CNET and CBS News Senior Producer Dan Patterson sat down with the Future Today Institute founder and quantitative futurist Amy Webb to discuss what artificial intelligence and data actually are, and how they work. The following is an edited transcript of the interview.
Dan Patterson: Okay, that tees up the conversation then about artificial intelligence and emerging technology. I say AI because that’s the thing that is on everybody’s tongue right now. Although you’ll also say that it’s quantum, and we’ll get there as well.
Amy Webb: Should we talk about what AI is?
Dan Patterson: Yeah, so help us understand what AI… When we say AI, that’s actually a cluster of different technologies, machine learning, natural language processing, and other technologies. So where is China today and where are they relative to the Western market? And can you help us better understand this notion that China only steals and they don’t innovate?
Amy Webb: Yeah. So let me first start by saying, the concept of artificial intelligence is not new. We’re just hearing a lot about it lately. And it almost sounds buzzword-y.
Dan Patterson: There was an AI winter and then an AI spring.
Amy Webb: Oh sure. And there was an AI summer, there are all seasons. You can trace the beginnings of artificial intelligence back literally thousands of years. And that’s because AI is very much tied to our brains and how we think. And philosophers and mathematicians and theologians have for centuries, been trying to figure out the relationship between mind and machine, how we think. It was in 1956, at a big meeting at Dartmouth, that the term artificial intelligence was first born. There was a workshop that happened that summer and a bunch of people got together. Cross discipline academics, researchers, who were trying to further the concept, the idea of this suite of technologies, which was not just a machine doing some cool stuff on our behalf. But really, and truly an entirely different and new era of computing. One in which our data is mined and refined, and others data similarly minded, refined for the purpose of making decisions about us, for us on our behalf. So that’s all AI is.
SEE: Artificial intelligence: Trends, obstacles, and potential wins (Tech Pro Research)
Dan Patterson: Okay, let’s pause there. Let’s define a term. What the hell is data?
Amy Webb: Yeah, that’s a good question. So the best way to think about data is all of the points of information that we are creating or generating either intentionally or unwittingly all the time.
Dan Patterson: Okay, fantastic definition sorry to interrupt. Let’s go back. So AI conference…
Amy Webb: So here’s the deal. Even Marvin Minsky and John McCarthy who coined that term AI knew, that the moment that a technology becomes indispensable and almost invisible, we no longer think of it as artificial intelligence, which means that these days, we’re all using AI time. It’s part of our everyday lives. So if you go on to your CBS app, and you’re downloading videos, part of the decision making authority that determined with the compression look like a lot of the technical pieces of getting that video from the studio to you, involved some algorithms. Some of your data, some of the network’s data-
In an automated way. That’s right. Here’s a classic example that also portends some challenges that we’re facing. So when I back my car into my garage, I listen to very loud music. We’re friends right? I listen to Soundgarden very loudly.
So here’s the thing. I don’t know if you should be reassured but here’s the issue. So I’m blaring music and feeling very happy about this. As I’m backing into my garage, my car now automatically turns the volume down. Oftentimes, right at the best parts of the song, which means that and the reason is because somebody somewhere decided that it would be best to have that automated feature available in the car. So that I don’t, I guess get into some car accident when I’m backing in because I’m distracted. It doesn’t take into account things like I haven’t been in a car accident when driving backwards to get into my garage. I haven’t been in a car accident in this car, there were no markers there that would suggest that I need to have the music turned down. But I no longer have the option to make that decision.
Dan Patterson: So the algorithm is de-optimized?
Amy Webb: Well, is it I mean, it’s de-optimized for me, it’s probably optimized for somebody who’s been in a lot car accidents, I guess is they’re backing into their garages. But that optimization is the problem, because when you are building out the AI ecosystem, when you’re building out applications, when you’re training machines, to learn to make decisions, what’s happening is-
Dan Patterson: Which requires data. Sorry, to interrupt I’m perpetually interrupting—
Amy Webb: So here’s the problem. Who are we designing for? Because you cannot accurately predict every single outcome for every single human, it is mathematically impossible. There’s too many variables, we don’t yet have the computers that can do this and also we’re capricious, right? So the best possible outcome is to optimize. Is to optimize that decision making. But who are we optimizing for? Most of the time, we’re optimizing for the people who built the systems. Who may or may not look like you, may or may not sound like you, and therein lies the problems.
Now you asked me a question about China a long time ago. And the difference between China’s innovation and some of what’s happening in the West. We must stop looking at China as a copy paste culture. That notion is outdated and it is dangerous. And the reason for that is, because if we continue to look only for copy paste signals, what we wind up missing is all of the innovation coming out of China in real time. We don’t even recognize it as innovation.
I would argue that China has built Road initiative, the pilot project that it has with 68 countries. The face to face diplomacy, not at the top levels of government only, but throughout the party business leaders. Lower level, lower ranking government people who are making constant trips to Africa, to Latin America to Southeast Asia, that is innovation. That some of the technology that we are starting to see, some of the workarounds for that hacking and breaches that we’ve started to see that are not in service of reproducing products that can be sold, but instead figuring out ways of leverage using our data, all of that is incredibly creative.
So we are in danger of losing ground to China, not just economically and politically, but in other areas of life, if we keep assuming that China is just stealing our ideas, and making lesser versions of them. That is not what’s going on anymore. And that hasn’t been the case in a long time and I would argue that is also not the case with AI. There are essentially nine big companies that control the future of artificial intelligence, because they are the ones that control the lion share of data. And you need data to train the systems and to get the systems to work. And then ultimately, to do things like prevent me from driving backwards into my garage while listening to loud music.
Dan Patterson: That is a fantastic circle. You almost took my pivot. Your book is called The Big Nine, and it is about nine major technology firms who are deeply invested, not just financially but intellectually in artificial intelligence. Some of those are familiar firms, Google, Amazon, some of those are Chinese firms, what are these companies And why are they so important?
Amy Webb: So there are nine big companies that I’ve actually been writing about and talking about for years. So in the United States, there are six Google, Amazon, Apple, Microsoft, IBM, and Facebook and in China they’re collectively known as the BAT. So this is Baidu, Alibaba, Tencent. It doesn’t mean that there aren’t other companies also working on AI. Salesforce are certainly doing some interesting things, so as Uber, I could make a long list. But for the most part, these are the companies that are developing the frameworks, they are developing the cloud services, they have the consumer applications that mine and refine our data. And they’re in very different situations.
So in the United States our six companies serve a few different masters. They serve the whims of Wall Street, and when DC happens to be paying attention for five minutes, they serve to some extent, the interests of people who are in Congress. Most of the time, that relationship is antagonistic more than anything else. But in China, it’s very different. So the BAT are independent companies, but they work in concert with the political leaders in Beijing. And let’s not forget there’re couples of things on the horizon in China in 1949, it’s the hundred year anniversary of the CCP. China has the largest social mobility, China has the largest number of people who are about to ascend to higher and higher socio economic positions at a scale that we have never seen before on the planet.
During a time in which it is pretty clear now that climate has changed, we’re going to probably run out of resources that we need. So there’s a lot of different things all happening at the same time. And President Xi Jinping, I think it was in earlier in 2018 and his party changed the regulations in China. There are no more term limits. So Xi Jinping is effectively president for life. So there’s a lot of change on the horizon, which is why when you think about AI, you can’t just brush it off and assume that the conversation is about killer robots and machines coming to take our jobs, right? AI is the next era of computing. It is not a single technology. And if you think of life as a gigantic, Venn diagram, there are a lot of intersecting vectors here that have to do with politics. And everyday life and privacy, many of geopolitics, geo economics, lots of different things.