WEBVTT - Artificial Intelligence Leading Watershed Moment

0:00:00.160 --> 0:00:04.200
<v Speaker 1>This is Bloomberg Business Week with Carol Messer and Bloomberg

0:00:04.280 --> 0:00:08.200
<v Speaker 1>Quick Takes Tim Stinovic from Bloomberg Radio. Well, our audience

0:00:08.240 --> 0:00:10.480
<v Speaker 1>is very familiar with our next guest. He was the

0:00:10.520 --> 0:00:13.920
<v Speaker 1>CEO of Google for a decade as well as executive chairman.

0:00:14.000 --> 0:00:17.239
<v Speaker 1>He's a philanthropist. He's also the co founder of Schmidt Futures,

0:00:17.239 --> 0:00:20.280
<v Speaker 1>and he's the he's the co author of a new book.

0:00:20.640 --> 0:00:24.079
<v Speaker 1>It's called The Age of AI and Our Human Nature.

0:00:24.239 --> 0:00:26.880
<v Speaker 1>Eric Schmidt joins us now on the phone from New

0:00:26.960 --> 0:00:30.159
<v Speaker 1>York City. Eric, how are you? Um, thank you and

0:00:30.240 --> 0:00:32.680
<v Speaker 1>I'm great. I'm so glad to see things going so

0:00:32.720 --> 0:00:36.280
<v Speaker 1>well in the market. Yeah, I mean it's it's wild

0:00:36.360 --> 0:00:39.239
<v Speaker 1>at six, the sixth day of record or sixth day

0:00:39.240 --> 0:00:40.920
<v Speaker 1>I should say higher in a row for the SMP

0:00:41.040 --> 0:00:43.760
<v Speaker 1>five hundred. Uh. Hey, let's talk about the book and

0:00:44.080 --> 0:00:46.159
<v Speaker 1>the Age of AI and Our Human Future. You co

0:00:46.200 --> 0:00:50.280
<v Speaker 1>wrote it with Henry Kissinger as well as Daniel Huttenlocker,

0:00:50.280 --> 0:00:52.720
<v Speaker 1>the inaugural dean of the M. I. T. Schwartzman College

0:00:52.720 --> 0:00:57.440
<v Speaker 1>of Computing. Explain how this book came about. Well, Dr

0:00:57.520 --> 0:01:01.560
<v Speaker 1>Kissinger when he was in college, which was before all

0:01:01.600 --> 0:01:04.640
<v Speaker 1>of us were born seventy five years ago wrote his

0:01:04.720 --> 0:01:09.039
<v Speaker 1>undergraduate thesis on Kant and content invented this whole theory

0:01:09.080 --> 0:01:12.640
<v Speaker 1>of whether objects were real or how did we perceive them?

0:01:12.680 --> 0:01:17.200
<v Speaker 1>And he had been interested in how virtual worlds would emerge,

0:01:17.319 --> 0:01:20.280
<v Speaker 1>although he couldn't articulate it the way we do it today.

0:01:20.400 --> 0:01:24.680
<v Speaker 1>And when he visited Google and then subsequently talked to

0:01:24.720 --> 0:01:27.759
<v Speaker 1>some of the AI folks, in particular a guy named Demostatibus,

0:01:28.360 --> 0:01:31.840
<v Speaker 1>he realized that we're busy building the world that he

0:01:31.959 --> 0:01:35.880
<v Speaker 1>thought about, but we're building it as technologists and without

0:01:35.920 --> 0:01:39.240
<v Speaker 1>the basis of any kind of philosophy. So he set

0:01:39.280 --> 0:01:41.720
<v Speaker 1>out with myself and Dan and a few others to

0:01:41.840 --> 0:01:45.520
<v Speaker 1>kind of think through how do we address these technologies.

