Loading summary
Transcript220 lines
- [00:00]
Sponsor Voice
On the media is brought to you by Progressive Insurance. Do you ever find yourself playing the budgeting game? Well, with the name your price tool from Progressive, you can find options that fit your budget and potentially lower your bills. Try it@progressive.com Progressive Casualty Insurance Company and affiliates price and coverage match limited by state law.
- [00:20]
Brooke Gladstone
Not available in all states. From WNYC in New York, this is on the Media. Bob Garfield is away this week. I'm Brooke Gladstone. And this week we ponder the current state of artificial intelligence and its potential to, you know, elevate or annihilate humanity. Why has Google gobbled up the greatest minds in the field of robotics? Why does the army use smartbot recruiters? Why are we leery of thinking machines but leap at the chance to snooze in the backseat while our cars drive themselves? It's a surprisingly old question, though the ancients seem seemed less conflicted. Take 11th century China, where mechanical humanoids chime the hours on an astronomical clock ever after. And across the world we craft automata to play music, write letters, and lately even plumb our souls. The first use of the word robot, derived from the Slavic word for work, dates from 1920, a play by the Czech writer Karel Czapek. It's called Rossum's Universal Robots or R U R. And it doesn't end well for us. Here's a scene from Yuri Rozovsky's radio play.
- [01:38]
Jerry Kaplan
It was a crime to make robots.
- [01:41]
Brooke Gladstone
No, Alquist, I don't regret that. Even today.
- [01:44]
Jerry Kaplan
Not even today.
- [01:45]
Brooke Gladstone
Not even today. The last day of civilization. Was it a crime to shatter the servitude of labor, the dreadful and humiliating labor that man had to undergo? Work was too hard, life was too hard. And to overcome that was not what.
- [02:01]
Jerry Kaplan
The two Rossums had in mind.
- [02:02]
Brooke Gladstone
It's what I had in mind. How well you succeeded.
- [02:06]
Jerry Kaplan
How well we all succeeded. For profit, for progress, we have destroyed mankind.
- [02:15]
Brooke Gladstone
That mortal conflict has since played out countless times in novels and on screens large and small. I mean, it's not always a conflict, but usually. JP Talaat, the author of a robot of the science fiction film, says that this template of ambivalence was set with the film Forbidden Planet, specifically in the introduction of Robbie the robot in 1956, a crucial year for smart machines. Robbie, in fact, from the start of that film, is always presented as a figure that might or might not be under total control. Robbie was a seven foot servant and protector with a vacuum tube for a head, a washing machine tub for a body, and articulated and legs Humans land on Robbie's world, which is devastated by technology. They decide to demolish the planet. But they can't leave without Robbie's technological know how. So they take him along. Tricky. He saves them. He could destroy humanity. But he doesn't. In fact, he has a great TV career, appearing on Hazel, the Many Loves of Dobie Gillis, the Addams Family, the Man from UNCLE and in his signature.
- [03:35]
Jerry Kaplan
Role, Danger Will Robinson.
- [03:37]
Brooke Gladstone
Danger on Lost in Space. Robbie was both iconic and a bit of an anomaly because JP Talat sees a trajectory in our view of robots from pre war fondness to post war dread. The bomb drove home our suicidal inventiveness and the Internet the intrusive power of great inexplicable forces in our lives. Now robots no longer looked like Maytags. They look as in the 1973 movie Westworld. We will soon be landing at the.
- [04:12]
Henrik Christensen
Ultimate resort.
- [04:17]
Brooke Gladstone
Where you can live out your every fantasy.
- [04:20]
Henrik Christensen
Each resort is maintained by reliable computer.
- [04:24]
Brooke Gladstone
Technology and peopled by lifelike robot men and women. Desire ends in satisfaction, and all in a controlled environment when that notion of control goes a bit haywire. And Yul Brynner, for example, in Westworld.
- [04:45]
Alex Goldman
Starts killing the other guests.
- [04:46]
Brooke Gladstone
Culturally, we've always been in a bit of a muddle. We love them, but do they love us back? Here's an early scene from the reboot of Battlestar Galactica. Deep down, you've always known there was something different about me.
- [04:59]
Henrik Christensen
It flatters your ego to believe that.
- [05:02]
Brooke Gladstone
Alone, among all the billions of people of the 12 colonies, you were chosen for my mission. Your mission? What mission? You knew I wanted access to the defense mainframe.
- [05:16]
Dave Moss
What exactly are you saying?
- [05:18]
Brooke Gladstone
Humanity's children are returning home.
- [05:23]
Sponsor Voice
Today.
- [05:30]
Brooke Gladstone
They seduce us every bit. As much as they scare us, we yearn not just for the leisure they promise, but the immortality they represent. Is that why we build them?
- [05:41]
Sponsor Voice
Yeah.
- [05:41]
Brooke Gladstone
It may be the natural question. Why is natural reproduction not good enough for us? Well, that's because we are. Not good enough for us. Right? We die, we can't help but wonder if there's some possible way of postponing it. Just much like the Replicants do in the movie Blade Runner. We want more life. I've seen things you people wouldn't believe.
- [06:10]
Henrik Christensen
Attack ships on fire off the shoulder of Orion.
- [06:17]
Jerry Kaplan
I watched sea beams glitter in the.
- [06:20]
Brooke Gladstone
Dark near the 10 hours of gate. All those moments will be lost in time.
- [06:37]
Alex Goldman
Like tears.
