The Great America Show - INTELLIGENCE WILL WIN-HUMAN OR MACHINE？

00:00:00.000 Hello, everybody. I'm Lou Dobbs, and welcome to The Great America Show. Great to have you with us.

00:00:05.360 A beautiful day in America. Chairman James Comer's investigation into the Biden family

00:00:11.300 moving full speed ahead. Senators Ron Johnson and Chuck Grassley also have their own investigation

00:00:17.140 away. The two senators honing in on Secretary of State Antony Blinken and his wife and their

00:00:24.900 contact with Hunter Biden relating to Burisma. Burisma is the Ukrainian energy company that

00:00:31.980 paid Hunter Biden and his associates $11 million. Joe Biden was Obama's vice president for four of

00:00:39.900 the five years that Hunter Biden sat on Burisma's board. In December of 2020, Blinken sat for a

00:00:47.620 transcribed interview with Senate investigators in which he was asked, did you ever talk with

00:00:53.820 Hunter Biden on the phone? To which he replied, not that I recall. He was then asked, did you have

00:01:00.240 any other means of correspondence with him? Emails or texts? Blinken replied, no. Emails uncovered on

00:01:09.000 the infamous Hunter Biden laptop show that Hunter Biden in fact did urge Blinken through his wife to

00:01:15.540 speak with executives of a Democrat consulting firm representing Burisma. The firm is called Blue Star

00:01:22.600 Consulting and Burisma Consulting and Burisma hired them to improve their image amid allegations of

00:01:28.380 corruption. It's unclear what role, if any, Blinken's wife played in that scheme. But at that time, she was

00:01:36.500 also working for the State Department. She was the Assistant Secretary of State for Educational and Cultural

00:01:43.160 Affairs. And turning now to our woefully woke military that is falling far short of its recruiting goals,

00:01:50.980 the Navy has responded and confirmed that they are now using a drag queen influencer as their so-called

00:01:59.080 digital ambassador to attract recruits to the Navy. That ambassador is Joshua Kelly, U.S. Navy Yeoman,

00:02:08.440 second class. He goes by the stage name Harpy Daniels. This should be, at best, an utter embarrassment to

00:02:16.360 Secretary of Defense Lloyd Austin and an insult to all who serve the U.S. Navy. And speaking of great

00:02:25.000 Americans who served, we asked Navy SEAL Robert O'Neill, the man credited with killing Osama bin Laden

00:02:31.520 for his take on the Navy's new recruitment strategy. Here he is. I don't have a problem with what anyone

00:02:38.200 does when the doors close behind them. But coming out front here and showing a drag queen as being 0.99

00:02:45.800 the head of our, a yeoman has the face of the toughest Navy in the world. China is literally 0.86

00:02:52.640 laughing at us. Russia is laughing at us. Why is there so many wars going on in the world? The world

00:02:56.680 is a better place when America is strong. We're supposed to be ferocious, not fabulous.

00:03:01.920 Rob O'Neill, a great American. He'll be joining us next week here on the Great America Show.

00:03:08.620 The NSA cybersecurity director has a warning for Americans. He says, buckle up for generative

00:03:15.420 artificial intelligence. Director Rob Joyce says the National Security Agency expects the generative

00:03:22.920 AI to be the source for a lot of scams and problems to come. To get a better sense of where we are right

00:03:29.760 now with artificial intelligence, we bring in our guest, James Barrett. James is a journalist,

00:03:36.560 a documentary producer, and author. He's also one of the first people to sound the alarm

00:03:42.260 on artificial intelligence. James, great to have you with us here on the Great America Show.

00:03:48.240 And let's start with your judgment about where we are right now as we're caught somewhere between

00:03:53.400 GPT-4 and 5 and a nation that doesn't truly understand what's happening with artificial

00:04:00.460 intelligence. Thank you. Thank you. It's great to be here. You know, I hate to be right all the time,

00:04:08.200 but most of the things I predicted in our final invention have come to pass. And right now,

00:04:13.120 we're experiencing this, you know, collective kind of head rush and disorientation as, you know, no longer

00:04:21.600 do we get LAMDA and DALI, then GPT-3, 4, and auto-GPT are thrown at us. And it's restarting this cycle of

00:04:32.200 shock and acceptance. But now this is the new normal, the exponential development of thinking machines.

