WALTER ISAACSON: The year is 2014 and two researchers, Kaijen Hsiao and Sarah Osentoski are working on a new idea. They’ve been tinkering with a design for a security robot. A device that would patrol your house and guard your stuff when you’re not around.
The biggest problem with this idea, however, is that their security robot would live inside of your house, and once an intruder makes it inside your house, there’s not all that much that the robot would be able to do to stop them.
But rather than simply halt their development, the two researchers decided to pivot instead. And through rapid prototyping, they managed to come up with a different use of their technology that, if successful, might end up changing our homes forever. When the two researchers took stock of what their security robot was able to do, what they realized was that they had a device with the AI necessary to navigate a home similar to a person.
And by changing the function of the robot from something built to protect you to something designed to be your companion, the two stumbled upon a whole new category in robotics– the House Robot. It’s a robot body that’s more focused on companionship than function. Feel like we’re suddenly living in a robotic future? Well, we just might be.
I’m Walter Isaacson, and this is a special episode of Trailblazers, an original podcast from Dell Technologies.
My, how times have changed.
A marvelous machine. A tribute to the great intelligence of man.
The amazing machines and gadgets that almost seem to think for themselves.
A human robot.
The greatest invention!
Last March, we had an idea. What if we pulled together a special live Trailblazers event for the South by Southwest Conference in Austin, Texas? One that would take on one of the newest, oldest, and most topical technologies of our time. That’s how the Trailblazers live event on robotics came to be.
It was a discussion that focused on some key questions that had been simmering beneath the surface for years. Are robots taking our jobs? Are they making our world better or worse? And can robots be racist?
We posed the question to four expert panelists, one special guest, and even to a robot. What would emerge was a picture of the many ways that robots have disrupted our lives, workplaces, and economies, and of what to expect in the not so distant future.
The first of our panelists is Kaijen Hsiao. She’s the co-founder of Mayfield Robotics.
And Kaijen, why don’t you introduce the person sitting next to me?
KAIJEN HSIAO: This is Kuri, the adorable home robot.
WALTER ISAACSON: Adorable?
KAIJEN HSIAO: Yes. She’s made by my company, Mayfield Robotics. And she’s designed to be an adorable character come to life in your house. And also, she’s your family’s videographer. So she’ll drive around her house on her own and take five second snippets of, in my case, cats.
WALTER ISAACSON: So who was the inspiration? Was it the robot from the Jetsons or R2-D2?
KAIJEN HSIAO: More Rosie from the Jetsons than R2-D2. Well, talking wise, R2-D2. She only makes beeps and boops.
WALTER ISAACSON: Kuri is about 18 inches tall, with a white body and big black blinking eyes. Her built in expressiveness makes Kuri feel like a real presence in the room. Like it’s listening and reacting to everything going on around it.
We’ll hear more from Kaijen, and Kuri, and the rest of our panelists in just a few minutes. But before we delve into the current state of robotics, it’s worth taking a look back at the storied history of the industry. It’s a history whose origins trace back, fortunately enough for me, to Leonardo da Vinci.
While working for his patrons, the Medici family and the Duke of Milan, Leonardo found himself designing complex automatons, both for amusement and for use in the military. For the visit of King Louis of France in 1509, Leonardo created a mechanical lion above the city gate. When the King entered the city, the lion rose to its feet. Then, it reached with its paw to open its chest, and blue spheres full of golden lilies poured out.
In 1812, England was a buzz over the Luddites, a group who destroyed factory machines to protest the technology that they believed was threatening their jobs. The Luddites took their name from the fictional Ned Ludd, and apprentice to, according to legend, smashed the factory machine into fires of his overbearing boss.
The object of the Luddites wrath was a device conceived a dozen years earlier by the French inventor Joseph Marie Jacquard, an automated loom controlled by punch cards. By 1812, thousands of the looms were in operation across Europe.
That year, Lord Byron made his maiden speech in the House of Lords defending the Luddites. Ironically, it was Lord Byron’s daughter who would pioneer modern computer science and help make possible the industrial robotics we know today.