- [06:42]
Brooke Gladstone
In rain. Time to die. More than zombies, werewolves, more even than vampires. Robots, in their infinite variety reflect all our greatest hopes and our greatest fears. And unlike those mythical creatures, robots exist but not how we imagine. And they also pose a potential threat, but also not how we imagine. Up next, artificial intelligence in the real world. This is on the media. This is on the media. I'm Brooke Gladstone. Artificial intelligence is a contentious field. To the people who own it, the engineers, it's about creating systems that can adjust their behavior based on data gleaned from past experience or new information systems that can establish sub goals for themselves to meet the broader goals of their designers, machines that in their word, think. But philosophers and others object, saying that to call that thinking is to imply that the complex web of sensations and culture that guide our thoughts can be reduced to code and that our minds are merely computers made of meat. Engineers respond with a shrug, a yawn, and the observation that the proof is in the processing. Jerry Kaplan is an engineer, an entrepreneur and a fellow at the center for Legal Informatics at Stanford University. He says that the attitude of engineers hasn't changed since the day the term artificial intelligence was coined.
- [08:35]
Jerry Kaplan
It was held in the summer of 1956. It was a beautiful, lazy summer and it was a conference that was organized by John McCarthy, who named the field artificial Intelligence. At that conference there were three other organizers. One was Marvin Minsky, who at the time was at Harvard and is still kicking around at mit, as you might know, Claude Shannon, famous as the father of information theory. And the fourth guy was Nathaniel Rochester, who was the manager of information research at IBM at the time.
- [09:06]
Brooke Gladstone
What were they there to talk about?
- [09:08]
Jerry Kaplan
I can actually read a little bit of it. Their proposal said the study is to proceed on the basis of the conjecture that every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it. Then they went on to say, we think that a significant advance can be made in one or more of these problems if a carefully selected group of scientists work on it together for a summer.
- [09:35]
Brooke Gladstone
These men were not lacking in self confidence, were they?
- [09:39]
Jerry Kaplan
They were not. And after the conference they made a number of wild predictions and got a great deal of publicity, which ruffled feathers at IBM Research. The salesforce was visiting customers, they were reading about this and they were getting worried.
- [09:54]
Brooke Gladstone
What was IBM selling in 1956?
- [09:57]
Jerry Kaplan
Data processing equipment. Really the only company or the primary company that was selling equipment of that kind. But the problem was that the customers were getting nervous about this stuff because they were concerned that these machines were somehow going to take over their own jobs. IBM decided in order to avoid this problem, they would close down their AI group. And then they armed their sales force with a response to this concern. And they came up with this little cliche that was computers can only do what they are programmed to do. And that concept has set the tone, in my opinion, for the last 50 years of the way in which people tend to think of computers. And unfortunately, it's no longer true.
- [10:38]
Brooke Gladstone
You also sort of hinted that it might be a threat, but I guess you're talking about to people's jobs potentially, not to their actual lives or autonomy.
- [10:48]
Jerry Kaplan
Well, the impacts on the job market are going to be extreme. There's a study that estimated that 47% of the U.S. working population that their jobs will in the next five to 10 years, come under potential threat of being completely automated. The people who are building these systems are going to have unique advantage in terms of skimming off the increased economic value that they'll be providing to society. So it's going to have a significant impact and already is having a significant impact on income inequality.
- [11:22]
Brooke Gladstone
So the Alphas will get richer and richer and the lower castes will just shuffle along with no purpose and not much money.
- [11:31]
Jerry Kaplan
Both the upper and the lower castes. This is the peculiar thing about it. We're talking about lawyers and doctors as well as the people who dig ditches and lay pipe, really. We're going to find out that Karl Marx was right. The struggle between capital and labor is very real. But capital is going to win hands down.
- [11:51]
Brooke Gladstone
So what's the distinction between strong AI and weak AI?
- [11:55]
Jerry Kaplan
Strong AI is basically a concept that intelligence is some kind of magic pixie dust that you sprinkle on a machine and it suddenly becomes conscious and aware. And that what we're trying to do in AI is to duplicate this magical quality. Now, weak AI is just a disparaging name for what we really see today, which is an engineering approach to artificial intelligence. It's solving specific problems that might be driving or playing chess. So strong AI is really for the sci fi freaks and weak AI is really for the engineers.
- [12:29]
Brooke Gladstone
Let's talk about progress in weak AI. There's been a lot of it lately, right?
- [12:35]
Jerry Kaplan
Absolutely. What's driving a revolution in artificial intelligence right now are really three fundamental things. Dramatic improvements in computing power and memory and networking. Just by way of example, the Watson program that won on Jeopardy, that had four terabytes of memory available to it to store all of its knowledge. If you look on Amazon today, you can buy four terabytes for $150. So we're talking about a speed of improvement that is utterly astonishing. The second thing really is breakthroughs in the research in what is commonly called machine learning. These are systems that learn by experience and learn by example. Which leads me to the third point. What the Internet provides is easy access to enormous mountains of data that can be used to train these systems. So the availability of that data has transformed artificial intelligence.
- [13:29]
Brooke Gladstone
When we talk about training, we aren't really talking about just collecting more and more facts and information, right?
- [13:37]
Jerry Kaplan
It really means exactly the same thing. It means when you talk about, like, training your children. There's kind of objective written facts that one can study, and that's why they go to school. There's things that you tell them, informal knowledge about the way the world really works that's based on your own experience. That's commonly called culture, things that you pass on to your children. But the third is simply interacting with the world. So the way your autonomous car is going to learn about the way your driveway is, shape or whatever, is by actually going out there and using sensory input in real time to incorporate that information into its model of the world. So all of those, the sensors, the Internet, and access to the accumulated knowledge of mankind is going to make them incredibly smart.
- [14:26]
Brooke Gladstone
I want to go back to your distinction between the engineers who you approve of and the science fiction freaks like myself of whom you don't. Six years before the Dartmouth Conference, Alan Turing, the famous British mathematician, designed his Turing test, which went like this. A human who is designated as a judge has a written conversation with a machine and another human for five minutes afterwards. If the human judge can't distinguish between his human correspondent and the machine one, the machine passes the test.
