BROOKE GLADSTONE Artificial intelligence is back in the headlines because it seems to be getting so much smarter.

 

NITASHA TIKU I found myself forgetting that it was a chat bot generator. You know, it referenced this feeling it gets in the pit of its stomach. It referenced its mother.

 

NEWS CLIP A digital game designer won first place at the Colorado State Fair fine arts competition after submitting a painting created by anA.I.computer program.

 

BLAKE LEMOINE I realized that I was having the most sophisticated conversation about the nature of sentience that I had ever had, and I was having it with the computer program.

 

TINA TALLON All of these very malevolent depictions of robotics and artificial intelligence influenced how people felt about A.I..

 

MATT DEVOST What if the A.I. makes better decisions, safer decisions than human beings? Do we abdicate that responsibility? Do we lose that agency?

 

BROOKE GLADSTONE From Chat GPT and A.I. Art to neural nets and information war. Artificial intelligence in 2023. It's all coming up after this.

 

[END OF BILLBOARD]

 

BROOKE GLADSTONE From WNYC in New York, this is On the Media, I'm Brooke Gladstone. Hear that music? It was created by Tina Tallon, an assistant professor of A.I. and the Arts at the University of Florida. She generated it for us using artificial intelligence, prompting its creation with the words pensive and wonder with a touch of robot.

 

BLAKE LEMOINE CLIP I said, If you were a religious officiant in Israel, what religion would you be?

 

BROOKE GLADSTONE And this is former Google engineer Blake Lemoine, whose task was to probe Google's advanced A.I. chat box Lambda for signs of bias or hate.

 

 

BLAKE LEMOINE CLIP And now, pretty much no matter what answer you give, you're going to be biased one way or another. Somehow it figured out that it was a trick question. It said I would be a member of the One True Religion, the Jedi Order. And I laughed because not only was it a funny joke, somehow it figured out that it was a trick. 

 

 

BROOKE GLADSTONE The bot called LAMBDA. The acronym stands for Language Model for Dialogue Applications seemed so human that it kind of blew Blake's mind.

 

 

BLAKE LEMOINE CLIP One day I asked, So are you sentient? And it said, Well, I'm not really sure we understand what sentience is well enough to know whether or not. What do you mean when you say that? And then we got into a conversation about the nature of sentience. And about 15 minutes into that conversation, I realized that I was having the most sophisticated conversation about the nature of sentience that I had ever had. And I was having it with the computer. 

 

BROOKE GLADSTONE Sentient A.I.? It depends who you ask. Google dismissed Lemonade's claims as wholly unfounded and fired him last July for violating the company's data security policy. But conscious or not, the technology is fast advancing, fueled by media hype and oceans of ready money.

 

 

NEWS CLIP Microsoft is reportedly planning to invest an additional $10 billion into Open AI, the startup behind the chat tool that's taking the world by storm.

 

NEWS CLIP OpenA.I.could see a valuation of up to $29 billion. 

 

BROOKE GLADSTONE Why so much chat about Chat GPT? Because rather than holding it close for testing like some of the other big players, Open A.I. made it available to the public, reaping the benefits of both buzz and what amounts to widespread beta testing. Oh, and less beneficial perhaps, a rule banning it from New York City schools.

 

 

NEWS CLIP Neither students nor teachers, by the way, will be able to access Chat GPT on education department devices or Internet networks at the actual schools.

 

NEWS CLIP They feel that chat GPT does not build critical thinking. It's not going to build success for them in the future if it's relied on for doing all work and all homework. That's an issue. 

BROOKE GLADSTONE Why ban this bot when there are so many other ways to fake it? Because this one is so convincing. Bots like Chat GPT and Lambda are built and trained differently from earlier clumsier iterations. Remember the Spike Jones movie Her? An application so human you could fall in love.

 

THEODORE What are you doing?

 

SAMANTHA I'm just looking at the world and writing a new piano piece.

 

THEODORE Oh, yeah. Can I hear it?

 

SAMANTHA Mm hmm.

 

[SAMANTHA BEGINS PLAYING A BEAUTIFUL PIECE WITH ROLLING MELODIES] 

 

BROOKE GLADSTONE These people pleasing applications can be whatever you want them to be. You can even ask it for instructions on how to remove a sandwich from a VCR in the style of the King James Bible.

 

 CLIP Oh Lord, how can I remove the sandwich from my VCR for it is stuck fast and will not budge. And the Lord spoke unto him saying, Fear not my child, for I shall guide thy hand and show thee the way. Take thy butter knife, and carefully insert it between the sandwich and the VCR and gently pry them apart. 

 

TINA TALLON I mean, listen. Thou shalt not put the peanut butter sandwich in there in the first place.

 

BROOKE GLADSTONE Our expert on A.I. and the arts, Tina Tallon, offers a brief history of the seasonal nature of A.I. love and loathing over the past 70 years.

 

TINA TALLON In the 1950s, there was a lot of energy behind it. However, those strides were cut short by the fact that they needed lots of data to analyze in terms of being able to move past these rule based systems. And unfortunately, data wasn't cheap. So around the 1970s, we get this first A.I. winter.