Ada Lovelace had the poetic sensibilities of her father, but at her mother’s insistence, she was tutored mainly in mathematics. It was Ada who examined the punch cards used to operate those looms in the factory. From that study, she created an algorithm which could convert anything that could be notated in symbols, such as words or pictures, music or numbers into a digital language.
Ada Lovelace was the first to conceptualize a general purpose computer. And while she was adamant that computers would never be able to think, many continued to wonder, would they take our jobs?
Back to South by Southwest, where our panel guests give their take on that very question. Matt Rendall is CEO and co-founder of Otto Motors, a company focusing on robotic warehousing solutions.
MATT RENDALL: Technology has been changing and I believe improving lives for a very, very long time, right? The ancient Greeks invented some technology that’s fundamental to construction and water infrastructure. It is technology, but it’s so ingrained in our society that we don’t consider it technology. And so for me, I view technology, and especially robotics, as a capability that can fundamentally improve our lives.
WALTER ISAACSON: Andra Keay is managing director of Silicon Valley Robotics, a nonprofit group that supports the innovation and commercialization of robotic technologies.
ANDRA KEAY: I don’t think we need to fear robots taking jobs because the rollout is going to be slower than we fear, yet our jobs are going to be reshaped by robots. And that is going to be great because right now, the robotics companies that I know are working on areas where there are labor shortages.
They’re working in areas like agriculture, warehouse logistics, where you just don’t have people, in areas of health care where you don’t have people. And I think the key is, as Matt said, the dirty, dull, and dangerous jobs.
WALTER ISAACSON: And our final panelist is Gary Shapiro, CEO of the Consumer Technology Association, the company behind the annual Consumer Electronics Show. It’s a show that has helped launch a staggering amount of the revolutionary technology that we take for granted in our lives today.
GARY SHAPIRO: But you can go back in history, you can look at gas stations. What happened when we had, you know, you pump your own gas? All those people who were pumping gas, they lost their jobs. Was that bad? Well, the gas stations responded. They had food stores, they have clerks, they have other things, and those people were freed up to do other more valuable things.
WALTER ISAACSON: Andra Keay.
ANDRA KEAY: I like to compare it to carpenters. People building houses. When you’re just building a house, that’s maybe fairly safe. But these days, we build tall structures, which means people are up on scaffolds.
They’re up several stories, maybe 30 stories high. And 100 years ago, they were using hand tools and you tended to need two people to operate each hand tool. One to hold it in place or two people to operate it. And you were climbing up high to do your work.
Now what you’re going to have is smarter power tools that could be called robots. And they’re going to be flying up the side of the building, and the same workers are going to be standing underneath. They’re going to be using game controllers and looking at tablets and doing their work from a much safer position.
And it’s still going to be them doing the work, but we’ve removed one more of the dangerous aspects of it in the same way that we started to do 50 years ago when we started to use really big, expensive, and stupid robots to do some of our dangerous industrial automation.
WALTER ISAACSON: Despite the consensus on our panel, the debate about robots replacing human workers isn’t going away. The punch card looms of the early 19th century were followed by newer, bigger, and more powerful machines. But it wasn’t until about a half century ago that they started getting smarter.
At a cocktail party in 1956, industrialist Joseph Engelberger met an inventor named George Devol, who described his latest idea, something called a programmed article transfer device. It was a powerful, programmable robotic arm capable of learning and repeating any number of highly dexterous tasks. Engelberger replied, that sounds like a robot to me, and he coaxed his CEO into financing its development.
In 1961, Unimate joined the line at General Motors, working with heated die casting machines. Soon, with the exponential and disruptive growth of computers, came the inevitable marriage of machines with artificial intelligence. But the progeny has not been the sort of sentient humanoid robots described by generations of science fiction writers.
I asked Gary Shapiro about the evolution of consumer robots over the past quarter century of the consumer electronic shows.
