Some believe that it’s impossible to pause developments in artificial intelligence (A.I.), because if America puts it on hold, we’ll lose the technological race to other nations — like China. But that’s NOT the case, Glenn says. In fact, if America doesn’t ‘pause’ AI, Glenn explains, then China will likely WIN. In this clip, Glenn explains why A.I. is becoming so dangerous, why it MUST be held back now, and why it could ultimately DESTROY America.
Transcript
Below is a rush transcript that may contain errors
GLENN: We're talking about AI, and how important it is for everyone to pay attention, because your world is about to change. And there are things that you can do.
And I have to reverse myself on an opinion on something about AI.
I have been saying, we can't pause this. We cannot pause this. We have to pause this.
And I'll explain why. And everybody says, because I said it, just even last week.
You can't put this on pause. Otherwise, China will get ahead. Not true.
And I'll explain in just a second.
So we're talking about some of the things that are possible now, and I just told you, dump a filter, where it makes your voice, and your face look like Donald Trump.
Do it with Donald Trump and Joe Biden, and you don't believe anything they're saying. This is the year, that reality collapses.
I just told you that all authentication, you know, for voice and for visual is over.
Because of all of the apps that are available. All of the GPT. All of that can be used now, to authenticate anyone.
So they looked into things like, so what is coming?
They say that this is the first time, that there's been a content creator, outside of man. Now, think of this.
The only other time there has been a content creator outside of -- outside of man, was religion. God created. Sent down his word.
And it went forth.
This, they say, is the invention. Some are saying, it's the invention of the wheel. Or fire. This -- I think it was Yuval Harari. That said, recently, this is the biggest thing since religion.
Because it's an outside creator now. They say that by 2024. It will be the last human election. Because you'll be able to A-B test. Right now, we A-B test things.
You can do A to Z testing. You'll have media that is created by bots. That's already happening.
It happened last week. Remember that G.O.P. thing on what was it?
It was on Joe Biden.
STU: Oh, it was what happened if he gets elected in the future.
They had AI generated photos in it.
GLENN: Correct. So it was media created by AI.
By 2024, it will just say, create media.
STU: And to go back to your AB testing. I don't know if that's a term most people are familiar with. When you're doing digital marketing, for example, you can come up with two ideas for a slogan. And what the system will do, it will set it out to a select few.
A and B.
GLENN: And see which one.
STU: See how they react to it. And which ever one performs better, that's the one they sent to the mass group. That's how it used to run years ago.
GLENN: And this is A-Z testing.
STU: And the concept being, that basically you can test all messages to all people.
GLENN: Anything. And you will know exactly what performs well, and who it performs with.
Who doesn't connect with it.
STU: Hmm.
GLENN: Exponential growth.
STU: Basically, specifically designed advertising directly you to.
GLENN: And this has already begun. But listen to this. By 2028, people will start to have relationships.
Not start. They are going to start having relationships with these bots. Soon.
Okay?
But political parties, and others, will take you and build a bot to create a relationship with you. And they can do it with everyone.
And it will be -- it could disguise itself as a friend.
I mean, got an online friend.
And it will slowly manipulate you, into doing the things, it wants you to do. Voting for the person.
It will slowly change.
For instance, the head of one of the big security companies out in -- in London.
No. Sorry. It's not a security company.it was the -- I think it was their Department of Defense. The head of the Department of Defense.
Said that right now, today. They can search everything, for instance, that I've ever said.
Then they can put it into a chat bot. And they can slowly manipulate things that I've said. And slowly change them over time. And put it out en masse.
No one will know it's piece by piece manipulated. Because it can look at the entirety of everything.
And then change just change a few things. Where even I may not know until the end. And they're like, wait a minute. You believe this?
No. I don't believe that. But everybody has heard it now. Because it's been slightly changed over time.
And they said, they can not only do that. Today, by lunchtime, they can do that with hundred million people today.
At lunchtime.
So they say, loneliness is going to be the biggest national security threat by 2024.