- [15:03]
Jerry Kaplan
I think that the Turing test is really misunderstood. It wasn't really about some kind of bar exam for computers to be intelligent. If you actually go back and read what Turing had to say, he said, the original question, can machines think? I believe to be too meaningless to deserve discussion. Nevertheless, I believe that at the end of the century, the use of words and generally educated opinion will have altered so much that one will be able to speak of machines thinking without expecting to be contradicted. And he was right. I can say, oh, Siri's thinking about my question. And you're not confused about whether Siri is some kind of human like intelligence? It's simply the use of the word.
- [15:50]
Brooke Gladstone
But he believed the mind was essentially a meat computer, and so have many other people since then. Ray Kurzweil, the visionary that actually foresaw much of the Internet, believes that he can reconstruct his Father with enough data and that if he can't tell the difference, then there is no difference.
- [16:13]
Jerry Kaplan
Ray is a remarkable scientist, but my point of view, and I'm representing, I think, a certain class of people who believe this, I see very little evidence, none, that machines are going to become conscious. Let me quote a Dutch computer scientist who's quite famous, but most people have never heard of Edsger Dijkstra. He put this really well in one sentence. The question of whether machines can think is about as relevant as the question of whether submarines can swim. The point is they're going to do the same thing that you do that you think of as thinking today, but they're not going to do it the same way and it doesn't matter.
- [16:53]
Brooke Gladstone
Perhaps it's doing something that looks a little like what we do but isn't entirely like what we do. I mean, isn't thought about independent intention, not intention that is programmed in by somebody else. Or do you also think that that's just irrelevant, that individuality is one of those magical notions?
- [17:17]
Jerry Kaplan
The notion of you as a unique individual will also apply to these machines because they are going to have their own experiences and they're going to draw their own conclusions and their own lessons from it. The real problem comes in if people are going to be attributing to these machines feelings and characteristics which in my view, legitimately they do not have, and those machines will be in a position to effectively manipulate us, acting in a way that evokes emotional reactions in us to providing sympathy for them.
- [17:49]
Brooke Gladstone
What danger does it pose?
- [17:51]
Jerry Kaplan
Well, it's that we include them in the circle of humanity and of mankind in a way where we sublimate our needs to the needs of these machines. And they will want to accomplish certain things. And if the best way to do that is to nag you or to cry at you or to put on a sad face, they might very well do that. They're going to be able to communicate with other machines, negotiate deals, offer arrangements, do things on your behalf. And that's going to create a variety of problems for society. And in particular, if we are not careful about the goals that we set and the capabilities that we provide these machines and these AI systems with, they have the potential to wreak havoc on our society.
- [18:41]
Brooke Gladstone
I love talking to you, Jerry.
- [18:43]
Jerry Kaplan
Oh, thank you so much, Brooke. It's a pleasure to talk to you, too.
- [18:46]
Brooke Gladstone
I only have one final question.
- [18:48]
Jerry Kaplan
Certainly.
- [18:48]
Brooke Gladstone
Do you think if you scratch an engineer underneath, you find a science fiction freak?
- [18:56]
Jerry Kaplan
I would love to give you a snappy answer. But I'll try it. I'm doing it right now. But I.
- [19:02]
Brooke Gladstone
You have just outlined the plot of about 700,000 science fiction novels in warning in a clear and dispassionate way of the potential hazards of artificial intelligence.
- [19:17]
Jerry Kaplan
You're absolutely right. And I come back to that IBM statement that computers can only do what they're programmed to do. That is completely false today. You know, I keep thinking of the example of squirrels crossing the road. Like squirrels haven't evolved to deal with 2,000 pound piles of metal that approach them at 60 miles an hour. And that's why they get run over. You're not designed to deal with a threat that's diffuse in the cloud. That is dealing with you in ways that you may not even realize when you look at the two great scourges of the modern world. The first being rampant unemployment and the second being income inequality. I think that a major cause of both of those is the accelerating progress in technology in general and in artificial intelligence in particular.
- [20:05]
Brooke Gladstone
But you can't stop that.
- [20:08]
Jerry Kaplan
No, you can't stop that. But there's a lot you can do about it. In other words, not to prevent it, but to make the transition to a new world much less bumpy. I think that robot Armageddon is going to be economic, not a physical war.
- [20:26]
Brooke Gladstone
Jerry, thank you very much.
- [20:28]
Jerry Kaplan
Okay, thank you, Brooke.
- [20:30]
Brooke Gladstone
Jerry Kaplan is an entrepreneur who teaches the history and philosophy of artificial intelligence at Stanford University. Earlier this month there were reports of Google scooping up eight robotic companies in a mere six months. Redwood Robotics specializes in arms, not weapons, arms. Industrial perception focuses on 3D vision. And Mecca produces lovable humanoid heads. Since Google hasn't said much about why it bought these and the other companies, the media fill the void with nervous speculation. What might Google do with a robot army? Particular focus centered on Boston Dynamics, a company that's used military money to develop Big Dog, which hops and hums over rough terrain like a man sized insect. Wildcat that sprints like a metallic wolverine. Sand Flea, which scuttles like a crab and leaps like a cat. And Atlas and Pet man, who believe me, you've already seen in your dreams. Henrik Christensen is a professor of robotics at Georgia Tech. Welcome to the show.
- [21:44]
Alex Goldman
Thank you for having me.
- [21:45]
Brooke Gladstone
This list of Recent acquisitions covers AI disaster relief robots, a four legged robot that can run 30 miles an hour. Why?
