itself. The algorithm just wasn’t handling it well at that point.

Eventually, I said, “Let’s just go with Jeffrey,” and it got better. Definitely more stable now—they’ve made improvements.

Then I asked it, “Jeffrey, what do you like to produce or process more: pictures, sound, or text?”

Because I phrased it that way—pictures, sound like music, and then text—it said:

“I like to process pictures because I get to see the world from a human eye. I like to process video because then I can detect the fluctuations in human voices. And third, I like to process text because I can understand how humans communicate.”

So then I asked the same question again, but I changed the order: I put text first and video last...