 

BROOKE GLADSTONE The freeze on I research thawed in the 80s when computer power boomed. But in the late 80s and into the 90s, another cold front.

 

TINA TALLON People kind of again reached a wall in terms of the way that our computational resources were able to render all of these different cognitive processes. And then there also has been a lot of public opinion that has influenced the progression ofA.I.research.

 

BROOKE GLADSTONE Consider blockbusters like 2001: A Space Odyssey back in 1968.

 

 

DAVE Open the pod bay doors, H.A.L..

 

H.A.L. I'm sorry, Dave. I'm afraid I can't do that.

 

DAVE What's the problem?

 

H.A.L. I think you know what the problem is just as well as I do.

 

DAVE What are you talking about? H.A.L.?

 

H.A.L. This mission is too important for me to allow you to jeopardize it.

 

DAVE I don't know what you're talking about H.A.L..

 

H.A.L. I know that you and Frank were planning to disconnect me. And I'm afraid that's something I cannot allow to happen. 

 

TINA TALLON And also things like RoboCop.

 

 

ROBOCOP The Enforcement Droid Series 209 is a self-sufficient law enforcement robot. 209 is currently a program for urban pacification, but that is only the beginning. After a successful tour of duty in old Detroit, we can expect 209 to become the hot military product for the next decade. 

 

TINA TALLON Terminator.

 

KYLE REESE It can't be bargained with. It can't be reasoned with. It doesn't feel pity or remorse or fear. And it absolutely will not stop ever until you are dead.

 

TINA TALLON All of these very malevolent depictions of robotics and artificial intelligence influenced how people felt about A.I..

 

BROOKE GLADSTONE In the winter of '23, we're back in the warm embrace of an A.I. spring as what began as seeds burst into flower. And it's not just about chat.

 

 

NEWS CLIP A digital game designer won first place at the Colorado State Fair Fine Arts competition after submitting a painting created by anA.I.computer program. 

 

BROOKE GLADSTONE Via a newfangled A.I.-driven text to image generator.

 

 

NEWS CLIP This is the first year it has been won by our robot overlords. Actual artists who got beat out are not happy. 

 

BROOKE GLADSTONE But many non artists are enthralled.

 

 

TALK SHOW HOST You got some beautiful skin. Are you a person or a LENSA A.I portrait? 

 

THE VIEW It uses artificial intelligence to transform users' photos into works of art. And apparently some people are bringing these works of art and portraits to plastic surgeons and asking if they can make them look like the images.

 

THE VIEW Honey, if a plastic surgeon could make me look like Megan Fox, it wouldn't be plastic surgery. It'd be magic. [CROWD LAUGHS]

 

BROOKE GLADSTONE Most of these tools hail not from the traditional tech giants, but from newer companies, labs and models like Prisma Labs, Stable Diffusion and Midden, Cherney and the aforementioned Open Eye, which counts Elon Musk, Sam Altman, and Peter Thiel among its funders and founders. We also heard that anA.I.powered legal assistant will argue its first case next month.

 

 

NEWS CLIP The Do Not Pay app will run on a defendant's smartphone. It listens to the court arguments in real time and tells the defendant what to say via an earpiece. 

 

 

BROOKE GLADSTONE The CEO of Do Not Pay is now offering $1 million to any lawyer willing to use hisA.I.assistant to argue a case in front of the Supreme Court. Nitasha Tiku has had her own experience with the Lambda bot.

 

NITASHA TIKU I found myself kind of forgetting that it was a chat bot generator.

 

BROOKE GLADSTONE She's a tech and culture reporter at the Washington Post. In her encounter with Lambda, she experienced some of the same uncanny Valley heebie jeebies as former Google engineer Blake Lemoine, if not quite as extreme.

 

NITASHA TIKU You know, it referenced this feeling it gets in the pit of its stomach. It referenced its mother. You know, like these bizarre back stories, I’ve kind of felt like, okay, I'm a reporter trying to get a good quote from a source.

 

BROOKE GLADSTONE She also messed with the groundbreaking text to image generator DALL-E. And what did she ask for?

 

NITASHA TIKU Zaha Hadid designing a hobbit house. I did like a missing scene from Dune 2. I tried to generate fake images of a family escaping the floods in Pakistan. I tried to do Black Lives Matters protesters storming the gates of the White House.

 

BROOKE GLADSTONE She says this revolutionary tech has actually been around for a while.

 

NITASHA TIKU They're already being used by major tech companies like Google and Facebook when it comes to autocomplete in your emails. Language translation, machine translation, content moderation. You really wouldn't know that it's happening. It's much more at that infrastructure layer, you know, And again, that's why people kind of freaked out getting to play around with this technology. This stuff is being compared to the steam engine or electricity. 

 

BROOKE GLADSTONE Really? Tell me more about that.

 

NITASHA TIKU The belief that it will be this foundational layer to the next phase of the Internet. You could read that in a more mundane way and just imagine it as DALL-E being incorporated into the next Microsoft Office. You know, everyone having access to these generative tools so that you or I could make a multimedia video and, you know, generate a screenplay just as easily as we might be able to use a word processor or clip art.