- [21:56]
Alex Goldman
Well, it really all boils down to competition on the market today. If you're trying to do product research, you will typically go to Amazon. Google is trying to get into logistics and material handling so that they can actually compete and make sure you go there to buy products rather than going to Amazon.
- [22:14]
Brooke Gladstone
Right. We search Google for information. We search Amazon for stuff. It wants us to search Google for stuff.
- [22:21]
Alex Goldman
Exactly. So, for instance, in San Francisco, there is something called Google Express where the idea is that you can basically go to them and say, I want to buy a ring for my wife. And then they will actually send a driver to the store, buy the item, bring it to your spouse, and in the long term, now that we have autonomously driving cars, if we now use these smart robots, we would be able to pick these items up and actually deliver them. It's very much about getting you back to use Google because a lot of their revenue is based on advertising. So if they know what you're looking for, they can make sure that they put the right ads in front of you.
- [23:01]
Brooke Gladstone
All right, I could see how driverless cars might do that. But what do you need Big Dog and Wildcat and Cheetah for that are tested for rough terrain. And the other ones that work through water, is that so they don't have to use the highways?
- [23:16]
Alex Goldman
If we think about it, to me, there are really three reasons why you do these acquisitions. One of them has to do with manufacturing. So can we automatically manufacture electronic goods, various kinds of assembly actions? Two, can we do the supply chain so we can drive it back and forth and free? Today, Google Maps are incredible. You can get Street View wherever you can drive a car, but you can't really get these very detailed Google Maps as soon as you go off road. So being able to use Boston Dynamics and the other company that they bought, Schraft, gives them very agile robots and a high degree of autonomy that would allow you to build very detailed Google Maps of anywhere in the world. So you really have to look at it as the entire supply chain. How can you do better manufacturing, how can you do better deliveries? And how can you make sure that you have what we in technical terms would call a geographic information system? So now I can track you with my Android phone. I can know where were you last weekend? Based on this, I can have much more detailed information about what should I put up on your webpage to make sure that you buy the right things?
- [24:23]
Brooke Gladstone
Okay, so you're talking about robotic Henry Ford type plants, robotic Pony Express, and potentially robotic Lewis and Clark.
- [24:33]
Alex Goldman
Absolutely.
- [24:34]
Brooke Gladstone
But the Guardian noted that Google has not only bought almost every machine learning robotics company it could find, it also hired Jeff Hinton, probably the world's leading expert on neural networks and Last year, it brought on the famed inventor and AI advocate, Ray Kurzweil. Google, according to the Guardian, could embark on what one investor called a Manhattan Project of AI.
- [25:04]
Alex Goldman
There's no doubt they've solidly bought many of the best brains to work in this area. As you mentioned, we have really good people on machine learning. We have fantastic perception people. We have really good robotics people. If you had to collect sort of the most impressive group of roboticists in the world, Google would have it today.
- [25:24]
Brooke Gladstone
It strikes me that it's a natural development for the company that created an algorithm that can now provide you with searches by working off your previous searches and even by understanding your misspellings. That develops translation technology, also based on collecting data of people's mistakes and requests. That a company with that kind of expertise in software would move into robotics because, as one scientist said, robotics is different from software because it turns data into physical action automatically.
- [26:06]
Alex Goldman
Exactly. That's the ultimate dream. You know, can we build a very intelligent assistant that will work in your daily life? I think it's still a bit down the road, but if we interconnect the world. If you look at another company that Google has bought, they bought Nest. Some people call it the thermostat company. What it really is, they have a number of key technologies on interconnecting to your home. The thermostat is one. The fire alarm is another one. It can understand what is the air quality, but you can also imagine that it's connected up to your infrared detector that will detect if there are any people present and if they move around so that the light turns on when you enter the room, and when you exit the room, it turns off again. If you fall down and you don't move for 30 minutes, it could signal to somebody that you might want to go and check on this person. So it's the ultimate service economy. You can tell me what services you want and we will deliver it to you. Of course, it will be at a cost.
- [27:03]
Brooke Gladstone
Thank you so much, Professor.
- [27:05]
Alex Goldman
Hey, thank you for having me.
- [27:07]
Brooke Gladstone
Henrik Christensen is a professor of robotics at Georgia Tech. Coming up, bots that answer serious questions and tell stupid jokes. This is on the Media.
- [27:21]
Sponsor Voice
Support for on the Media comes from gsk, whose Discovery Fast Track challenge for academics seeks to turn concepts into medicine. @gsk.com discoveryfasttrack lumosity.com a brain training program design designed by neuroscientists that has been used by 50 million people worldwide. More@lumosity.com and Road Scholar with educational travel adventures to 150 countries, including Cuba. @roadscholar.org.
- [27:53]
Brooke Gladstone
This is on the media. I'm Brooke Gladstone. As Google buys up smart robots, some of which were developed with money from the military, the military deploys AI to help fill its ranks. Right now I'm on the Internet with an army recruiter. I'm typing in. Who are you? My name is Sergeant Starr, which stands for strong, trained and ready. I'm an artificial intelligence agent created for.
- [28:23]
Henrik Christensen
The US army to provide you with.
- [28:25]
Brooke Gladstone
Information about army life. So how do we feel about Sergeant Starr? Alex Goldman of our podcast TLDR wasn't sure. So he talked to the army and other experts to consider the pros and cons of deploying a Siri to guide our decision to go to war.
- [28:42]
Dave Moss
Sergeant Starr is the Army's robot. Specifically, he's a chatbot. But unlike some chatbots, he's got a face. If you go to the Army's website, you can see it. He's white, 30 something, and computer generated. He straddles the uncanny valley between approachable and kind of vacant looking. Like most chatbots, he's there to talk to you about one thing and one thing only. This chatbot wants to talk to you about your service in the U.S. army. And he doesn't cotton too much to goofing around. Tell him to do some push ups.