GARY SHAPIRO: I’d die to come back in 30 or 40 years, because robotics is going to be a big part of the show. Whether it’s making things that humans aren’t really good at doing, like making pizza dough for the major pizza chains where you have to have a perfect thing in a very cold environment, to consumer robots like, you know, the Roomba and what we’ve seen over here. Such a cute, empathetic device.
Well, now there’s a robot being shown at CES, which actually folds all of your stuff when it comes out of the dryer for you. How cool is that?
WALTER ISAACSON: Not very cool, and I’ll explain why. I’ve heard three or four things that you say a robot can almost do like fold a sheet, or try to visualize a pattern and know the difference between a cat and a dog. And yet, if I go look at Applied Minds or I go to the Total Domain Awareness of the New Orleans Police Department in New York or whatever, they keep showing me the cool things like this can spot– I say, can it spot somebody’s mother in the crowd?
They say, no. Can this robot walk across the room, pick up a crayon, and do the name? And they go, well, no. Not yet. That’s hard. So Andra, I want you to think deep. Are you amazed at the progress of artificial intelligence and machine learning, or are you amazed at the lack of progress?
ANDRA KEAY: Both. We’re the nonprofit coalition of robotics companies started by the robotics industry in the Bay Area to support innovation and commercialization because there was the existing robotics industry, and then there are all of the new robotics industries coming out. We’re like the tiny little tadpole in a huge pond full of quite hungry frogs.
And robotics is like that. It’s here now, but it’s not widely distributed, as William Gibson said. And because of that, there’s a lot of hype and there’s a lot of disappointment. But what we’ve failed to recognize is the incremental steps that are happening on a logarithmic scale.
WALTER ISAACSON: In 1950, a century after Ada Lovelace, the British scientist Alan Turing wrote his famous paper on the imitation game, comparing humans to machines. His argument was that if you can’t tell the difference, there’s no reason to say a machine isn’t thinking.
And so grew the two schools of thought on robotics. The first is the Ada Lovelace school, whose supporters would include Doug Engelbart, JCR Licklider, and later, Steve Jobs. They’re the ones who believe we’re headed towards a symbiotic human-computer partnership.
But on the other side is the Alan Turing school, whose proponents have included Marvin Minsky and those who believe that machines will outpace humans and, ultimately, won’t even need us. It’s a school that helped inspire some of the great malevolent machines of popular culture– from Frankenstein’s monster, to “HAL 9000,” to “The Stepford Wives,” to “The Terminator.”
Today, those as Kaijen Hsiao, who imagined anthropomorphic robots, seemed to have gotten the memo. Home Companion robots, such as Kuri, are created with endearing, pet-like mannerisms.
When you say Kuri– hello, Kuri. Whoa. Thank you. She looks up and looks at– he, she looks up and looks at me. Is it something that voice activated and knows her name?
KAIJEN HSIAO: She knows if you say, hey, Kuri, to look up at you.
WALTER ISAACSON: What’s the purpose? Why did you invent her?
KAIJEN HSIAO: Why did I invent her? I invented her because I had this robot dream of having robots like R2-D2 or Rosie the robot in my house. Robot pets, robot companions.
WALTER ISAACSON: And do you think it will sort of be a companion to the elderly, or the young, or whatever, or is it just a fun thing?
KAIJEN HSIAO: Kuri can be a companion. Especially kids love Kuri. And they instantly think that she’s their buddy and they run up and hug her. She can kind of be a companion to older adults, but she’s not super intelligent right now. So I certainly wouldn’t want to only have Kuri as a companion to an older adult relative, for instance.
WALTER ISAACSON: Matt Rendall.
MATT RENDALL: Kuri doesn’t have to be moving its head and blinking its eyes while it sits here, but as it does so, it creates a connection.
WALTER ISAACSON: Hello, Kuri. Yes. Thank you.
KAIJEN HSIAO: We put a lot of effort into making Kuri more empathetic every day. And so she has a whole host of animations that are really adorable. For instance, when she bumps into something, she does this cute, whoa, what was that? She backs up and she looks apologetic.