00:04:38.400 And of course, they have huge commercial potential. They have reckless emergent properties, and they're

00:04:45.380 they're shaking everybody up. Well, I'm amongst those. I will tell you, I'm shaken up by thinking

00:04:53.380 machines that can outthink me and produce papers in a matter of moments that would take any

00:05:02.120 grad student at least a couple of weeks to do a similar job. We're watching the potential

00:05:09.240 here of replacing just about everything we accept as normal, deep research, time taken.

00:05:19.880 There are multipliers of all kinds, whether in depth, whether in time, in knowledge. It is just

00:05:27.900 striking to think that I've got an app on my phone, that I can ask it any question, I will get an

00:05:34.440 intelligent answer. It may not be the right answer, and it may not be the one I sought. But it's

00:05:39.720 remarkable what we can do just simply with an app right now. What in the world lies ahead?

00:05:46.960 Well, it's, you know, I'm a skeptic of AI, but I'm just as fascinated as you are. I play with

00:05:53.980 ChatGPT all the time. I don't really use it for writing, but I ask it questions, and it's an

00:05:59.760 excellent, if you ask it to connect two ideas, you know, I'm just making something up like, you know,

00:06:06.820 chrome mating patterns and climate, it will write a persuasive and well-thought-out essay.

00:06:15.960 The danger, I think, that we all have is when we receive writing or speech from something, we

00:06:25.400 perceive that there's a mind behind it. And that's just our, you know, we've never heard,

00:06:31.220 we've never talked to a mind that wasn't a human mind. And so we give it a lot of authority. And as

00:06:37.340 you pointed out, ChatGPT is wrong about 25% of the time. I did an experiment with it where I asked

00:06:44.600 a question, and I said, provide citations. And then I looked up the citations, and they were

00:06:50.020 completely made up. Authors, book titles. It was amazing. It was just completely, they're just

00:06:57.140 completely wrong. So even the head of OpenAI said, don't trust this for anything important.

00:07:05.580 It's not ready yet.

00:07:06.960 You know, being wrong is one thing, but I've seen a number of accounts with those who are

00:07:13.940 spending some considerable time and have great expertise, talking about the hallucinations of AI

00:07:21.740 and various forms. And it doesn't seem to matter which machine, which example of AI is being

00:07:31.900 interacted with. It is a common, apparently, a common response endemic to AI at this stage.

00:07:45.380 At this stage, yes, absolutely. Because, you know, these large language modules,

00:07:51.620 they use deep learning techniques and massive amounts of data to summarize and generate and

00:07:59.520 predict new content. They're prediction machines. They predict what is the next best word,

00:08:06.000 what is the next best sentence. And they base that on the prompts that you give them, like,

00:08:12.120 you know, what are the mating habits of ravens and crows? The trouble is, they don't actually know

00:08:19.920 anything. There's nothing inside. There's no mind. There's no sense of self. There's nothing like

00:08:25.280 consciousness. They have the trappings of language, but not any of the understanding. So

00:08:29.700 you know that you can put a cup on your table and you can cup your hands.

00:08:35.780 These large language modules don't know what a cup or a table or hands are. 0.73

00:08:40.780 They're very excellent mimics. And that's why, to them, saying something that's untrue has absolutely no,

00:08:48.280 no, there's no demerits for that. It's not, it's not looking for truth or falseness. It's,

00:08:55.200 it's scanning the, uh, vast amount of, of, uh, text and books and, and articles. It's, and, and the

00:09:03.240 whole, the whole contents of the internet. And it's looking for a statistical alignment of, of words,

00:09:09.640 but it doesn't care if it's, if it's true or not. And it has no way, no mind to evaluate its truth.