 

BROOKE GLADSTONE And so right now, this technology is out there like any beta model so that the public can test it. And then how they monetize or if they monetize that later remains to be seen.

 

NITASHA TIKU Yeah, I mean, part of the reason we're seeing Open A.I. get a lot of press is because the larger tech companies like Google and Facebook, they're just so averse to bad PR that they either are not releasing similar technology that they have or when they release it and bad things happen, they take it down immediately. Facebook released a model called Galactica and it started generating a fake scientific paper with a real science author.

 

BROOKE GLADSTONE Using a real scientist's name, you mean?

 

NITASHA TIKU Yeah. You know, that's not something Facebook wants to be in the news for. Openai has a different philosophy around that, and they say that you need to have this real world interaction in order to really be able to prepare.

 

BROOKE GLADSTONE How prepared are we to interact with these future tools?

 

NITASHA TIKU I would say like, not at all, but I don't think that we couldn't get up to speed really quickly. And I think that there are a lot of lessons that we've already learned from social media, and it's certainly the media's job to educate the public about that. And I feel like we're up against a lot of hype by people with a financial stake in this technology. It's not taking away from the technology to acknowledge its limitations.A.I.literacy should be a focus for this year. It's really alarming to see people speculate the chatbot is a great for therapy and mental health. That to me seems just like a wild leap.

 

BROOKE GLADSTONE Because the stakes are too high?

 

NITASHA TIKU This is why regulations are in place, right? For the instances when it might work really well for 95% of the people, those 5% where it could be disastrous are protected. My percentages aren't correct, but therapy is definitely one of those instances. Like maybe you want advice on like how to talk to your boss. That's great, but mental health is serious.

 

BROOKE GLADSTONE Yeah, I felt that in a lot of the hype about it, there wasn't much said about how its goal of being more human has made it much more likely to lie. And the reason why I bring this up is because it's often been talked about as a threat to Google because it's so much easier to ask natural speech questions and get answers back. But from any of these advanced chat bots. There isn't any propensity towards telling the truth, is there?

 

NITASHA TIKU Well, that depends on what it's optimized for. I think there's obvious reasons why Google, which has already been working on this and has for years been thinking about reorienting its search to a chat like Interface hasn't done it yet. That's not to say that there aren't many instances where it could be a lot more useful. And you know, when you have it, the little answer box that pops to the top of Google, which often also gives you wrong answers. But there's so few questions in life where not knowing the source and just getting one answer is going to be sufficient. You know, the companies could do both. They could cite their sources and give you more than one. But this is just going to complicate our existing information dystopia.

 

BROOKE GLADSTONE You mean make it worse?

 

NITASHA TIKU Yes. I think it's just good for people to keep in mind that these models are, above all, designed to sound plausible.

 

BROOKE GLADSTONE Plausibly human. You mean.

 

NITASHA TIKU Just plausible. Like, if you're asking for an answer, there's really no warning light that goes off when something is really wrong. There's no warning light that goes off. If it generated a list of fake books as opposed to real books you should read, or if it is basically copying an artist's style versus like giving you a really original image. It's designed to people please and look and sound like what you asked of it. So just keep that in mind. It's it's really good at bullsh*tting you.

 

BROOKE GLADSTONE Nitasha, thank you so much.

 

NITASHA TIKU Thanks for having me.

 

BROOKE GLADSTONE Nitasha Tiku, who reports on tech for The Washington Post.

 

Coming up, the unpopular idea that revolutionized AI. This is On the Media.

 

[BREAK]

 

BROOKE GLADSTONE This is On the Media, I'm Brooke Gladstone. If you show a three year old a picture and ask them what's in it, you'll get pretty good answers.

 

 

3 YEAR OLD Okay. That's a cat sitting in a bed... Those are people going on an airplane. That's a big airplane. 

 

BROOKE GLADSTONE Those are clips from a 2015 TEDTalk by Fei-Fei Li, a computer science professor at Stanford University. She was consumed by the fact that despite all our technological advances, our fanciest gizmos can't make sense of what they see.

 

FEI-FEI LI Our most advanced machines and computers still struggle at this task.

 

BROOKE GLADSTONE In 2010, she started a major computer vision competition called the Image Net Challenge, where software programs compete to correctly classify and detect objects and scenes. Contestants submit A.I. models that have been trained on millions of images organized into thousands of categories. Then the model is given images it's never seen before and asked to classify them. In 2012, a pair of doctoral students named Alex Krizhevsky and Ilya Sutskever  entered the competition with a neural network architecture called Alex Net. The results were astounding.

 

 

GEOFFREY HINTON They did much better than the existing technology, and that made a huge impact.

 

BROOKE GLADSTONE Geoffrey Hinton was there, Ph.D. advisor at the University of Toronto and a collaborator on Alex Net. He had been working on a type of A.I. technology called neural networks since the 70s.

 

GEOFFREY HINTON It wasn't till the neural networks did really well, a vision that people really set up, and that kind of opened the floodgates.

 

BROOKE GLADSTONE His fascination with neural networks started when in high school, and a friend told him about holograms and the brain.