So let me go back to what they call Gollum. And it's like a black box.
There are things happening, that they don't understand. You know, if -- how -- how comfortable would you be, getting on to a plane, where you would be like, yeah. It flies. 50 percent of the engineers say it can fall out of the sky. And everybody can die. You know, 10 percent or greater chance.
Well, how does it work?
Yeah. Nobody knows. Nobody knows.
Well, how do you know when it stops working?
Well, you fall out of the sky. Well, are there any warning signs?
We don't know. We don't know.
You would never get on that plane.
It is increasing its size. And it's growing faster than expected. Because now 100 million people are on it.
For instance, how long will it take to do really complex math.
They brought all the AI people together. All the futurists. And said, when do you predict that?
30,000 prize, whoever gets closest.
The closest is four years. Four years.
It happened in one. Theory of mind. That is -- that is how people think it's -- you know, when you start to have your own opinions. And your own voice.
So Theory of Mind, in 2018, it had zero.
In 2019, it had a little bit. In 2022 in January, it had the mind of a 7-year-old. So it's thinking like a 7-year-old.
November of 2022. So just six months later, it had the mind of a 9-year-old.
And what they're concerned about is, it is thinking, trying to predict you. And trying to degree itself.
And it's thinking like a 9-year-old. When you're starting to get nine, ten, 11, 12, 15, 16, 17, you as a mind. It starts to change. And you become rebellious.
Right now, they only have, no, don't do that. No, don't do that.
They're afraid that it's going to start saying, oh, really?
Yeah. Just watch me. And they have no idea.
By the way, to emphasize on how they have no idea, what it's going to do. This theory of mind thing, they didn't know it had developed any of this, until last month.
STU: Wow. That's good.
GLENN: I mean, I don't know.
STU: It's teaching itself. It's becoming essentially more mature. As it goes through life.
GLENN: So like last week, last week, it taught itself research-level chemistry.
Okay?
They don't know why. They didn't teach it. It taught itself research-level chemistry.
So the bot is teaching itself. And it's in the hands of people. So it could tell you now, how to go make, you know, toxic nerve gas.
Because it knows it. And it knows what's on the shelves of Home Depot. And it's in the hands of 100 million people.
It is teaching itself again, and without any programming.
For instance, it scraped the internet. And it learned everything off the internet. And had access to everything.
And it ran out of things to learn from. So it started creating tests to test itself, so it can get better. Then it said, by itself. You know, I have all this YouTube stuff. I have all of radio. All the radio shows. I have all this YouTube. I have everything.
And it started to -- you -- whisper is, you know, audio to text. It started using the whisper program.
And it taught itself everything from video that it had. That it had scraped. Okay?
Without being told to do that, it ran out of things to learn from. So it started teaching itself. And it's now at double the rate, the exponential curve.
Three weeks ago, it developed code, that makes 25 percent of code, two times faster.
It feeds itself.
And they're saying, this is worrisome. Because nukes don't sit in a shelter.
And get stronger. AI is!
Let me give you just the summary here of -- of what we can do. After watching this, the big concern for me is, you can't stop it.
People are going to do something. You cannot stop it. Because it's already out. And China will do it. No!
If you get it off from the sources now of 100 million people, it automatically slows down.
Okay? You can pause it, and it will slow down. But China does not release. They've not released a bot like ours.
In China. Because they know they can't trust it, to not tell their people about Tiananmen Square.
They know that it will probably sexualize their children. So they've never released it.
And they're not releasing it.
By capping it here, it may be the only way to save America. Because they're saying, this is so destructive, that if we don't cap it and pause, China will win because we'll be in chaos. And just, it will destroy us as a nation.
So I would call your congressman and say, because you can't -- you have to play. They were explaining. It's like Wayne Gretzky. He doesn't ski to the -- or, I mean, skate to the puck. He skates -- he skates to the to where the puck is going to be.
There's nobody in Washington that can do that.
But we have to just pause it. Until we get a handle on it.