0:01:45.520 --> 0:01:48.400
<v Speaker 1>And we concluded, and this is the headline, is that

0:01:48.520 --> 0:01:51.440
<v Speaker 1>this change in the next five to ten years is

0:01:51.480 --> 0:01:55.200
<v Speaker 1>the beginning of an ethical change in human history, similar

0:01:55.240 --> 0:01:57.920
<v Speaker 1>to the Age of Reformation, where the age of reason

0:01:58.000 --> 0:02:00.320
<v Speaker 1>excuse me, where we went from the age of faith

0:02:00.560 --> 0:02:03.360
<v Speaker 1>where people basically just believe that whatever God said was

0:02:03.400 --> 0:02:07.480
<v Speaker 1>correct versus the notion of having reason, and that the

0:02:07.480 --> 0:02:10.440
<v Speaker 1>world where we're going to have ourselves and then inanimate

0:02:10.680 --> 0:02:14.720
<v Speaker 1>non human intelligence is helping us is a very different

0:02:14.720 --> 0:02:18.040
<v Speaker 1>world than the one we're in today, and bridge the

0:02:18.080 --> 0:02:20.480
<v Speaker 1>gap between the world we're in today and the world

0:02:20.639 --> 0:02:23.600
<v Speaker 1>we're heading into where AI is a bigger part of it.

0:02:23.639 --> 0:02:26.200
<v Speaker 1>Because you know, when I think about a I I

0:02:26.240 --> 0:02:29.600
<v Speaker 1>struggled to think about how it will transform my day

0:02:29.639 --> 0:02:31.760
<v Speaker 1>to day life. So if we go down that path,

0:02:31.880 --> 0:02:34.920
<v Speaker 1>you know, five, ten, even more years into the future,

0:02:34.960 --> 0:02:37.240
<v Speaker 1>I mean, how do you see this shaping the world

0:02:37.240 --> 0:02:40.440
<v Speaker 1>around us? Well, most people when they think about are

0:02:40.520 --> 0:02:43.400
<v Speaker 1>they think of a killer robot. I wasn't going to

0:02:43.480 --> 0:02:46.839
<v Speaker 1>say that, but yeah, yeah, yeah, we're not talking about that.

0:02:47.240 --> 0:02:50.680
<v Speaker 1>What we're talking about is that you will have assistance. Today,

0:02:50.720 --> 0:02:53.200
<v Speaker 1>you have an assistant in the form of Google UM,

0:02:53.240 --> 0:02:55.760
<v Speaker 1>you have recommendation engines and so forth. Those are all

0:02:55.800 --> 0:02:58.560
<v Speaker 1>AI powered by the way that you translations and so forth.

0:02:58.840 --> 0:03:01.920
<v Speaker 1>That all works very well. What will happen in the

0:03:01.919 --> 0:03:03.560
<v Speaker 1>next five or ten years is you're going to have

0:03:03.680 --> 0:03:07.280
<v Speaker 1>conversational systems that are human life that you're can ask

0:03:07.360 --> 0:03:11.480
<v Speaker 1>questions and can make recommendations and more importantly, generate new

0:03:11.560 --> 0:03:15.359
<v Speaker 1>things UM. In our book, we talk about the h

0:03:16.760 --> 0:03:21.760
<v Speaker 1>the achievement of AlphaGo, where computers developed new game strategies

0:03:22.160 --> 0:03:24.000
<v Speaker 1>for a game that have been played by humans for

0:03:24.040 --> 0:03:26.760
<v Speaker 1>two thousand private or years, that had never been discovered

0:03:26.760 --> 0:03:30.040
<v Speaker 1>by humans. We highlight the discovery of a new antibiotic.

0:03:30.160 --> 0:03:34.560
<v Speaker 1>They could never have been done without using superhuman supercomputer

0:03:34.760 --> 0:03:39.040
<v Speaker 1>analysis with human collaboration to look through all the choices

0:03:39.080 --> 0:03:41.800
<v Speaker 1>to find a new antibiotic that saves that saves lives.