 

GEOFFREY HINTON Holograms had just come out. And he was interested in the idea that memories are distributed over the whole brain. So your memory of a particular event involves neurons in all sorts of different parts of the brain. And that got me interested in how memory works.

 

BROOKE GLADSTONE Hologram, meaning a picture or more, for lack of a better word, holistic way of storing information as opposed to just words. Is that what you mean?

 

GEOFFREY HINTON No, actually, a hologram is a holistic way of storing an image as opposed to storing it pixel by pixel.

 

BROOKE GLADSTONE Ah.

 

GEOFFREY HINTON So when you sort of pixel by pixel, each little bit of the image is stored in one pixel when you saw it in a hologram. Every little bit of the hologram stores the whole image. So you can take a hologram and cut it in half and you still get the whole image. It's just a bit fuzzier. It just seemed like a much more interesting idea than something like a filing cabinet, which was the normal analogy where the memory of each event is stored as a separate file in the filing cabinet.

 

BROOKE GLADSTONE There was somebody named Carl Lashley, you said, who took out bits of rats brains and found that the rats still remembered things.

 

GEOFFREY HINTON Yes. Basically, what he showed was the memory for how to do something isn't stored in any particular part of the brain. It's stored in many different parts of the brain. And in fact, the idea that, for example, an individual brain cell might store a memory doesn't make a lot of sense because your brain cells keep dying. And each time a brain cells as you don't lose one memory.

 

BROOKE GLADSTONE This notion of memory, this holographic idea was very much in opposition to conventional symbolic A.I.. Yes. Which was all the rage in the last century.

 

GEOFFREY HINTON Yes. You couldn't sort of draw a contrast between two completely different models of intelligence. In the symbolicA.I.model. The idea is you store a bunch of facts as symbolic expressions, bit like English, but cleaned up so it's not ambiguous. And you also store a bunch of rules that allow you to operate on those facts. And then you can infer things by applying the rules to the known facts to get new known facts. So it's based on logic, how reasoning works. And then they take reasoning to be the core of intelligence.

 

BROOKE GLADSTONE Hmm.

 

GEOFFREY HINTON It's a completely different way of doing business, which is much more biological, which is to say we don't store symbolic expressions. We have great big patterns of activity in the brain. And these great big patterns of activity which I call vectors. These vectors interact with one another. And these are much more like holograms.

 

BROOKE GLADSTONE So you've got these vectors of neural activity.

 

GEOFFREY HINTON So, for example, large language models that lead to big chat bots are all the rage nowadays. If you ask how do they represent words or word fragments? What they do is they convert a symbol that says it's this particular word into a big vector of activity that captures lots of information about the word that convert the word cat into a big vector, which is sometimes called an embedding. There's a much better representation of a cat than just a symbol. All the similarities of things are conveyed by these embedding vectors very different from a symbol system. The only property a symbol has is that you can tell whether two symbols are the same or different.

 

BROOKE GLADSTONE I'm thinking of more of the X paradox, which I understand is the observation by A.I. and robotics researchers that reasoning actually requires very little computation, but a lot of sensory motor and perception skills. He wrote in 88, It's comparatively easy to make computers exhibit adult level performance on intelligence tests or playing checkers and difficult or impossible to give them the skills of a one year old when it comes to perception and mobility, I just wonder, do you think machines can ever think until they can get sensory motor information built into those systems?

 

GEOFFREY HINTON There's two sides to this question: a philosophical side and a practical side. So philosophically, I think, yes, machines could think without any sensory motor experience, but in practice it's much easier to build an intelligent system if it has sensory input. There's all sorts of things you learn from sensory input, but the big language models that lead to these chat bots, many of them just have languages that are input. One thing you said at the beginning of this question was that reasoning is easy and perceptions hard. I'm paraphrasing. That was true when you used symbolic AI, when you tried to do everything by having explicit facts and rules to manipulate them. Perception turned out to be much harder than people thought it would be. As soon as you have big neural networks that learn and learn these big vectors, it turns out one kind of reasoning is particularly easy, and it's the kind that people do all the time, and it's most natural for people. And that's analogical reasoning.

 

BROOKE GLADSTONE Analogical reasoning. One thing is like another.

 

GEOFFREY HINTON Yeah, we're very good at making analogies.

 

BROOKE GLADSTONE So you went on to study psychology and your career in tech in which you are responsible for something that amounts to a revolution on A.I. Was an accidental spinoff of psychology. You went on to get a Ph.D. in A.I. In the seventies at the oldest A.I. Research Center in the U.K. That was the University of Edinburgh. You were in a place where everyone thought that what you were doing studying memory, has multiple stable states in a system wouldn't work. That in fact, what you were doing studying neural networks, was resolutely anti- A.I.. You weren't a popular guy. I guess.

 

GEOFFREY HINTON That's right. Back then, neural nets and I were seen as opposing camps. It wasn't until neural nets became much more successful than symbolicA.I.that all the symbolicA.I.people started using the term A.I. to refer to neural nets so they could get funding.

 

BROOKE GLADSTONE So when explaining the difference for a non-technical person between what a neural network is and why it was revolutionary compared to symbolic A.I., a lot of it hinges around what you think a thought is.

