Bold (10 page)

Even better, for entrepreneurs interested in building Watson-backed business, Cane was stunned by how easy it was to work with IBM. “They provided so much support and guidance,” he explains, “that we were able to build our entire Watson-powered prototype in two weeks.”

One of this book's core goals is to point out those pivotal moments when a technology becomes ready for entrepreneurial prime time. Watson in the cloud, tied to an openly available API, is the beginning of one such moment, the potential for a Mosaic-like interface explosion, opening AI to all sorts of new businesses and heralding its transition from deceptive to disruptive growth. Attention, exponential entrepreneurs: What are you waiting for?

And everything we've just covered is here today. “Soon,” says Ray Kurzweil,
⁴⁰
“we will give an AI permission to listen to every phone conversation you have. Permission to read your emails and blogs, eavesdrop on your meetings, review your genome scan, watch what you eat and how much you exercise, even tap into your Google Glass feed. And by doing all this, your personal AI will be able to provide you with information even before you know you need it.”

Imagine, for example, a system that recognizes the faces of people in your visual field and provides you with their names. This shouldn't be too much of a mental stretch, as these capabilities are already coming online. Now imagine that this same AI also has contextual understanding—meaning the system recognizes that your conversation with your friend is heading in the direction of family life—so the AI reminds you of the names of each of your friend's family members, as well as any upcoming birthdays they might have.

Behind many of the AI successes mentioned in this section is an algorithm called Deep Learning. Developed by University of Toronto's Geoffrey Hinton for image recognition, Deep Learning has become the dominant approach in the field. And it should come as no surprise that in spring of 2013, Hinton was recruited, like Kurzweil, to join Google
⁴¹
—a development that will most likely lead to even faster progress.

More recently, Google and NASA Ames Research Center—one of NASA's field centers—jointly acquired a 512 qubit (quantum bit) computer manufactured by D-Wave Systems to study machine learning. With lightning speed, this computer can tackle face and voice recognition, as well as understanding biological behavior and the management of very large systems. “The tougher, more complex the problem,” says Geordie Rose,
⁴²
D-Wave's cofounder and CTO, “the better the results. For most problems, it was eleven thousand times faster, but in the ‘more difficult' category it was thirty-three thousand times faster. In the ‘most difficult' category, it was fifty thousand times faster.” So when Stark asks JARVIS to look at a massive amount of imagery data and pick out certain faces in the crowd, well, JARVIS is probably using qubits.

Why am I telling you about artificial intelligence aided by quantum computers? Not because I expect you to start developing these machines or using quantum computing (though a new SU start-up called 1Qbit
⁴³
has created an online user-interface that would allow an entrepreneur to get access to a D-Wave machine via the web). Instead, the point is that AI has been in a deceptive phase for the past fifty years, ever since 1956, when a bunch of top brains came together for the first time at the Dartmouth Summer Research Project
⁴⁴
and made a “spectacularly wrong prediction” about their ability to crack AI over a single hot New England summer. But today, couple the successes of Deep Learning and IBM's Watson to the near-term predictions of technology oracles like Ray Kurzweil, and we find a field reaching the knee of the exponential growth curve—that is, a field ready to run wild in disruption.

So what does this mean to you, the exponential entrepreneur? This is a multibillion-dollar question. But as you try to find answers, remember that JARVIS is essentially the ultimate user interface, democratizing every exponential technology and giving all of us access to Stark-like capabilities.