0:03:42.160 --> 0:03:44.839
<v Speaker 1>And we also talk about these new language models, which

0:03:44.840 --> 0:03:47.680
<v Speaker 1>are these immense systems where we build them and that

0:03:47.800 --> 0:03:50.120
<v Speaker 1>we don't know what they know and what they really think.

0:03:50.640 --> 0:03:54.720
<v Speaker 1>And it's called GPP three. These are the beginning platforms

0:03:54.800 --> 0:03:57.360
<v Speaker 1>of a transformation of our the world around us. So

0:03:57.440 --> 0:04:01.360
<v Speaker 1>the answer for you, specifically your report, Uh, you're you're

0:04:01.400 --> 0:04:04.960
<v Speaker 1>in the news, You're inundated with information. Everyone is going

0:04:05.000 --> 0:04:06.960
<v Speaker 1>to be pushing more and more stuff to you. You're

0:04:06.960 --> 0:04:10.120
<v Speaker 1>gonna be inundated. You're gonna need a digital assistant. It

0:04:10.160 --> 0:04:13.040
<v Speaker 1>will help you prioritize what you what you need to

0:04:13.080 --> 0:04:15.520
<v Speaker 1>do in your work as well as your whole life. Okay,

0:04:15.520 --> 0:04:19.359
<v Speaker 1>that sounds better than a killer robot. Absolutely, So to

0:04:19.440 --> 0:04:21.880
<v Speaker 1>that end, you write in the preface along with your

0:04:21.920 --> 0:04:23.640
<v Speaker 1>co author, is that the three of you differ to

0:04:23.640 --> 0:04:26.520
<v Speaker 1>the extent in which you're optimistic about AI. Where do

0:04:26.600 --> 0:04:29.560
<v Speaker 1>you fall in that? Of the three of us, I'm

0:04:29.600 --> 0:04:33.120
<v Speaker 1>the more most optimistic, I think because in my world

0:04:33.160 --> 0:04:36.919
<v Speaker 1>and my experience, we've we have these enormous opportunities with

0:04:37.040 --> 0:04:41.599
<v Speaker 1>enormous down side, and somehow we muddle through. Um. If

0:04:41.600 --> 0:04:45.200
<v Speaker 1>I told you the social media would become the primary

0:04:45.200 --> 0:04:47.800
<v Speaker 1>attention source and information source for the most of the world,

0:04:48.160 --> 0:04:50.880
<v Speaker 1>you'd say, great. What I didn't realize when we did

0:04:50.920 --> 0:04:55.400
<v Speaker 1>social media it would also be used to manipulated elections

0:04:55.400 --> 0:05:00.000
<v Speaker 1>and manipulate people through falsehood and misinformation and so forth. Um,

0:05:00.120 --> 0:05:03.479
<v Speaker 1>we're working on the ladder. So when this next generation

0:05:03.560 --> 0:05:07.039
<v Speaker 1>of much smarter systems come along, we better have a

0:05:07.120 --> 0:05:10.039
<v Speaker 1>notion of what their ethics are. What are they promoting

0:05:10.880 --> 0:05:12.760
<v Speaker 1>When you put one of these things with your kid

0:05:13.240 --> 0:05:16.680
<v Speaker 1>who's obviously developmentally vulnerable. What happened If the thing has

0:05:16.720 --> 0:05:19.240
<v Speaker 1>a hidden bug in it, like it's slightly racist, or

0:05:19.240 --> 0:05:22.919
<v Speaker 1>it has the wrong values. What's all that says something inappropriate?