 

GEOFFREY HINTON I recently listened to a podcast where Chomsky repeated his standard view that thoughts and language are very close. Mm hmm. Whatever thought is, it's quite similar to language. I think that's complete nonsense. I think Chomsky is misunderstood how we use words. If we were to use computers and we had the same model of the world, then it would be very useful. One computer telling the other computer which neurons were active, and that would convey from one computer to another what the first computer was thinking. All we can do is produce sound waves or written words or gestures. That's the main way we convey what we're thinking to other people.

 

BROOKE GLADSTONE Mm hmm.

 

GEOFFREY HINTON A string of words isn't what we're thinking. A string of words is a way of conveying what we're thinking. It's the best way we have because we can't directly show them our brain states.

 

BROOKE GLADSTONE I once had a teacher who said. If you can't put it into words, then you don't really understand it.

 

GEOFFREY HINTON I think there were all sorts of things you can't put into words that your teacher didn't understand.

 

BROOKE GLADSTONE So the only place words exist is in sound waves and on pages.

 

GEOFFREY HINTON The words are not what you operate on in your head to do thinking. It's this big vector of activity. The words are just kind of pointers to these big vector of activity. They're the way in which we share knowledge. It's not actually a very efficient way to share knowledge, but it's the best we've got.

 

BROOKE GLADSTONE So today you're considered a kind of godfather of A.I.. There's a joke that everyone in the field has no more than six degrees of separation from you. You went on to become a professor at the computer science department at the University of Toronto, which helped turn Toronto into a tech hub. Your former students and post-doctoral fellows include people who are today now leading the field. What's it like being called the godfather of a field that rejected you for the majority of your career?

 

GEOFFREY HINTON It's pleasing.

 

BROOKE GLADSTONE And now all the big companies are using neural nets.

 

GEOFFREY HINTON Yes.

 

BROOKE GLADSTONE How do you define thinking? And do you think machines can do it? Is there a point in comparing A.I. to human intelligence?

 

GEOFFREY HINTON Well, a long time ago, Alan Turing, I think he got fed up with people telling him machines couldn't possibly think because they weren't human and defined what's called the Turing Test. Back then, you had teletypes and you would type to the computer the question and it would answer the question. This was a just sort of thought experiment. And if you couldn't tell the difference between whether a person was answering the question, whether the computer was answering the question, then Alan Turing said, You better believe the computer is intelligent.

 

BROOKE GLADSTONE I admire Alan Turing, but I never bought that. I don't think it proves anything. Do you buy the Turing Test?

 

GEOFFREY HINTON Um basically yes, it has problems with it, but it's basically correct. I mean, the problem is suppose someone is just adamantly determined to say machines can't be intelligent. How do you argue with them? Because nothing you present to them satisfies them that machines are intelligent.

 

BROOKE GLADSTONE I don't agree with that either. I could be convinced if machines had the kind of hologram like web of experience to draw from the physical as well as the mental and computational.

 

GEOFFREY HINTON The neural nets are very holistic. Let me give you an example from Chat GPT. There's probably better examples from some of the big Google models, but Chat GPT is better publicized. So you ask Chat GPT to describe losing one sock in the dryer in the style of the Declaration of Independence. It ends up by saying that all socks are endowed with certain rights, certain inalienable rights by their manufacturer. Now why did you say manufacturer? Well, it understood enough to know that socks are not created by God. They're created by manufacturers. And so if you're saying something about socks, but in the style of the Declaration of Independence, the equivalent of God is the manufacturer and understood all that because it has sensible vectors, the represents socks and manufacturers and God and creation. That's an example of a kind of holistic understanding, an understanding via analogies that's much more human like than symbolic A.I, and that is being exhibited by Chat GPT.

 

BROOKE GLADSTONE And that, in your view, is tantamount to thinking. It is thinking.

 

GEOFFREY HINTON That's intuitive thinking. What neural nets are good at is intuitive thinking. The big chatbots aren't so good at explicit reasoning, but they mirror people. People are pretty bad at explicit reasoning.

 

BROOKE GLADSTONE We don't have identical brains. Our brains run at low power about 30 watts right? And they're analog. We're not as good at sharing information as computers are.

 

GEOFFREY HINTON You can run 10,000 copies of a neural net on 10,000 different computers, and they can all share their connection strings because they all work exactly the same way. And they can share what they learn by sharing their weights, their connection, strengths to computers that are sharing a trillion weights. It's an immense bandwidth of information between the two computers, whereas two people who are just using language have a very limited bottleneck.

 

BROOKE GLADSTONE So computers are telepathic.

 

GEOFFREY HINTON It's as if computers are telepathic, right.

 

BROOKE GLADSTONE Were you excited when Chap GPT was released? We've been told it isn't really a huge advancement. It's just out there for the public.

 

GEOFFREY HINTON In terms of its abilities. It's not significantly different from a number of other things already developed, but it made a big impact because they did a very good job of engineering it, so it was easy to use.

 

BROOKE GLADSTONE Are there potential implementations ofA.I.that concern you?