00:09:16.260 And no way to do that. It separates from what we think of coding, uh, programming, uh, and producing

00:09:26.260 a result in inner in human interaction with our machines. But the power can't, seems like it is

00:09:35.020 quickly slipping from the programmer operator user, uh, to the machine itself, to the CPU that is becoming

00:09:42.740 something, uh, organic unto itself. It seems to be that way. And, you know, but there's a, there's a

00:09:50.620 big gap. Now what, what, what all these companies really want to do, they're making these, they're

00:09:55.600 making these, uh, these image generators and word generators, and they're really powerful and really

00:10:00.840 persuasive, but where they're headed is called AGI. As you probably know, artificial general intelligence,

00:10:06.740 intelligence, and that's human level intelligence in a machine. That's job number one for all these

00:10:11.420 companies for, for, for Google, for open AI, for deep mind, for meta. And they've said as much,

00:10:17.740 they've said that they want to get there because it's, it would be, it would be a huge, a huge product.

00:10:24.000 Um, but you know, they're getting there and as fast as they can, but there's a lot missing in these large

00:10:31.200 language modules before they can become as, before they can become smart. I mean, they have to know

00:10:37.920 about the world and they don't know anything about the world. Um, they need to be connected to what

00:10:43.580 they call a, uh, uh, just a, a, a database of common sense knowledge and, uh, or they need to be

00:10:51.720 embodied in a robot and learn about the world by going out and looking at it and seeing it and hearing it.

00:10:57.060 So, and, and for that, you can also use old fashioned programming. Um, they're not going

00:11:03.300 to get all the way through neural networks, which is what these large language modules are based on,

00:11:08.340 but they need old fashioned rule-based programming to, to complete the, the circle, to get to human

00:11:15.480 level intelligence. And, and I want to share with everyone the subtitle to your, your terrific book,

00:11:24.040 our final invention. And I want to, and I want to mention, if I may, the publication date. And by the

00:11:29.800 way, uh, we recommend our final invention, uh, James Buront's, uh, terrific book on AI, uh, to you.

00:11:38.440 It's available, uh, on, on, uh, Amazon, but here's the, here is the subtitle. Are you ready folks?

00:11:46.600 Artificial intelligence and the end of the human error. Now that's an attention grabber, uh,

00:11:54.040 James, that, that, uh, if, if you didn't get, if you didn't pick up on our final invention,

00:11:59.660 uh, the end of the human era, uh, is a, uh, a stone cold, uh, inspiration to, to read, read,

00:12:07.160 read, uh, give us your sense of, uh, you, you mentioned that you've had many of your predictions

00:12:14.620 and forecasts come to, to fruition to reality. Uh, how likely is it that we will see artificial

00:12:22.920 intelligence at a level sentient, uh, independent, uh, in the next decade?

00:12:31.500 I'd say it's very high. I used to say, uh, 2029, which happens to be Ray Kurzweil's anticipation.

00:12:39.220 And he said, by 2029, we should have at the price of a computer, a, uh, basically a, a human

00:12:46.340 level brain in a machine. Um, so I think we, we'll get to that and to, to AGI human level

00:12:54.900 intelligence within a decade. But the next step is the really, the really sensitive one.

00:12:58.840 And we call that the intelligence explosion and you can already see it forming. Now you have, uh,

00:13:04.220 you have, uh, chat GPT four and people have already said that chat GPT four will help program

00:13:10.520 chat GPT five. And there's also another GPT called auto GPT, which, which improves its own

00:13:16.660 programming. And there's talk about, you know, you know, how you improve the programming of

00:13:21.480 one of these models is you generate more data. And right now, believe it or not, we're running

00:13:26.800 out of data. So, um, every question you ask chat GPT becomes part of its data for training,

00:13:33.040 but if it could make its own data, if it could make its own unbiased data and, you know, good,

00:13:39.100 clean data, then it could improve its own intelligence. And that's what the intelligence

00:13:43.940 explosion. It really is. It's, it's, it's a idea that's been around since the 1960s and says,

00:13:49.040 if we create machines that are as smart or smarter than us, uh, they'll be able to, to do many things

00:13:57.240 that we do better than us. One of those things critically is artificial intelligence research

00:14:03.680 and development. So right now we're developing, uh, machines that are, that are going to be good

00:14:09.440 at artificial intelligence research and development. And that's, that's the recipe for the intelligence

00:14:14.860 explosion. Now, what happens after that, you know, do we come through? Okay. Because, you know,

00:14:19.700 one day we're playing with chat GPT or it's, you know, it's, it's much smarter cousin in five years.