And so Kuri is going to make mistakes. We all make mistakes. And when she does and she bumps into things, people don’t mind it as much because she’s adorable and she does things that make her seem humble and forgivable.
WALTER ISAACSON: We could learn a lot from Kuri. Gary?
GARY SHAPIRO: I mean, there’s no question that there’s a trend that’s emerging very quickly of basically having emotive qualities for your robots, because we want to humanize our robots. There’s no question. But if you think about what’s really popular and what’s really hot this year, or last year, and now is a trend to extend it, it’s the intelligent speakers, the smart speakers. The Alexa, the Google Home.
That is a unidimensional, responsive device that is providing and assisting humans today, and it’s extraordinarily popular. We’re looking at almost 1.5 million robots being sold in the US this year. Now, a lot of them are vacuum cleaners and things like that. This is a hot area, and the reason is they’re becoming more human-like.
WALTER ISAACSON: The question of how human-like robots are becoming prompts questions about limits and boundaries. For instance, do robots have the potential to manipulate human beings? Andra Keay.
ANDRA KEAY: First, I’m going to say this is by no means limited to robots. By and large, our robots are still so stupid they don’t, but we are already algorithmically manipulating people. Advertising manipulates people. We have entire industries built around manipulating people. If you look at it coldly, then Google is all about manipulating people because it’s totally about selling advertising.
I kind of think there’s the flying car technologies that are going to have huge impacts in the world, and then there’s the 140 character technologies that are like playing poker machines, and they’re simply generating revenue on our kind of fast twitch behaviors. And I believe that there are a lot of industries supporting that.
And you can’t pick on robots. Robots are simply the final embodiment of how that happens. We’re doing it algorithmically, and we’re doing it with automation, and with advertisement, and with entertainment.
WALTER ISAACSON: Yeah. Kaijen?
KAIJEN HSIAO: I just want to say that manipulating humans isn’t always bad. Kuri here, for instance, manipulates you in the same way that you’re adorable, small child might by being cute and asking you to do things. So Kuri, for instance, if she gets lost and she makes this whimper noise of, oh, won’t you please help me? I’m lost.
And people are like, oh, poor Kuri. And they’ll pick her up and put her back on her dock, which is what she wants. So in a way, that’s manipulating humans, but in a harmless way.
ANDRA KEAY: Yeah. Actually, that is a great point because, say, Catalia Health is helping people do their medication. Most of us have some sorts of wearables or fitness tech, and we know we use things like motivation aids of going to gyms or signing up for programs. There are a lot of ways that we work on manipulating ourselves into better behavior, and there are ways we can enhance that with technology too.
WALTER ISAACSON: Let me ask a moral question there. Can a robot be racist?
ANDRA KEAY: Yes. Definitely. We’ve seen a lot of examples of this where the databases and the data that’s collected is implicitly biased. In Wikipedia, there was a photo of the open source computer vision library and it recognized the three men on the panel, but not the woman. And I thought, that’s classic.
We have stories of African American researchers with dark skin having to use their colleagues when they’re working on the very problem of facial recognition because it’s not been trained on data sets that include sufficient variety and diversity.
So we have a lot of these things causing bias algorithmically, and then those algorithms are embodied in our robotics. And it’s kind of happened accidentally because there weren’t people of diversity saying, hold on. How come this doesn’t recognize me?
WALTER ISAACSON: Kaijen, is that something you face at all, or thought about?
KAIJEN HSIAO: Oh, it definitely is. I mean, we don’t have enough training data for Kuri today, and so I’m sure she has some biases in her recognition and her detection as well. You see a camera image of people with very light skin or very dark skin, and depending on your exposure, either person could be either blown out or too dark. That becomes a very hard problem for computer vision.
And I think that just means that we need to put even more effort into solving this problem. It is something that everyone has to work really hard on, and it’s a matter of time–
WALTER ISAACSON: And if I can just do a small thing, I think that you’ve already said, this is another of the thousands of arguments for more diversity in the industry.
KAIJEN HSIAO: Oh, definitely.