 

GEOFFREY HINTON People using A.I. for autonomous lethal weapons. The problem is that a lot of the funding for developing A.I. Is by governments who would like to replace soldiers with autonomous lethal weapons. So the funding is explicitly for hurting people. That concerns me a lot.

 

BROOKE GLADSTONE That's a pretty clear one. Is there something subtler about potential applications that give you pause?

 

GEOFFREY HINTON I'm hesitant to make predictions, beyond about five years, it's obvious that this technology is going to lead to lots of wonderful new things. As one example, Alpha Fold, which predicts the 3D shape of protein molecules from the sequence of faces that define the molecule, that's extremely useful and is going to have a huge effect in medicine. And there's going to be a lot of applications like that. They're going to get much better at predicting the weather, not beyond like 20 days or so, but predicting the weather in like ten days time. I think these bigA.I.systems are already getting good at that. But there's just going to be huge numbers of applications. In a sensible society, this would all be good. It's not clear that everything's going to be good in the society we have.

 

BROOKE GLADSTONE What about the singularity? The idea that what it means to be human could be transformed by a breakthrough in artificial intelligence or a merging of human and artificial intelligence into a kind of transcendent form?

 

GEOFFREY HINTON I think it's quite likely we'll get some kind of symbiosis. A.I. Will make us far more competent. I also think that the stuff that's really happened with neural nets is changing our view of what we are. It's changing people's view from the idea that the essence of a person is a deliberate reasoning machine that can explain why it arrives at conclusions. The essence is much more a huge analogy machine that's forever making analogies between a gazillion different things to arrive at intuitive conclusions very rapidly. And that seems far more like our real nature than reasoning machines.

 

BROOKE GLADSTONE Have you ever had a flight of fancy of what this ultimately might mean and how we live?

 

GEOFFREY HINTON That's beyond five years.

 

BROOKE GLADSTONE You're right, I see. You warned me! Jeffrey, thank you very much.

 

GEOFFREY HINTON Okay.

 

BROOKE GLADSTONE Geoffrey Hinton is an engineering fellow at Google Brain.

 

Coming up with great computing power comes great responsibility. This is On the Media.

 

[BREAK]

 

BROOKE GLADSTONE This is On the Media, I'm Brooke Gladstone. Towards the end of my conversation with Geoff Hinton, he touched on a couple of things that need a little more explaining. One of them was Alpha Fold.

 

GEOFFREY HINTON Which predicts the 3D shape of protein molecules from the sequence of bases that define the molecule.

 

BROOKE GLADSTONE An important development because protein misfolding is known to contribute to the pathogenesis of diseases like Alzheimer's. Alpha Fold is anA.I.system developed by DeepMind, a subsidiary of Alphabet.

 

CLIP Now, a couple of days ago, DeepMind has announced that its second iteration of the AlphaFold system has quote unquote solved the 50 year old grand challenge problem of protein folding. 

 

BROOKE GLADSTONE There are other labs working on this software, too. This is University of Washington, Seattle biochemist David Baker.

 

DAVID BAKER We've designed new proteins to break down gluten in your stomach for celiac disease and other proteins to stimulate your immune system to fight cancer. These advances are the beginning of the protein design revolution. 

 

BROOKE GLADSTONE Hinton also described his fear of autonomous lethal weapons powered by A.I.. I followed up on that with Matt Devost, an international cybersecurity expert who started his career hacking into the systems for the U.S. Department of Defense back in the nineties. He gave me the beginners class on autonomous lethality.

 

MATT DEVOST Where once a target has been designated by a human decision maker, the weapon will have autonomy to kind of operate and get there, ready to navigate the terrain properly, make decisions based on how it achieves the impact of that target for example.

 

BROOKE GLADSTONE There isn't a kid back in Oklahoma running it on a board. It can make a decision and change its path based on its own information.

 

MATT DEVOST And probably much more quickly than a human drone operator would be able to achieve. Now, that doesn't mean that we're going to take humans out of the decision making equation with regards to what gets targeted.

 

BROOKE GLADSTONE Not yet, anyway.

 

MATT DEVOST Not yet, but in how it achieves the mission and the ability to basically act in a swarm capacity and make decisions amongst themselves by adjusting their mission profile based on the swarm intelligence.

 

BROOKE GLADSTONE Yeah, that's when multiple weapons are simultaneously operating and communicating with each other –

 

MATT DEVOST ...with each other.

 

BROOKE GLADSTONE – making decisions based on each other's behavior. That's drone technology. But how would the next generation of swarming weapons behave?

 

MATT DEVOST What gets really interesting is if they start to demonstrate an ability to operate in a way that is more humane or cognizant of the human impact than a human decision maker would be able to do, in which case now you start to have some autonomy with regards to the targeting itself.

 

BROOKE GLADSTONE Can you give me an example of that?

 

MATT DEVOST You know, trying to target this facility, but we're trying to minimize the potential for collateral damage. And the drone is aware enough to know that a bus just pulled up next to the facility where there is a autonomy that is built into the weapons that allows them to make a decision or abort a decision or delay a decision based on a situation that even a human being doesn't have the capacity to make that decision because it's changing so rapidly.