00:14:27.100 And then the next thing we know, we're sharing the planet with something that's a thousand or a

00:14:31.560 million of times more intelligent than we are. And we don't know how to do that. We don't know how to,

00:14:38.100 uh, share the planet with something smarter than us. You know, I, I was lucky enough to interview

00:14:42.940 Arthur C. Clarke, the science fiction writer years ago. Uh, people don't know he was also a scientist,

00:14:48.040 right? Um, but he said, uh, you know, we share the planet or we steer the future,

00:14:53.260 not because we're the fastest creature or the strongest creature, but because we're the most

00:14:58.000 intelligent when we share the planet with something more intelligent than we are, they will steer the

00:15:03.000 future. So that's, you know, that's where we're headed. That's, that's where the, uh, end of the

00:15:08.860 human era comes from. We have a couple of really big problems to solve, or I'm afraid it's, you know, 0.63

00:15:15.540 it's going to be lights out for, for our species. You know, I hadn't heard, uh, Arthur C. Clarke's

00:15:22.860 name, uh, for a very long time. I had the opportunity to spend some time with him, uh, years, uh, years ago

00:15:29.580 when, whenever he came to town to talk about, uh, all, all of his splendid writings, uh, beautiful

00:15:37.860 writings, including of course, uh, space odyssey and the computer to think how prescient he was in so

00:15:47.740 many areas, uh, is, is humbling. But, uh, anyway, I, I, a great, a great man, a great, uh, author, uh,

00:15:55.920 and he absolutely, he absolutely, uh, thought, and one of the other things we, we, when I interviewed him,

00:16:01.560 he said, intelligence will win out in whatever form. Intelligence is the superpower of, of our

00:16:07.780 planet. And, uh, if it's, if it's machine or human, the greater intelligence will win out.

00:16:13.380 Well, let's, let's continue this conversation, uh, with, with James Barrett. Uh, he is the author

00:16:19.920 of our final invention, artificial intelligence and the end of the human era. And we're going to

00:16:25.720 find out more about the explosion of both intelligence and knowledge. And is there a

00:16:31.420 countervailing influence available or imaginable, uh, to AI and what looks to be an almost certain

00:16:41.280 path to its dominion over our planet? We're coming right back with James Barrett. We're back. We're

00:16:47.380 talking with author James Barrett and his book is our final invention, artificial intelligence and the

00:16:53.380 end of the human era. Uh, welcome back. And James, let's talk about that, uh, intelligence explosion,

00:17:00.820 uh, the ever, the ever, uh, growing need on the part of AI to find more data, a greater,

00:17:08.620 ever greater knowledge base. Is there any countervailing, any possibility of a countervailing

00:17:15.460 influence against what will be the awesome intelligence and capacity of AI in whatever

00:17:23.140 form it takes, let's say over the next 10 to 20 years? Sure. Fortunately, there is, there is hope.

00:17:29.520 And I'm actually writing a book proposal for a, for a follow-up to our final invention

00:17:34.060 called the intelligence explosion. Um, what, what, what has to happen is these companies have to,

00:17:41.440 you know, there was recently the future of life Institute released a open letter saying, let's,

00:17:46.500 let's pause development for, for six months. Another, uh, prominent AI theorist,

00:17:51.620 Eliezer Yudkowsky said, no, let's just stop it until we understand, because here's the thing,

00:17:57.060 even the experts say, we don't understand what's going on inside these large language modules.