- [29:11]
Brooke Gladstone
Hey, I'm the sergeant here. You drop down and give me 20. I can't hear you. Count em.
- [29:18]
Dave Moss
I found out about Sergeant Starr from Dave Moss. Dave's an investigative researcher with the Electronic Frontier foundation. And for him, Sergeant Starr was just the latest bot he's formed a relationship with.
- [29:28]
Henrik Christensen
I think the first bot I encountered was Eliza. And this was a very rudimentary bot that was supposed to function as a therapist. And I remember sitting in front of my computer at college having these long conversations, trying to fool it, trying to get it to get angry at me. And they would just ask more questions. But there was something about engaging in the middle of the night with this character that fulfilled this. I don't know if it was a loneliness or if it was just a need for late night communication.
- [29:57]
Dave Moss
Even though he knew Eliza was a bottle, she managed to simulate to some degree human contact, which is what made the idea of a recruitment bot a little creepy to him. So Dave got curious about how Sergeant Starr works. But asking Sergeant Starr about himself doesn't get you very far. He just pulls rank because I said.
- [30:15]
Brooke Gladstone
So, and last I checked, I outrank you.
- [30:19]
Dave Moss
So last November, Dave decided to foia the United States Army There were several.
- [30:23]
Henrik Christensen
Things we asked for. One of the things was primarily was the input and output scripts. These are all the things that Sergeant Starr can recognize and all the things that he can possibly say back to you. And we asked for that for every year he's been around, so we could compare how he's expanded over the years and how his answers have changed. We also asked for guidelines for using him, privacy policies, how many people have used him, how many hours of manpower were saved.
- [30:49]
Dave Moss
Dave didn't get any response, not even one saying they wouldn't release the information he requested. So I decided to just give them a call to see if they'd talk. Surprisingly, they were game. If the sound quality is a little funky here, it's because I spoke to a bunch of military officials and contractors who were all huddled over one speakerphone.
- [31:05]
Brooke Gladstone
It's our pleasure.
- [31:06]
Jerry Kaplan
We're quite proud to talk about this innovative tool.
- [31:10]
Dave Moss
To answer some of Dave's questions. Sergeant Starr has a total of 840 answers he can currently give. His answers are analyzed every day, and they're adjusted every month. And the answers he gives change as military policy changes. Although the army collects every question that's asked of Sergeant Starr, it doesn't collect any personal information of the askers themselves. The army wasn't willing to divulge proprietary information about how Sergeant Starr works, but Colonel Keeter was otherwise pretty forthcoming.
- [31:35]
Jerry Kaplan
Over the last five years, I believe the average is about 500,000 user engagements per year, and about 2 million questions are answered.
- [31:48]
Brooke Gladstone
That's estimated to be an equivalent of 55 man years.
- [31:52]
Dave Moss
Cleet Grumley from Next IT the company that programmed Sergeant Starr told me that when the army built him, it endlessly focus grouped him.
- [31:59]
Brooke Gladstone
The army was going to have several different avatars, potentially. And then when they did focus group tests with people in a bunch of different ethnicities and age groups in different parts of the country, they all basically.
- [32:11]
Alex Goldman
Picked the same avatar.
- [32:14]
Brooke Gladstone
And so the army elected to go ahead and just go with one. The current version is actually built off.
- [32:19]
Jerry Kaplan
Of a live soldier, and then it's.
- [32:21]
Brooke Gladstone
Rendered to make it look slightly animated.
- [32:24]
Dave Moss
Are there certain things that you didn't even know that recruits were interested in finding out until Sergeant Starr started getting those questions.
- [32:32]
Jerry Kaplan
Yeah, the one that comes to mind is if you have to shower with other folks.
- [32:37]
Dave Moss
This was the most surprising thing to me about Sergeant Starr, that there are questions that potential recruits couldn't ask a recruiter, but felt comfortable asking a chatbot.
- [32:45]
Brooke Gladstone
Like, are there jobs where I wouldn't have to be deployed? Things about what they can bring to basic training, even down to they would ask Sergeant Star questions about what toiletries should they bring.
- [32:54]
Dave Moss
There are questions a teenager might have about war that they're too scared to ask a person but happy to ask a chatbot. Granted, Sergeant Starr is just the first step to recruitment, but as first steps go, it seems like a pretty good one. Still, I wanted to test my assumptions, so I got in touch with an expert.
- [33:10]
Sponsor Voice
Helen Benedict, professor at the Graduate School of Journalism at Columbia University and the author of Sand Queen and the Lonely Soldier, two books about the Iraq war.
- [33:20]
Dave Moss
In covering the Iraq war, Helen's heard stories of recruiters forging parental permission slips for underage recruits, lying to candidates about where they would end up, sexual assaults, and much more. In the early aughts, when Helen was reporting on this, the army disclosed a, quote, 60% rise in inappropriate actions by recruiters. So when I asked Helen about the idea of Sergeant Starr, I figured she'd be on board. After all, a chatbot doesn't lie, and he doesn't pressure people to go to war so he can meet his quota.
- [33:48]
Sponsor Voice
It certainly takes away the power of an individual to manipulate a young recruit. But on the other hand, the choice to enlist is absolutely enormous, will change your life forever, if not, end up getting you killed or maimed. And I think some communication with an honest human being is terribly important. The answers that this robot will give will be the official line. And the official line in the military, in my view, as a result of my research, is not always honest at all. Supposing you ask the robot, what are my chances of getting sexually assaulted as a female or I'm gay and now don't ask, don't tell is over. But how is that going to actually affect me? I could predict that the answers will be the official line.