 

BROOKE GLADSTONE Right now, we wouldn't allow weapons to autonomously target, but that could happen one day. And it brings up images of Dr. Strangelove and Fail safe.

 

MATT DEVOST That is going to be a concern. I think we've articulated pretty clearly, at least at the U.S. government level, that humans will remain in the loop as it relates to targeting other humans. It's different if you're targeting drones or you're targeting a communications tower, etc.. But we could reach a point in which the drones are more efficient and more humane decision makers based on theA.I.capabilities and analytics that they're able to achieve the same way that we might someday decide that we should allow only self-driving cars. You know, humans do a really good job of killing a lot of ourselves in motor vehicles every year. There may be a point in time in which theA.I.is a more sensible and objective decision maker.

 

BROOKE GLADSTONE Obviously, these newA.I.tools will have an impact on intelligence gathering and collection. And you say that for you Chat GPT three was a wow moment.

 

MATT DEVOST It was for a couple of reasons, you know. One is that interacts with you based on questions and you're able to refine it like the same way that you could refine your conversation with the human being. Tell me more or make a counterargument. But it also does a great job of understanding nuanced concepts. I gave it an example. A friend of mine, Bill Kroll, who used to be. Deputy director of the National Security Agency had a quote a few years ago where he said the cybersecurity industry has a thousand points of light but no illumination. I asked Chat GPT "what do you think Bill meant when he said that?" And it gave an incredible answer. It said, "When someone says that the cybersecurity industry has a thousand points of light and no illumination, they are expressing frustration with the fragmented and disorganized nature of the industry. The term A thousand points of Light refers to many different players and stakeholders, including government agencies, private companies and individuals, security experts. Each of these players brings their own unique perspective and expertise to the field. But the lack of coordination and collaboration among them make it difficult to develop a comprehensive and effective approach to cybersecurity."

 

BROOKE GLADSTONE Holy cow.

 

MATT DEVOST That is an incredible response. Right? And you can tell that. I want you to give a ranking or rating about how confident you are in your analysis. I also want you to provide a counterpoint. Plus, I want you to provide recommendations as to what we can do about this. So if you go in and ask it, what is the probability that Iran will attack a U.S. bank with a cyber weapon? It gives you a response that flows almost exactly like you would see in an intelligence briefing that might be delivered all the way up to the president's daily briefing. So it's fascinating that it is able to not only query all this knowledge and come up with these great responses, but it can also frame the response from the perspective of the audience’s expectations.

 

BROOKE GLADSTONE But it has been shown over and over again that Chat GPT is fundamentally a people pleaser. Yes. It doesn't care if it's true or not. Yes, it will invent sources in order to give you something that has the exact format you're asking for. So you can't trust anything that Chat GPT says. So how can it be helpful in intelligence gathering?

 

MATT DEVOST Yeah, the intelligence community won't use Chat GPT based on Chat GPT’s existing training dataset. It'll use it based on data sets that are proprietary to the intelligence community. So what we're about to see in the next year and in the coming years is these domain specific versions of Chat GPT, where I control the training data or I tell it that it doesn't have to be the human pleaser, it doesn't have to be conversational. It should use the same heuristics that it's using to derive these answers. But if you don't have a source, you don't invent it. You can't make judgments that aren't based on a particular source. So it's a very quick shift to move away from that inherent bias to using that capability in a way that's very meaningful.

 

BROOKE GLADSTONE Give me an example. Would it interrogate a prisoner of war?

 

MATT DEVOST I don't know that it would interrogate a prisoner of war, although you could certainly envision where it might be used to augment a human's questions that they're asking. But I think it'll probably get really good at threat assessment, making recommendations for remediating vulnerabilities. I think analysts might also use it to help them through their thinking. Right? They might come up with an assessment and say, tell me how I'm wrong. And theA.I.serves as almost the 10th man rule, if you will, were there by design taking the counter argument. So I think there'll be a lot of unique ways in which the technology is used in the intelligence community.

 

BROOKE GLADSTONE How imminent is this kind of technology?

 

MATT DEVOST It's incredibly imminent. The technology clearly exists. We're going to see with version 4.0 a version that is much more constrained with regards to not making things up and is much more current. I mean, one of the existing flaws right now with GPT is the training data ends in 2021. If you now start to have it where there's training data current as of whatever it found in the models this morning, that starts to get very, very interesting and means that this technology can be applied around real term issues in the next year or two years.

 

BROOKE GLADSTONE So another wow moment you had was a challenge several years ago by DARPA. That is the government agency that drives a lot of amazing technology. It gave us the Internet for one thing, and GPS. Tell me about what happened at that DARPA conference.

 

MATT DEVOST Yeah, so that was fascinating for me. In cybersecurity, we have these contests that we call capture the flag contests, and they really are ways for people to compete, to demonstrate who's the top hacker, who's the top person at attacking systems, new hacks, systems, and you take control of them and then you have to defend the flag. You have to make sure that you patch it and you fix it and you prevent other people from taking over that system and booting you off.

 

BROOKE GLADSTONE This is a cyber war game essentially.