00:18:02.340 They're not programmed the way we used to program. They ingest information and they organize it

00:18:07.340 themselves. If you were to open it up and peer inside, you'd see a lot of decimals and, and random

00:18:13.060 marks. And it's really the way that it encodes the vast amounts of, of data it ingests, but, but it's not

00:18:20.580 programmed and you can't like insert some code to make it safe. So what, what's that, what that's

00:18:28.100 called is interpretability. We need, we need to be able to, to interpret and to explain what's going

00:18:34.260 on inside these machines. And, you know, the head of, uh, open AI and, uh, Stuart Russell, who's one of

00:18:40.740 the world's preeminent, uh, AI, AI thinkers. He wrote this code of the standard texts on, on artificial

00:18:47.300 intelligence called a modern approach. He says, you know, we, we, we don't understand what's going

00:18:52.740 on inside and we're headed for what he called a Chernobyl size disaster. Now we, there are ways around

00:19:01.220 that, uh, that conclusion. We can, we can, we can, uh, so a slow down and B insist that, that, uh, AI

00:19:12.900 makers are able to explain how the systems are working. And what that may mean is going back to

00:19:19.860 last generation AI techniques like expert systems, symbolic AI, case-based reasoning.

00:19:26.260 And these, these old techniques allow programmers to interpret and explain how the systems work,

00:19:33.300 making them more predictable, making them safer than models based on deep, deep learning.

00:19:39.940 Um, some experts advocate a hybrid approach of both deep learning and older techniques.

00:19:45.620 So you don't lose all the awesomeness of, of the large language modules models.

00:19:50.740 Um, that may be a good, that may offer a route to solving, you know, to, to solving mankind's

00:19:57.620 problems. I mean, there are many benefits to artificial intelligence, but there's also some,

00:20:03.220 some harrowing risks. And what we need to do is, as you said, we need to, we need to mitigate them.

00:20:08.900 And, and just knowing where to start with a, with a reliable mitigation, a workable mitigation is going to

00:20:17.220 be a fascinating enterprise all, uh, all into itself. Uh, we, you mentioned the pause that had been called for by

00:20:24.420 some what 1400, uh, uh, leaders in technology to, to moving from, uh, GPT four to five, uh, or wherever the

00:20:36.580 research is at this point, Bill Gates, I thought it was interesting. He, he absolutely rejected the idea of that pause

00:20:43.700 on AI says just, it's impractical and impossible. I don't often agree with Bill Gates about, uh,

00:20:52.020 a number of things, but the, the reality is I can't see, uh, the CCP and China and Beijing saying,

00:21:00.020 you know, yeah, that's a good idea. Let's pause. We don't want to have additional advantages in our

00:21:05.380 technology to, uh, to win over the world. Uh, your thoughts about the nationalism that persists,

00:21:13.700 the, uh, the idea that there's going to be a global governance, which will be rejected,

00:21:18.660 I think by most, uh, Americans just by nature, uh, and the efficacy of, of what would result from a pause

00:21:27.940 or from just plunging ahead? Well, you're right. Getting China to the table would be tough,

00:21:35.540 but there's a cultural thing about China that, that we need to keep in mind that the, the,

00:21:41.780 the programmers in China right now, there's a chip embargo. So they don't have the chips to make a chat

00:21:46.980 GPT. They don't have the Nvidia chips and, uh, they're starting their own multi-billion dollar chip

00:21:52.820 industry because right now they're dependent on us. So they won't be coming up with one of these,

00:21:57.180 these things soon. Secondly, there's a cultural thing where they don't want to be seen as to do

00:22:03.980 anything to, that would threaten the premier. So they, they will not make something that is more

00:22:10.460 powerful. I'm not kidding. They will, they will probably not make something willingly or wittingly

00:22:15.820 that is more powerful than, than, uh, than the premier of China. So it's not a, it's not a given

00:22:23.580 that they will eat our lunch as, as people are afraid. Um, we do, we, we, that we already seem