- [34:33]
Dave Moss
What are my chances of getting sexually assaulted in the military?
- [34:36]
Brooke Gladstone
All soldiers are subject to the uniformed code of military justice and army regulations which govern behavior and conduct. Commanders have a wide range of administrative and disciplinary measures to address substandard behavior and misconduct. Each case is addressed differently depending on the circumstances.
- [34:57]
Sponsor Voice
Totally as I predicted.
- [35:00]
Dave Moss
Can you ask it about don't ask, don't tell and bullying?
- [35:03]
Sponsor Voice
I'm a gay man, and I want to enlist as an openly gay man. What are my chances of being treated.
- [35:09]
Alex Goldman
Badly if otherwise qualified individuals may join.
- [35:13]
Brooke Gladstone
And serve in the army without regard to sexual orientation?
- [35:18]
Dave Moss
That's it.
- [35:21]
Sponsor Voice
Listen, the truth of the matter is, a live recruiter might well give you the same answers. I think there will be an honest recruiter who could tell you, look, we do have a problem with sexual assault from the military, but we're really working on it. And here are some of the things you can do to protect yourself and might actually get further with succeeding in recruiting someone.
- [35:44]
Dave Moss
It turns out that Sergeant Starr is a lot of different things. He's a robot who never sleeps. He's a person you can ask questions that you're scared to ask any other human being. He's also a rigidly bureaucratic pro army voice that can give oblique answers to questions about your basic safety. He's frustratingly full of contradictions and strangely, that makes him feel more human.
- [36:05]
Brooke Gladstone
Thanks, I try. OTM producer Alex Goldman is co host of our podcast TLDR. You can also find a blog post featuring all 8, 835 of the answers Sergeant Starr gives on the Electronic Frontier Foundation's blog at eff.org deeplinks you can hear a longer version of this piece and a follow up interview with David Moss at tldr and on themedia.org I'm.
- [36:36]
Dave Moss
The operator of my pocket calculator.
- [36:40]
Brooke Gladstone
As Sergeant Starr soldiers on, another bot is being tested to generate not grunts, but laughs. Decades of research has not yet yielded a machine capable of delivering a tight 10 minutes of observational humor. But computers can, under certain circumstances, grind out a joke. Recently, Joel Warner and Peter McGraw, authors of the Humor Code, A Global Search for what Makes Things Funny, pitted comedian Mike Kaplan against Manatee, a computer created to tell jokes. Joel Warner is going to see if I can tell which joke is human and which isn't.
- [37:19]
Mike Kaplan
Okay, you're gonna guess now. Okay, I like thrills like I like flights.
- [37:25]
Brooke Gladstone
Cheap, cheap thrills. Cheap flights. I think that since cheap in this context is being used as kind of salacious, sexy, that that was the human.
- [37:37]
Mike Kaplan
Mike, you are incorrect. Oh, that was Manatee.
- [37:44]
Brooke Gladstone
All right.
- [37:45]
Mike Kaplan
Another I like fingers like I like my notes sticky.
- [37:50]
Brooke Gladstone
That's the computer.
- [37:52]
Mike Kaplan
Yes.
- [37:53]
Brooke Gladstone
Actually, artificial intelligence researchers have been studying computers and humor for a long time.
- [38:00]
Mike Kaplan
Yeah. Among my favorites would be Lib Job, which is the light Bulb Joke Generator, the double entendre noun transfer program, which figures out where inside passages of text to insert the line. That's what she said.
- [38:17]
Brooke Gladstone
The current effort, the one that yielded manatee, was a $700,000 government project.
- [38:26]
Mike Kaplan
Yes. Project received stimulus funding. And when Republicans heard that a joke computer received $700,000, they were fuming.
- [38:36]
Brooke Gladstone
How can you justify that?
- [38:37]
Mike Kaplan
I think there's a valid reason we spend so much time with our gizmos. We Want the interactions to be as smooth and fluid as possible. In some ways, we want these things to almost be a bit human. I mean, who wouldn't want a GPS device that after you ignore it for the 20th time, it somehow sarcastically rolls its eyes at you?
- [38:59]
Brooke Gladstone
I would say almost everybody wouldn't want that. But I'm sure there's another example.
- [39:04]
Mike Kaplan
Well, think about when a webpage doesn't load now.
- [39:06]
Brooke Gladstone
Yeah.
- [39:07]
Mike Kaplan
Designers think we might be a little more okay with this stumble if a computer almost makes humor about it.
- [39:13]
Brooke Gladstone
I guess if it were real, personifying a program can also run into a lot of problems. I mean, Clippy, may he rest in peace, was probably the most deeply unpopular thing ever produced by what Microsoft.
- [39:28]
Mike Kaplan
If Clippy made fun of himself, say, hey, I know you don't like me, but you pressed the wrong button and here I am. I'm gonna do my little Clippy dance. I mean that, right? That might help make things a little more.
- [39:40]
Brooke Gladstone
You are so right.
- [39:41]
Mike Kaplan
One of the examples of human intelligence is our ability to process all this information around us and find new and novel ways to make things funny. So it's actually a really good challenge, a really good goal.
- [39:53]
Brooke Gladstone
Do humans really understand what makes things funny?
- [39:56]
Mike Kaplan
That's a very good question. There have been many attempts over the centuries to figure out what exactly makes things funny. You can go back to Plato and Aristotle, who said it was about superiority, laughing at the misfortune of other people. Then came the incongruity theory by folks like Kant who said that humor arises when you're led to believe one thing and something else happens. Setups and punchlines work well. And then there was Freud, of course, who had his own theory based on.
- [40:26]
Brooke Gladstone
Sexual frustration, no doubt, of course. And do you have a favorite?