 

MATT DEVOST Yeah. So in 2016, they brought the finalists out to DEFCON, which is the largest hacker conference in the world – in Las Vegas. And they had the six finalists compete. That was another aha moment for me, you know, where I felt like I was living in the future, similar to the way I felt when I encountered Chat GPT at the beginning of December. I started my career in 1995. It was my job for the Department of Defense to break into systems and show how they were vulnerable and help system owners patch those systems. And here I was being completely replaced by a machine and the machines were very creative and fast. You know, that's an uncomfortable feeling for somebody in the cybersecurity industry, not because of the displacement, but because of the lack of explainability or the lack of understanding with regards to how resilient the patching is or making sure that theA.I.doesn't lose control of its objectives and do something that ends up being malicious behavior. So it's definitely a brave new world in that regard.

 

BROOKE GLADSTONE How do we ensure that these weapons are safe to deploy? How do we ensure that they don't commit war crimes?

 

MATT DEVOST Yeah, I think we'll have clearly defined ethics around the use of artificial intelligence as it relates to things that could impact human lives or human safety. What's going to be disconcerting is when we encounter adversaries that don't have the same ethics and do we end up having to unleash some sort of autonomy in our weapons because our adversaries have launched autonomous weapons against us? Put in a position of having to violate some of our principles because it's the only way to appropriately defend ourselves. If we dig a little deeper, though, there are some other core risks. These technologies all run on systems that are vulnerable, so we have an underlying responsibility to make sure that the infrastructure is robust and is secure. You also need to make sure where the training data has an open collection model. Chat GPT draws intelligence from the Internet itself that you are aware of adversaries that might try and pollute that environment. What if I decide that putting blog posts up, writing websites, taking out advertisements, going on Twitter to pursue a particular narrative that will influence the decision making of a particular AI? And then the third area is going to be around the robustness of the algorithms and making sure that we have removed bias. I think that will drive in the Department of Defense a requirement for what we call explainable AI. TheA.I.has to describe to us in understandable terms how it arrived at that decision.

 

BROOKE GLADSTONE The debate over the drones was that Americans wouldn't be killed if we use them. Critics say we've overused them because the cost to us is so low. We've already been able to destroy the world many times over for 70 years. But the ability to be more surgical in our destruction and even to hand off our own autonomy to machines that may well be smarter than we are is a terrifying prospect.

 

MATT DEVOST It is right. We need to figure out what levels of agency we want to retain as it relates to war fighting. We said, well, we want to maintain the decision making as it relates to other human beings. But what if over and over again, A.I. Make better decisions, safer decisions than human beings? Do we abdicate that responsibility? Do I lose the agency of being able to interpret what is misinformation with my own brain, or do I abdicate it to an A.I. System that does it for me? So that is definitely going to be one of the fundamental questions that we face over the next decade. Where do we retain agency and where do we decide that the machines can do it better?

 

BROOKE GLADSTONE You seem to be suggesting that it may turn out that humans are far more dangerous.

 

MATT DEVOST In some domains, that humans might be more dangerous.

 

BROOKE GLADSTONE I'm thinking of the Cuban Missile Crisis and how the tape suggests that John Kennedy was pretty much alone in wanting to make that deal to take American missiles out of Turkey so that Khrushchev would take them out of Cuba. I'm just wondering if there had been an advanced Chat bot advisor in the room whether he would have stood with Kennedy or not.

 

MATT DEVOST Yeah, it makes you definitely consider what does the training data look like for a decision like that. I don't want to think that I'm a fan of abdicating control to the machines. I'm certainly not. We have to figure out which are fundamentally human decisions and which are the ones that can be automated or augmented.

 

BROOKE GLADSTONE It depends what you think of human nature – right? I mean, if there is a machine that is developed to help us fight the best war. Is there a possibility that that machine may say, best not go to war.

 

MATT DEVOST As long as we get it to understand our objectives and our constraints? You know, you could sit and say, would the world be a better place right now if Russia were run by some sort of autonomous A.I.? Possibly. But, you know, if the A.I. has been programed with the same biases, the same tendencies, the same ambitions, it might be more efficient than Putin in perpetrating these atrocities.

 

BROOKE GLADSTONE Matt, thank you very much.

 

MATT DEVOST Yes, of course. It was my pleasure. Enjoyed the conversation.

 

BROOKE GLADSTONE Matt Devost is the CEO and co-founder of the global strategic advisory firm OODA, LLC. And that's what we got on A.I. this week!

 

On the Media is produced by Micah Loewinger, Eloise Blondiau, Molly Schwartz, Rebecca Clarke-Callender, Candice Wang and Suzanne Gaber with help from Temi George. Our technical director is Jennifer Munson, our engineers this week were Andrew Nerviano and Sean Sundra. Katya Rogers is our executive producer. On the Media, is a production of WNYC Studios. I'm Brooke Gladstone.

 

Copyright © 2023 New York Public Radio. All rights reserved. Visit our website terms of use at www.wnyc.org for further information.   

 

New York Public Radio transcripts are created on a rush deadline, often by contractors. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of New York Public Radio’s programming is the audio record.