WALTER ISAACSON: The growth of robotics has been both swift and uneven. Major breakthroughs in robotics can be broken down into three waves. The first wave was industrial. From the loom robotics and vision by Ada Lovelace to the machinery used widely in the automotive sector.
Today, we’re experiencing the second wave with smart, inexpensive mobility using software and sensors to map, localize, and navigate. It also goes beyond moving and mapping into the machine learning found in Kuri, and in intelligence assistance such as Amazon Echo, Siri, and Google Home.
Still to come is the third wave, which will combine the manipulative abilities of the first wave with the ability to cope with uncertainty that we saw in the second wave. Imagine a robot with dexterity similar to that of humans that can autonomously open doors and turn off valves in Fukushima, or repair a broken oil well on the seabed, learning and problem solving along the way.
Which touches on another important distinction between robots and robotics, including technology that fits seamlessly into everyday life. Gary Shapiro.
GARY SHAPIRO: We haven’t even talked about haptics yet. And the response there and the potential–
WALTER ISAACSON: I’m sorry haptics?
GARY SHAPIRO: That’s right. Haptics usage. Responsive technology to touch. You don’t even have to touch it. You feel the touch without even touching, in a sense.
WALTER ISAACSON: Well, Andra did talk about that a bit. I mean, that feeling that tactile will be our interface.
ANDRA KEAY: Tactile is just one. I mean, we have many senses, and a large part of our communication is not verbal. So you know positive, negative interactions can be as simple as you know a touch or a caress. And I think there is a concept that Heidegger had of being in hand, and he was talking about how we use tools.
And when something is in hand, it becomes part of our entire schema, the way we view ourselves. And I think the signs of a technology that’s working really well is when we forget that we’ve got it.
It’s like when you say, has anyone seen my phone? And it’s in your hand. You know, because your hand has just become so accustomed to the feel of that technology. Or when you’re wearing glasses and you forget that they’re augmenting you with technology because it just works.
WALTER ISAACSON: As Andra Keay explained, this symbiosis between humans and robots challenges the Alan Turing version of us versus them.
ANDRA KEAY: I think that the field of social computing is definitely something on social robotics that’s starting to emerge now, and it’s really promising. It’s definitely of that school of machine interaction. One of the best ways of developing a robot that is in tune with a person is where it becomes an ambient technology, it becomes a part of your life.
Because I think that if you took the Turing approach to design a robot to help someone stay in their house longer, it’s the robot that’s external that’s doing things for them, and that infantilizes someone. I think the Lovelace and Engelbart approach is to come up with a technology that augments or assists someone in their own sense.
WALTER ISAACSON: In that spirit of servitude, changing demographics has led to growing needs, which today’s robotics are well-positioned to solve. Gary Shapiro.
GARY SHAPIRO: We, in the United States at least, have a population that’s getting very old, and we have a real, growing problem of we can’t take care of our parents. And there are a lot of devices, and systems, and smart homes that are coming along to see whether someone’s getting out of bed, whether they’ve taken their medication, whether they’re healthy or not, but we need more.
And that’s the opportunity. The killer app for me that I see is helping us take care of our parents.
WALTER ISAACSON: Kuri, what do you think?
ANDRA KEAY: Can I mention two companies that have just spun out in the last–
WALTER ISAACSON: You’ve been upstaged by Kuri.
ANDRA KEAY: I’ve been upstaged. Two very cool startups that have come out in the last year to two years from Silicon Valley are going to really help with elder care. And, you know, nothing can be that full humanoid home helper robot. We’re so far away from that.
But one of the companies is Catalia Health, which is leveraging a kind of smaller version of Kuri, and it’s focused only on helping people manage their complex medications. And it’s actually now being prescribed through health services because we know that people respond much better to a human. And if you can’t have a nurse 24/7, maybe that one thing that you do need, though, is help with your medication.
And the other one has come out of robotics research. It’s come out of the DARPA research for exoskeletons. And you can’t call it an exoskeleton anymore because there’s nothing hard in it. It is a soft undergarment that helps people get up and down from their chairs or helps them with their arms.