Camel racing is a centuries-old tradition in the Middle East, but it's an activity primarily reserved for large festivals. Yet, in the past half century, the sport has been transformed into both a mainstay of Arab culture—think the Kentucky Derby for sheikhs—and one of the richest sports on Earth. It's the jockeys who have changed the most. Twenty years ago, camels were ridden by children—the lightest possible riders—but general principle, injury, and death led to a humanitarian outcry. So both the UAE and Qatar banned the practice, instead replacing children with an even lighter saddle occupant—the robot jockey.
⁴⁵

Today, in camel racing, robot jockeys are the norm. Exactly like traditional jockeys, these robo-replacements sit on a saddle, steer with
the reins, and prod with a whip. To prevent the camels from being frightened by their cyborg occupants, designers found that humanlike features—a mannequin face, sunglasses, a hat, traditional racing silks, and even the traditional perfumes used by human jockeys—help keep the animals calm. The latest robot jockeys are small, about a foot high, and light, weighing between five and eight pounds, with skinny hinged arms that control the reins and whip. There's even a speaker on the robot so camel owners can issue commands to their animals via walkie-talkie as they follow along on an outside track (in air-conditioned SUVs).

Of course, our point isn't that there's a bevy of entrepreneurial possibility in camel racing. It's that robotics, another exponential technology long mired in deception, is now heading for disruption. According to a report by the Littler Workplace Policy Institute:
⁴⁶
“Robotics is the fastest growing industry in the world, poised to become the
largest
in the next decade.” Which is to say, robot jockeys are just the beginning.

Consider Baxter,
⁴⁷
the brainchild of legendary roboticist Rodney A. Brooks, Panasonic Professor of Robotics (emeritus) at MIT and cofounder of iRobot (creator of the Roomba). With a humanoid design, a nine-foot wingspan, and a tablet computer for a face, Baxter looks like something out of a cartoon. Grab one of his arms, for example, and Baxter will turn his head in your direction, the tablet computer displaying a pair of wide-open eyes to demonstrate interest. But what is most exciting about Baxter is his user interface.

Unlike most industrial robots, Baxter is human-safe. Getting in a room with a typical six-axis car-building robot is a good way to get dead—which explains why most industrial robots are cordoned off from humans. But Baxter doesn't need a cage. Sensors detect when the robot hits something unexpected and stops the motion immediately, so “he” can't hurt you.

Moreover, Baxter has an elegant and simple user interface. Instead of a complicated code-based programming, it learns through guided imitation. Simply move the robot's arms through the motions you want him to replicate, and presto, he's programmed. And with AI
soon coming online, it won't be long before putting Baxter through his motions will be replaced by simply having a conversation with him. “Hey, Baxter, could you put this tire on that car?”

“Baxter is a big step forward,” says Dr. Dan Barry,
⁴⁸
head of robotics at SU. “It's the first robot that bridges the gap between mindless, repetitive, robust, single-purpose industrial robots and intelligent, widely sensing, situationally aware, computationally complex, delicate research robots.” More important, Baxter is the kind of robot that entrepreneurs can now build businesses around. Case in point: Digital Apparel, a Bay Area clothing start-up, plans to do 3-D scans of their customer's bodies, then use those scans as a pattern for cutting and stitching denim to make perfect custom-fitting jeans. And what robot will Digital Apparel use to help assemble your jeans? You guessed it, Baxter.

Besides user-friendly robotic interfaces, we're also seeing exponential progress in robotic agility and mobility. Enabled by a new generation of sensors and actuators, and driven by near-infinite computing and artificial intelligence, there's a Cambrian explosion
⁴⁹
in robotics, with species of all sizes, shapes, and modes of mobility crawling out of the muck of the lab and onto the terra firma of the marketplace. Festo, for one example, has created a robot that flies like a bird. Boston Dynamics, for another, now makes robots that can climb, crawl, jump, and hop, and all while carrying heavy loads (some bots can manage over a hundred kilograms of weight). These “Sherpa-bots” can traverse boulder-strewn hillsides, balance on sheets of ice, and even jump from the ground to a rooftop three stories up.

But what has been relatively slow progress—run out of university labs and funded by government grants—took a quantum leap forward in late 2013, when Amazon announced it was going into the drone business
⁵⁰
and Google announced the acquisition of eight robotics companies (including Boston Dynamics).
⁵¹
With the big dogs in the game, progress is coming even faster.