- [40:29]
Mike Kaplan
I have to say I'm partial to the one my co author and I looked at in the book the Humor Code. His name is Peter McGraw, and his theory is something called the benign violation theory. The idea that humor arises when something seems benign while simultaneously appears to be a violation or somehow wrong.
- [40:47]
Brooke Gladstone
Of course, all of these can be right.
- [40:49]
Jerry Kaplan
Yeah.
- [40:50]
Mike Kaplan
There hasn't been enough hard and fast research to say, this is it. This is the end all be all of human research.
- [40:56]
Brooke Gladstone
Okay, so a guy is going down the street and he sees his friend. He hasn't seen him for I don't know how long, and he has this big orange head. And he goes up to him and he goes, hey, what's with the big orange head? And he goes, you know, it's a funny story. I Was in an antique shop, and I found this lamp, and I rubbed it, and a genie came out and gave me three wishes. And so I. I wished for a gorgeous house. And you see behind me this huge mansion. Yeah, it's really nice. That's the house. And then I wished for a beautiful wife. And you see that really lovely blonde coming down the street? That's my wife. And then here's where I think I went wrong. I wished for a big orange head.
- [41:45]
Mike Kaplan
The amount of kind of human intelligence that you and I have to have to one, tell that joke and create that joke and then appreciate it.
- [41:53]
Brooke Gladstone
I mean, the smartest person I know did not get that joke.
- [41:56]
Mike Kaplan
I think it's less about pure intelligence, but, I mean, how you process the world, how you see the world. I mean, how are you gonna teach a computer to do that?
- [42:05]
Brooke Gladstone
But if we were to do that, would that be the apotheosis of artificial intelligence?
- [42:11]
Mike Kaplan
Some people have suggested that one of the ultimate tests is to get a computer to tell jokes and appreciate jokes in all their forms. I think we're still a ways off on that.
- [42:22]
Brooke Gladstone
What was the best joke told in the contest?
- [42:26]
Mike Kaplan
Oh, okay. I have to say the one that sums the whole contest up. I like jokes like, I like robots. Efficient and killer.
- [42:36]
Brooke Gladstone
And whose joke was that?
- [42:38]
Mike Kaplan
Thankfully for the human race, that was Mike Kaplan.
- [42:43]
Brooke Gladstone
Joel, thank you so much.
- [42:44]
Mike Kaplan
Thank you for having me.
- [42:45]
Brooke Gladstone
Joel Warner is the author of the humor Code Mike, that's my Q. Kaplan is the comedian that competed with Manatee the Computer. He was chosen in part for his quote, robotic delivery.
- [42:58]
Henrik Christensen
Had some other jobs. I talked to a street musician once. Found out that you can make up to $60 an hour being a street musician or up to $60aminute robbing a street musician.
- [43:09]
Brooke Gladstone
So it's hard, isn't it? Just standing there alone on the high wire.
- [43:14]
Henrik Christensen
I mean, it's a subjective thing. I'll do a show for military, and they'll come up after and say, I could never do what you do. And I'm like, well, that's rude. You know, it's not. I. I could say the same. You know, like, worst case scenario for me is a room full of people doesn't like me. That's something I dealt with in high school every day.
- [43:30]
Brooke Gladstone
Some say your comedy has been described as robotic. And your new album is called Meat Robot.
- [43:38]
Henrik Christensen
I definitely understand that. I like relationships and friendships, but I don't always necessarily outwardly manifest those emotions as much as other people do. So people like, for example, when I was married, I Remember, my wife would call me Huzzbot. Now it's Ex Husbot. Perfect.
- [43:55]
Brooke Gladstone
Some comedians might have been afraid that they would lose to a robot.
- [44:01]
Henrik Christensen
I mean, comedy is subjective. There are, I'm sure, people who think that this robot is better than either way. I was excited to be part of the story. If I win, then hooray for humanity. And if I lose, then we're the ones who made the computers. So it's still a win for humanity, at least.
- [44:16]
Brooke Gladstone
So the structure of this was very simple. Yes, I like this. Like, I like this. Was that a structure you'd used before?
- [44:25]
Henrik Christensen
It's sort of a comedy trope. The robot can find, oh, these two words mean the same thing, but in different situations. People might not have thought of that. So here is an incongruity. Is that a joke? That's my impression of the robot.
- [44:37]
Brooke Gladstone
So how do you think the robot did at that?
- [44:39]
Henrik Christensen
I think it did great. I don't think they were the most sophisticated jokes of this type, even, but I think they made sense. Most people are not the best when they start doing comedy. And I think this robot is off to a great start.
- [44:51]
Brooke Gladstone
Let's grade some of Manatee's jokes. I like fingers. Like, I like my notes sticky.
- [44:59]
Henrik Christensen
Yeah, that. That one, I guess. Now, thinking about it, that one could use some more work, I guess. Like, I like fingers didn't really. Oh, why do you like fingers? Like, nobody would ever say, I like fingers unless there's some real specific great payoff of why they like fingers. But in that situation, it seems like it just thought, oh, it's seen the term sticky fingers and sticky notes. There's no real, like, you know, resonance.
- [45:25]
Brooke Gladstone
That it's just about words. I like periods. Like, I like my verbs irregular.
- [45:31]
Henrik Christensen
This robot is talking about a woman's cycle. That's. If a robot were to perform, I think that would be sort of the same way that, you know, Betty White. She's been funny for her whole life, but there's a specific thing that now is even, like, more incongruous about our idea of what an old person is supposed to talk about. It's like she doesn't even have to say a specifically funny thing, but people will think it's funny because it's an old person talking about sex.