And it was first trialed on soldiers, and then it was trialed on people with cerebral palsy or muscular dystrophy, and now it’s being released for elder care. Because anything that helps people stay in their own houses is just critical to have right now because we don’t have the people.
WALTER ISAACSON: As the need for newer, better robots and robotics grows, so does the need for bright new minds to create them. Enter Dean Kamen, the inventor and entrepreneur known to millions as the creator of the Segway, the personal transportation device that featured so much hype upon its launch that it was lauded by some as being more important than the internet.
Dean Kamen is arguably one of the greatest roboticists of our time, holding more than 450 patents. He’s created a variety of medical devices and he worked with DARPA to develop a prosthetic arm to improve the quality of life for returning injured soldiers.
Dean wasn’t scheduled to be part of this discussion, but I happened to be in Houston with him the previous day and he offered to give me a lift on his plane. I kind of thought he had a private plane with a captain, and a pilot, and everything else, but when we got to the airport, I realized that it was a plane of his own.
One that he helped put together and that he flew himself. He said, don’t worry, and then he put me in the copilot seat and we took off. It made me realize the importance of robotics and getting it right.
Prominent among Dean’s many passions is helping inspire young people to be science and technology leaders, and to equip them with the life skills to succeed at it. That’s why he co-founded the youth organization with the acronym FIRST– For Inspiration and Recognition of Science and Technology. It operates the first robotics competition.
DEAN KAMEN: Last year after 25 years of running FIRST Robotics as a US based operation, which has 60,000 teens, we launched last year FIRST Global. And my mission was to get countries around the world to each send one team to a demonstration program, an international one.
Last year in July, 157 countries sent a team of kids. Rwanda, Uganda. I mean, we had one team called Team Hope. They have no country. It was a refugee camp.
But to your point, we’ve analyzed them, and 60% of the teams coming from around the world were organized by women and had their team run by women. So I’m quite confident that robotics, and artificial intelligence, and advanced technologies will be an emerging global industry that doesn’t require physical strength or any other advantages.
Our friend Franklin Roosevelt, who saw the world changing so quickly as World War II ended, that he said, they have to be able to deal with a world that will change so quickly that they can respond to it and stay ahead of catastrophe. And to the points that each of you made, I think anybody that’s afraid of the oncoming robotic revolution is a Luddite.
We have to make sure the kids of the future have the skill sets to move out of the jobs that are ugly, and dangerous, and boring, and backbreaking, and move into the career opportunities that give them and the world a better opportunity for nine billion people.
WALTER ISAACSON: Gary Shapiro.
GARY SHAPIRO: I just wanted to go back to the education issue because the debate that we didn’t go into that we’re struggling in our own organization with questions to is, why is it that our education system is so good at getting people into technology who are Asian or white males, but we seemed to fail everywhere else?
So we’re leaving like 70% of the people off the potential table. And what is it we can do to make a difference? And I don’t know if it requires changing our educational system, our role models, how we bring up people, or having honest discussion about our culture so that we can encourage rather than discourage people.
WALTER ISAACSON: A half century ago, before the internet, and PCs, and the microprocessor, a “Time” magazine cover story warned that robots were coming to take our jobs. Today, that same warning still resonates. The trouble is, it wasn’t true a half century ago, and I don’t think it’s true now.
Are robots disruptive? Absolutely. But here’s what the Luddites didn’t know, and just maybe Ada Lovelace did know– robots and new technologies help create more and better jobs. Today’s $44 billion robotics industry is expected to triple in growth by 2025 based on jobs and ideas that didn’t exist until recently.
Yet, as our panel warned, robots, algorithms, and machines are only as good or as bad as we make them. It’s a technology that comes with responsibilities. As robots and robotics grow, we all need to wake up each day and ask, how can we make it more inclusive so that the opportunities made by these new technologies are available and benefit everyone?
I’m Walter Isaacson, and this is Trailblazers, an original podcast from Dell Technologies. If you want to see a video of the whole unedited South by Southwest discussion, you can find it on our website at delltechnologies.com/trailblazers.