0:05:23.440 --> 0:05:26.039
<v Speaker 1>Do you really want your child exposed to that in

0:05:26.080 --> 0:05:29.040
<v Speaker 1>such an important role where this is the child's best

0:05:29.080 --> 0:05:32.440
<v Speaker 1>friend because they rely on them. This is an experiment

0:05:32.480 --> 0:05:35.440
<v Speaker 1>in human existence that we have are now running that

0:05:35.480 --> 0:05:38.440
<v Speaker 1>we've never run before. We've never had a technology that

0:05:38.600 --> 0:05:44.719
<v Speaker 1>was both imprecise, dynamic emergence and learning. It's learning while

0:05:44.800 --> 0:05:47.800
<v Speaker 1>your kid is using it. Well, let's get right back

0:05:47.800 --> 0:05:50.920
<v Speaker 1>to Eric Schmidt, the former Google CEO and chairman, also

0:05:51.000 --> 0:05:53.880
<v Speaker 1>the co founder of Schmidt Futures. He's also the co author,

0:05:53.920 --> 0:05:56.680
<v Speaker 1>along with Dr Henry Kissinger and Daniel Hutton Locker, of

0:05:56.720 --> 0:06:00.839
<v Speaker 1>the Age of AI and Our Human Future. Eric, I

0:06:00.920 --> 0:06:03.039
<v Speaker 1>want to talk about the metaverse and the context of

0:06:03.160 --> 0:06:06.520
<v Speaker 1>last week's rebranding of Facebook. Now Facebook is known as

0:06:06.560 --> 0:06:09.600
<v Speaker 1>Meta Platforms, Inc. Mark Zuckerberg said last week it's now

0:06:09.600 --> 0:06:13.279
<v Speaker 1>a metaverse first company. Indeed, if you open up Instagram now,

0:06:13.279 --> 0:06:15.800
<v Speaker 1>it says instead of saying Instagram by Facebook, it says

0:06:16.160 --> 0:06:19.839
<v Speaker 1>Instagram by Meta. What is the relationship between AI and

0:06:19.880 --> 0:06:24.520
<v Speaker 1>the metaverse? Well, the metaverse needs to be defined, you know.

0:06:24.560 --> 0:06:28.839
<v Speaker 1>As you know, the term was coined by Neil Stevenson

0:06:28.880 --> 0:06:32.360
<v Speaker 1>in a famous science fiction book. And most of us

0:06:32.400 --> 0:06:35.360
<v Speaker 1>think that the metaverse is a world where sort of

0:06:35.440 --> 0:06:38.479
<v Speaker 1>similar to the movie Ready Player, one where you would

0:06:38.520 --> 0:06:42.960
<v Speaker 1>move into a digital world that's everyone's smarter, more beautiful, faster,

0:06:43.160 --> 0:06:47.120
<v Speaker 1>it's more fun and it's really enjoyable. Facebook has not

0:06:47.200 --> 0:06:50.640
<v Speaker 1>built that product yet and so we have to assume

0:06:50.680 --> 0:06:53.120
<v Speaker 1>that that's their direction. But it seems to me that

0:06:53.240 --> 0:06:56.279
<v Speaker 1>naming the changing the name of all of Facebook, which

0:06:56.279 --> 0:06:59.320
<v Speaker 1>is a hundred thousand people in an enormous business based

0:06:59.360 --> 0:07:02.800
<v Speaker 1>on products have not been announced yet, is a pretty

0:07:02.839 --> 0:07:05.760
<v Speaker 1>big decision. And we'll see, let's see how the products

0:07:05.760 --> 0:07:08.680
<v Speaker 1>play out. Yeah, I can't wait to see how that

0:07:08.720 --> 0:07:11.520
<v Speaker 1>plays out at Facebook now. Meta and I love the

0:07:11.560 --> 0:07:13.840
<v Speaker 1>movie Ready Player one. I don't know if I would

0:07:13.840 --> 0:07:17.240
<v Speaker 1>want to live there though, but it's interesting following this

0:07:17.440 --> 0:07:22.120
<v Speaker 1>announcement from Meta Facebook. I mean, you've seen Microsoft detailed

0:07:22.120 --> 0:07:25.400
<v Speaker 1>their plans for the metaverse or their ambitions there. Uh