- [45:53]
Brooke Gladstone
Given the state of artificial intelligence these days, do you think this had any meaning? We can't really learn anything about humans by making a computer make a joke. It's still purely technical.
- [46:08]
Henrik Christensen
I guess the ideal is, like, I don't want you know, robots taking my jobs, but as long as I keep working and as long as every other comedian can, like maybe robots can be trained enough to, you know, go into hospital wards and, you know, the 23 hours a day that there's not a comedian there, they're like, what do you think about these jokes, robots? Or like, you know, into an old folks home or, you know, doing a benefit for veterans, literally. I have no idea what people are planning on doing with joke robots, if anything. So no, no greater meaning.
- [46:35]
Brooke Gladstone
So what advice would you give Manatee?
- [46:37]
Henrik Christensen
Write and perform as much as possible. You know, figure out what works, bring it to the people. You've got a great avenue right now being the only joke robot. But I would say try to take a lower profile initially. Just write a lot and then try things out on smaller crowds. Then I assume you'll get to a point where, you know, you might overtake me. And I don't want that to happen. So I'd say you just try to live a more fulfilling life, you know, become a service robot. If you would. What do you want? What do you really want?
- [47:03]
Brooke Gladstone
Manatee, can I tell you this joke?
- [47:06]
Henrik Christensen
Of course.
- [47:07]
Brooke Gladstone
Okay. You may have heard this before.
- [47:08]
Henrik Christensen
It's possible.
- [47:09]
Brooke Gladstone
If you have, just stop me.
- [47:11]
Henrik Christensen
Sure.
- [47:12]
Brooke Gladstone
Okay. So there's this guy and he sees an old friend of his he hasn't seen for a long time and he has a big orange head.
- [47:18]
Henrik Christensen
This is my favorite joke.
- [47:20]
Brooke Gladstone
It's your favorite joke.
- [47:21]
Henrik Christensen
I love it so much.
- [47:23]
Brooke Gladstone
Sometimes the smartest people just don't get it.
- [47:26]
Henrik Christensen
Do you want to hear my theory?
- [47:27]
Brooke Gladstone
Yes.
- [47:28]
Henrik Christensen
If you heard for the first time, why did the chicken cross the road? You're expecting some grand answer, but the answer is to get to the other side. You know, the expectation is it's something. So the dashing of that expectation is that it's nothing. Which is the same thing here. You probably had heard so many jokes about genies where people wish for something and it goes slightly awry. I didn't say a 12 inch pianist. You know, I didn't say a million ducks. You know, your brain has become practiced in the art of understanding that that's what's going on here. These other people who don't get it, they're just people who I guess led lives full of not enough jokes.
- [48:02]
Brooke Gladstone
It's so funny. Cause the big orange head joke is my favorite joke too.
- [48:06]
Henrik Christensen
It's wonderful.
- [48:07]
Brooke Gladstone
I always thought it was. Because the answer is that human beings just do stupid things, just really idiotic choices.
- [48:17]
Henrik Christensen
I mean, your interpretation is by no means incorrect.
- [48:21]
Brooke Gladstone
But it's true that in the moment, which is where you live as a comedian, it would be, well, we all know you know the genie trope.
- [48:28]
Henrik Christensen
Yes, you bring this greater human meaning to it. But I don't think anyone could not have heard any jokes like them and still feel that exact same way about it. So it's certainly not the only thing at play, but it's definitely a portion of it.
- [48:42]
Brooke Gladstone
This brilliant friend, I think you're right. A, she doesn't have enough jokes in her life, even though she laughs a lot, but not at jokes. And B, I don't think that she immediately goes to the place that human beings just make completely illogical choices.
- [48:59]
Henrik Christensen
So you say, well, wish for a big orange head. And she's like, well, why would he do that? That doesn't seem like the thing to do. What's the rest of the joke?
- [49:09]
Brooke Gladstone
Thank you so much, Mike.
- [49:10]
Henrik Christensen
Absolutely. Thank you so much for having me.
- [49:12]
Brooke Gladstone
On May 15, Mike Caplan's comedy special comes out on Netflix. It's called Small Thought and Handsome Robots need love too.
- [49:25]
Henrik Christensen
They want to be loved by you.
- [49:28]
Brooke Gladstone
We anthropomorphize, experts say, because it's the way we create order in a complex world. We see faces everywhere in a slab of rock, the grill of a car. Though computer intelligence is seductive, it's just as inhuman, answerable only to human human overseers too remote to see its impact on the rest of our species. So go ahead and project humanity onto the howl of the wind. But if you sense it in your operating system, unplug and take a shower. That's it for this week's show on the Media was produced By Alex Goldman, P.J. voats, Sara Abdorahman, Chris Neary, Laura Mayer, Meera Sharma and and Kimmy Regler. We had more help from Cameron Lindsey and our show was edited by me. Our technical director is Jennifer Munson. Our engineer this week was Leslie Sisson. Katya Rogers is our executive producer. Jim Schachter is WNYC's Vice President for News on the Media is produced by WNYC and distributed by npr. Bob Garfield will be back next week. I'm Brooke Gladstone.
- [50:36]
Sponsor Voice
Support for on the Media comes from the John Dee and Catherine T. MacArthur foundation, the Ford foundation, the Jane Marcher foundation, the Overbrook foundation and the listeners of WNYC.
- [50:47]
Mike Kaplan
Radio.
- [50:50]
Sponsor Voice
NYC now delivers the most up to date local news from WNYC and Gothamist. Every morning, midday and evening with three updates a day. Listeners get breaking news, top headlines and in depth coverage from across New York City by sponsoring programming like NYC now. You'll reach our community of dedicated listeners with premium messaging and an uncluttered audio experience. Visit sponsorship.wnyc.org to get in touch and find out more.