And the resulting change will be considerable. Robots don't unionize, don't show up late, and don't take lunch, yet Baxter can work an assembly line for the equivalent of $4
an hour.
⁵²
A 2013 report from the Oxford Martin School concludes that 45 percent of American jobs are at high risk of being taken by computers (AI and robots) within the next two decades.
⁵³
Good or bad, this same trend is evident around the world. In China, Foxconn, the Chinese electronics manufacturer that builds Apple's iPhone, made news in 2013 when the skyrocketing demand for cell phones led to labor disputes, reports of harsh working conditions, even riots and suicides. In the aftermath of these reports, Foxconn's president, Terry Gou, said he intended to replace one million workers with robots over the next three years.
⁵⁴

Besides replacing our blue-collar workforce, over the next three to five years, robots will invade a much wider assortment of fields. “Already,” says Dan Barry, “we're seeing telepresence robots transport our eyes, ears, arms, and legs to conferences and meetings. Autonomous cars, which are, after all, just robots, will [start to] chauffeur people around and deliver goods and services. Over the next decade, robots will also move into health care, replacing doctors for routine surgeries and supplementing nurses for eldercare. If I were an exponential entrepreneur looking to create tremendous value, I'd look for those jobs that are least enjoyable for humans to do. . . . Given that the global market for unskilled labor is worth many trillions of dollars, I would say this is a huge opportunity.”

So how does an entrepreneur take advantage of this opportunity? Well, as a June 3, 2013, article in
Entrepreneur
explained:
⁵⁵
“Startup infrastructure dedicated to robotics is likewise emerging: hacker spaces (Robot Garden), accelerators (Robot Launchpad) and even a dedicated venture capital firm, New York City–based Grishin Robotics, founded last June by Russian internet entrepreneur Dmitry Grishin.” These advances are proof that a field long out of reach for most entrepreneurs is now open for business. “We're seeing very early indicators that this market is coming into fruition immediately,” says Jon Callaghan, a founding partner in the early-stage tech venture capital firm True Ventures, in that same
Entrepreneur
article. “It's super early, but it will hit very, very quickly, and we'll look back on 2013 . . . as a year for robotics coming into its own.”

Throughout the past few chapters, we've been examining exponentials poised to explode over the next three to five years and seeing how these technologies reinforce and empower one another—the rise of cloud computing enables more capable and ubiquitous AI, which in turn allows the average entrepreneur to program robots. To close this chapter, we're going to examine synthetic biology, a technology that's a little further out—say, five to ten years—but is still transitioning from deception to disruption.

And it's going to be a sizable disruption.

Synthetic biology
⁵⁶
is built around the idea that DNA is essentially software—nothing more than a four-letter code arranged in a specific order. Much like with computers, the code drives the machine. In biology, the order of the code governs the cell's manufacturing processes, instructing it to make specific proteins and such. But, as with all software, DNA can be reprogrammed. Nature's original code can be swapped out for new, human-written code. We can co-opt the machinery of life, telling it to produce—well, whatever we can think of.

In itself, this idea isn't new. With genetic engineering, we've been inserting a gene or two from this organism into that organism—such as taking the DNA that makes jellyfish glow and, as South Korean researchers did in 2007, inserting it into cats to create, you guessed it, glow-in-the-dark cats.
⁵⁷
The difference with synthetic biology is that it's not just individual letters being swapped out—it's whole genomes.

“Synthetic biology is essentially genetic engineering gone digital,” explains synthetic biologist and Autodesk distinguished researcher and S. U. professor Andrew Hessel.
⁵⁸
“Used to be all this stuff was done by hand in the lab, with enormous expense and high error rates. Today we manipulate DNA with computers, using programs that function much like word processors. Mix and match genetic code, spell and error check, shuffle bits around—it's becoming drag-and-drop easy.”

This increase in simplicity and accessibility has opened the door to
a wonderland of possibilities. New fuels, foods, medicines, construction materials, clothing fibers, and even new organisms are all in the offing. Everything we now manufacture industrially, we'll eventually be able to assemble biologically.