00:22:30.380 to have a lot of buy-in from the, the, the heads of, uh, of some of these, of the, of the big tech

00:22:36.460 companies. So, you know, six, I'm not sure if six months is going to be enough, but China is, 0.98

00:22:43.580 is they're, they're tech technologically extremely astute. They know what's happening and they,

00:22:50.300 they know what's coming now, boy, you know, maybe we can get them to the table because they know that

00:22:56.540 what's best for, for, uh, for, for China in AI is best for all of us. What's best for 0.95

00:23:04.220 slowing down the intelligence explosions, slowing down that this development is not just good for

00:23:09.980 us, but good for them. Um, and another way to look at is this, if China did win and were suddenly

00:23:17.580 ruled by, you know, the, the guy with the biggest AI, well, that wouldn't be, would that be any worse

00:23:23.980 than being ruled by, ruled by AI? You know, we, we'd be, we'd be, we'd be under, uh, a tyranny

00:23:32.060 one way or the other. So I'm trying to find, I'm trying to find in that James, some cold

00:23:38.780 comfort, at least, uh, but the choice you offered, uh, is not, you know, it doesn't warm my heart.

00:23:46.460 Uh, it's, and it's precisely what I think concerns most Americans. Uh, and that is we have a

00:23:53.900 unique, uh, uh, life, uh, in America, a lifestyle, a way of life, uh, and it is decentralized. And

00:24:03.980 here we are, we thought that, uh, the CPU in all those desktops would be decentralizing and

00:24:10.140 democratizing, uh, to all the population. It hasn't quite worked out that way, but it's better than it

00:24:16.940 was. And it does open up avenues for far more people. But once we go to AI, we are talking about,

00:24:24.700 if not the singularity, we're talking about high, high dense levels of centralization of both

00:24:32.300 capacity and power. Well, yeah, absolutely. I mean, look at it this way. Our, our lives are in the hands

00:24:40.620 of about five CEOs and they weren't elected and, you know, I didn't vote for them. And what gives

00:24:47.340 them this power? Here's the, there's a, there's a, there's a, there's a kind of a, a metaphor that

00:24:52.380 people use a lot. And I like it. It's about getting on an airplane. Let's, let's say, uh,

00:24:57.820 you're about to get on an airplane when you're told half of the engineers who designed it, believe

00:25:02.460 it has a 10% chance of crashing. Furthermore, the airplane fails the explainability test.

00:25:08.780 It's a black box system and none of the engineers can explain how it works.

00:25:13.260 Then there's the problem of control. The longer the plane stays in the air, the more prone it is

00:25:18.300 to unpredictable, uncontrollable behavior. And the engineers themselves, each has a sorted

00:25:24.860 rap sheet of rotten behavior, like lying to Congress, employing predatory business practices,

00:25:30.460 infringing on copyright, copyrights and publishing incendiary fake news. Now, would you get on that airplane?

00:25:38.940 Well, since you put it in such an inviting, uh, inviting terms, I'm, I'm thinking probably

00:25:43.900 I'll catch the next flight. Yeah. Well, here's the, the bad news is you're on it already. And so is

00:25:49.500 everybody. And so is everybody you care about. We're on this plane. It's, it's, uh, you know, we're,

00:25:55.900 we're, we're rocketing ahead as fast at warp speed with these architectures. And we simply don't

00:26:05.180 understand how they work. Well, we're, we're talking with author James Barat, who has, uh,

00:26:11.020 offered us a thought experiment that, uh, I don't think it leads to a lot of places pleasant,

00:26:16.300 but we'll find out more and what we can do to keep this, uh, this flight, uh, airborne. Uh,

00:26:23.340 when we continue with a brief message from our sponsors, right after this, we're coming right

00:26:27.020 back. Stay with us. We're back with author James Barat and documentary, uh, uh, Tarion as well.

00:26:35.100 Uh, his book is our final invention, artificial intelligence and the end of the human era.