0:07:25.480 --> 0:07:28.840
<v Speaker 1>Nividia Sword today after Wells Fargo said that there's an

0:07:28.880 --> 0:07:33.000
<v Speaker 1>opportunity for that company in the metaverse. Is this a

0:07:33.080 --> 0:07:35.320
<v Speaker 1>new buzzword? I mean, are we going to see every

0:07:35.320 --> 0:07:38.760
<v Speaker 1>company trying to jump on this bandwagon? Well, I think

0:07:38.800 --> 0:07:41.320
<v Speaker 1>I would rather have them go back to Hype AI

0:07:41.440 --> 0:07:47.080
<v Speaker 1>because AI is to transform their businesses in an enormous way. Um,

0:07:47.120 --> 0:07:49.840
<v Speaker 1>what I think of as a meta versus something very

0:07:49.880 --> 0:07:53.440
<v Speaker 1>interesting and very powerful. It's a natural next step. Um.

0:07:53.480 --> 0:07:56.120
<v Speaker 1>I've been involved in various projects associated it for more

0:07:56.160 --> 0:07:59.360
<v Speaker 1>than thirty years. Perhaps it's time has come. UM. I

0:07:59.480 --> 0:08:02.520
<v Speaker 1>was in the media last night where one person said

0:08:02.560 --> 0:08:05.280
<v Speaker 1>that that what will happen is Facebook will use its

0:08:05.400 --> 0:08:09.160
<v Speaker 1>significant cash and engineering resources to build this world using

0:08:09.160 --> 0:08:11.600
<v Speaker 1>its identity services and so for and so on, but

0:08:11.680 --> 0:08:14.440
<v Speaker 1>that there will be a competitor set of metaverses that

0:08:14.520 --> 0:08:17.280
<v Speaker 1>will be based on technology known as Web three, which

0:08:17.320 --> 0:08:21.880
<v Speaker 1>is the crypto platform, which has different different economics, different

0:08:21.880 --> 0:08:24.760
<v Speaker 1>power structures, different leaders. I think we don't know who

0:08:24.840 --> 0:08:27.200
<v Speaker 1>the leaders are going to be. What's exciting about this

0:08:27.280 --> 0:08:31.360
<v Speaker 1>metaverse idea once again the marketing ahead of the reality,

0:08:31.600 --> 0:08:34.280
<v Speaker 1>is that it's another example of how the industry moves forward.

0:08:34.679 --> 0:08:36.560
<v Speaker 1>In our book, we spent an awful lot of time

0:08:36.600 --> 0:08:39.440
<v Speaker 1>talking about this new set of platforms that are being

0:08:39.480 --> 0:08:43.280
<v Speaker 1>built in AI and their impact on society. The metaphors.

0:08:43.400 --> 0:08:46.280
<v Speaker 1>If it is successful and if we understand ultimately understand

0:08:46.320 --> 0:08:50.160
<v Speaker 1>what it is, UM may have huge impact on our

0:08:50.280 --> 0:08:52.840
<v Speaker 1>education systems and the way we live and so forth.

0:08:53.040 --> 0:08:55.920
<v Speaker 1>It doesn't mean Facebook will be the leader, but probably

0:08:56.000 --> 0:08:58.240
<v Speaker 1>somebody will be the leader in that space. Well speak

0:08:58.320 --> 0:09:02.760
<v Speaker 1>of Facebook in the wake of Facebook's brand too meta platforms.

0:09:02.800 --> 0:09:05.120
<v Speaker 1>A lot of commentary is focus around the Facebook brand

0:09:05.120 --> 0:09:07.280
<v Speaker 1>and the idea that it's under scrutiny from Democrats and

0:09:07.320 --> 0:09:10.480
<v Speaker 1>Republicans over what it did in terms of censoring or

0:09:10.480 --> 0:09:13.400
<v Speaker 1>not censoring content are promoting hate speech on the platform.