00:26:40.860 Uh, you've put together a, a, a, a, a challenging, uh, experiment. Uh, we're aloft. Uh, we have all of

00:26:49.900 the conditions that you laid out. Now, what do we do, James? Well, we, we, the, I, I, I'm the,

00:26:58.940 usually the last to recommend regulation. Um, but you know, right now there's a bill in front of

00:27:06.700 Congress called the algorithm accountability act of 2022, and that requires, uh, impact assessments

00:27:13.820 of AI systems to check for biases and effectiveness and presumably safety. These regulations are not

00:27:20.060 moving very fast. Um, but, but, you know, we need some oversight. These, you know, even, even, uh,

00:27:28.860 even, uh, the heads of some of these countries say, or companies say we need oversight. I heard an

00:27:34.220 amazing thing from the CEO of Google who said, uh, yeah, Sundar Pichai said, society isn't prepared

00:27:41.900 for the rapid advancement of AI, but establishing common sense precautions is not for a company to

00:27:47.980 decide. So he completely, completely said, it's not our responsibility. Now, when, when, when we release

00:27:56.300 a new food, a new drug, we have the FDA that gives it a thorough checkout. When we,

00:28:02.780 when we introduce a new airplane, the FAA has to approve it. Um, the IAEA, the international

00:28:08.060 atomic energy agency has to approve any kind of, uh, development in, in, in nuclear energy or nuclear

00:28:14.700 weapons, but we have nothing. It's bizarre given the stakes that the industry has been left to self

00:28:20.620 regulate. Um, you know, there's an issue. There's, there's an issue in that too, that, uh,

00:28:28.860 uh, everyone's talking gently about our talking around it. And that is how many minds do you think,

00:28:37.180 uh, are extant in American society right now who can fully comprehend and engage the first,

00:28:46.060 the concept of AI and who truly understand it, uh, uh, scientifically, uh, and operationally.

00:28:54.620 Oh, you know, I, I, I, I, there are so few, you know, I would say that you could, you could fit them

00:29:01.260 in a phone booth. My problem with that is James, I'm a populist. I really believe in the consent

00:29:08.460 of the governed. I believe in our constitutional Republic. And I don't believe that there is a

00:29:13.900 existing at this moment, another government system that comes anywhere close to assuring

00:29:21.260 individual rights as does the American, the American governmental system, uh, and political

00:29:28.380 system is as flawed and messed up as it is right now. I still think it's, uh, the done correctly.

00:29:35.260 It's the best there is. I don't know a way, uh, to preserve it. If we have so few who can comprehend

00:29:41.740 and, and navigate, uh, the future. That's, that's a really good point. I'm, I'm with you. I'm a populist

00:29:48.540 too. And that's why I wrote our final invention to be read by anybody. Um, I tell you, if you,

00:29:55.180 if you stroll through a bookstore and you pick up an AI book, it's generally hard going. And I really

00:30:01.180 try to make our final invention, uh, not, not, you know, like I really targeted like ninth or 10th grade

00:30:09.180 reading level, which happens to be where the New York times is and a lot of, a lot of other, uh,

00:30:15.180 papers. I'm doing the same thing with, um, the, the, the, the followup to our final invention,

00:30:21.420 because here's the thing you're asking about how we, how are we going to change things? How are we

00:30:25.100 going to save ourselves? And I think it comes down to this. We have to save ourselves. That means we

00:30:29.740 have to put pressure on politicians. We have to do it, you know, vote with our vote. We have to, um,

00:30:35.900 look at politicians. You know, the white house has a AI bill of rights. It's non-binding guidance on the

00:30:41.980 use of AI systems. It doesn't have any power. There's no money behind it. Just like there's

00:30:46.620 no money behind the algorithm accountability act, but we need to insist on that. You know,

00:30:52.860 we need to say, Hey, we're the voters. We don't want the, the power, uh, the, the, the fate of our

00:31:00.620 lives to be in the hands of five, you know, CEOs, all of whom have getaway houses in the desert. And

00:31:07.180 I'm not even kidding. You know, everybody, you, everybody who gets on TV, who's a CEO of one of

00:31:12.620 these companies has a place to escape to or try to escape to if, if the stuff hits the fan.

00:31:19.740 Yeah. It's is every, and every American should be so blessed. Uh, nothing is more concerning to

00:31:28.300 me right this moment than to hear the department of Homeland security secretary, Alejandro Mayorkas

00:31:36.220 come back again with wanting to control freedom of expression, freedom of thought,

00:31:43.660 and to do so by creating yet again, another task force. And the purpose of that task force, James,

00:31:50.060 how to use artificial intelligence to protect critical infrastructure, including screening cargo.

00:31:57.580 And they doesn't mention people, but it would be people as well

00:32:02.060 to ferret out any threat to quote unquote, the system. And suddenly, uh, DHS, which should,

00:32:10.700 in my opinion, be well down the number of agencies and departments that should be working with AI

00:32:16.380 AI wants to have AI chat GPT in several months to, so they can test how to use the technology to

00:32:26.140 protect the homeland as they would put it. I always hear them do that with an accent when I think of

00:32:31.820 them describing it as such, uh, that's deeply concerning to me because they haven't, they don't

00:32:37.900 have officials in the department of Homeland security who are capable of even the most rudimentary cyber, uh,

00:32:44.620 protection as they've proved in election after election.

00:32:48.700 I, I totally agree with you. It's, it's a giant mistake. You know, they did, they, they released

00:32:53.820 chat GPT onto the web, which is a big mistake as far as I'm concerned, because as chat GP develops,

00:33:00.460 and they're also, they're also trying to imbue it with initiative. So it's not inert. It's not just

00:33:06.060 sitting there waiting for a prompt. It's actually taking action and getting things done. Well, you don't

00:33:11.420 want that. You don't, you don't want to release it onto the internet. Um, and it was somebody that

00:33:15.900 wasn't thinking very hard about, about, about the risks who did that. Um, you don't want, I, you don't

00:33:23.740 want to put it in the hands of, uh, of Homeland security to guard anything because right now it's so

00:33:29.580 full of holes. It's just a giant security breach waiting to happen. Um, you know, it would, it would take

00:33:36.860 a bureaucrat to come up with that idea. Exactly. Exactly. Uh, James, this has been a delightful,

00:33:45.260 uh, conversation with you. I appreciate it so much. You're taking the time to educate, uh, educate us

00:33:51.020 and to illuminate AI. Uh, we always give our guests the last word here. Uh, and, and if we may, your

00:33:58.780 concluding thoughts, uh, here on the great. And the pleasure has been mine. Thank you very much for

00:34:03.820 having me on. Um, I would say to everybody, these concepts, while it is hard to, to, to program this

00:34:11.420 kind of AI, uh, and it's not up to everybody, but it is up to everybody to get involved, to start to

00:34:17.420 understand the basics of what it is about this era. We're going through a period that's as transformative

00:34:23.420 as electricity or the internet. And it's up to each individual in this Republic to, to get educated so

00:34:31.180 that they can, they can force their, and by, by peaceful voting, they can force their politicians

00:34:37.180 to take action to regulate. I just want to say again, what a pleasure. It's been talking with

00:34:42.780 you and thank you for being with us. And I hope you'll come back soon. And, uh, we want to recommend

00:34:48.300 James book, our final invention, artificial intelligence and the end of the human era.

00:34:55.020 We recommend it to you highly available on amazon.com straight away, uh, and fire up your

00:35:02.380 Kindle and let's go get them. Thank you. Thanks for being with us here. And please join us Monday

00:35:09.580 when our guest will be Ed Rollins, the savant of political strategy and the New York post,

00:35:16.380 Michael Goodwin. Join us for all that. Have a great weekend and we'll be right back Monday.

00:35:22.220 Until then, thanks. God bless you. And may God bless America.

The Great America Show - May 05, 2023

INTELLIGENCE WILL WIN-HUMAN OR MACHINE？

Episode Stats

Summary

Transcript