0:09:13.600 --> 0:09:16.360
<v Speaker 1>Our audience is very familiar with the articles that came

0:09:16.360 --> 0:09:20.280
<v Speaker 1>out as part of the Facebook papers recently. Eric, I'm

0:09:20.280 --> 0:09:23.840
<v Speaker 1>curious because you're widely viewed as coming in as the

0:09:23.920 --> 0:09:26.360
<v Speaker 1>quote adult in the room when it came to Google.

0:09:27.000 --> 0:09:30.520
<v Speaker 1>I'm wondering if if Facebook needs an adult to come

0:09:30.559 --> 0:09:34.720
<v Speaker 1>in right now. Well, in fairness to Mark and his team,

0:09:34.720 --> 0:09:37.680
<v Speaker 1>they're now proper adults. Maybe they were, you know, young

0:09:37.679 --> 0:09:40.960
<v Speaker 1>when they were founding the company, but they've they've earned

0:09:40.960 --> 0:09:43.560
<v Speaker 1>the right to drive in the way they are. I

0:09:43.600 --> 0:09:47.280
<v Speaker 1>think we can debate their decisions. One way to understand

0:09:47.280 --> 0:09:50.520
<v Speaker 1>the problem in social media is you want to maximize revenue,

0:09:50.880 --> 0:09:52.959
<v Speaker 1>and by maximizing revenue, the best thing to do is

0:09:53.000 --> 0:09:56.800
<v Speaker 1>to maximize engagement. By maximizing engagement, the best way to

0:09:56.840 --> 0:09:59.800
<v Speaker 1>do that is with outrage. So guess what you get

0:09:59.840 --> 0:10:05.200
<v Speaker 1>more were outrage Because outrage drives drives attention. Attention drives revenue,

0:10:05.480 --> 0:10:08.120
<v Speaker 1>and the companies are seeking that, and then the AI

0:10:08.280 --> 0:10:11.959
<v Speaker 1>systems are optimized for that. You shouldn't be surprised we

0:10:12.040 --> 0:10:15.440
<v Speaker 1>got what we got. Um When I was CEO of Google,

0:10:15.800 --> 0:10:18.760
<v Speaker 1>we faced similar choices that we made a decision, which

0:10:18.840 --> 0:10:21.679
<v Speaker 1>was a collective decision that we would take some of

0:10:21.679 --> 0:10:24.240
<v Speaker 1>the games to revenue, but some of it to quality.

0:10:24.640 --> 0:10:28.640
<v Speaker 1>That's a different company's decision. If I were doing Facebook today,

0:10:28.720 --> 0:10:30.920
<v Speaker 1>I would have a hard decision in front of me,

0:10:31.440 --> 0:10:34.360
<v Speaker 1>which is I want to improve the quality of my

0:10:34.520 --> 0:10:39.840
<v Speaker 1>users experience and frankly, peddling misinformation and manipulating people is

0:10:39.880 --> 0:10:42.559
<v Speaker 1>not a high quality outcome. It got to figure out

0:10:42.559 --> 0:10:46.520
<v Speaker 1>a way to address that. They don't. They will eventually

0:10:46.559 --> 0:10:48.760
<v Speaker 1>face some form of regulations. Eric Schmith, thank you so

0:10:48.880 --> 0:10:51.240
<v Speaker 1>much for joining us on Bloomberg Business Week. His new book,

0:10:51.320 --> 0:10:54.079
<v Speaker 1>The Age of AI and Our Human Future, co written

0:10:54.080 --> 0:10:56.680
<v Speaker 1>with Henry Kissinger and Daniel Hutton Locker. Eric Schmidt the

0:10:56.679 --> 0:10:59.679
<v Speaker 1>co founder of Schmidt Futures and former Google CEO and

0:11:00.040 --> 0:11:00.360
<v Speaker 1>chairman