Monday, December 28, 2009

Singularity U Wrap-up

I submitted my last blog post after days 6 and 7, but SU is a 9-day program. What happened during the last two days? Sorry, I'm not at liberty to discuss it. Executive director Salim Ismail asked me not to write about the last two days, and I agreed not to do so without consulting first with the participants. There's no great mystery behind his request - just that the program ended by exploring some controversial issues, and Ismail didn't want people to feel inhibited by the presence of a reporter in the room.

Since the Executive Program ended, I've been pitching magazines on a feature article about it. No takers so far - fingers crossed - but for the time being, I'm saving further thoughts for publication.

Suffice it to say that the SU Exec Program was an energizing week-plus full of fascinating information, bold ideas, and warm comaraderie. The people, especially, were great, from faculty to administration to the students themselves, who were, to a one, bright, ambitious, principled, and tremendous fun to hang out with. I hope to see them all again before too many centuries pass.

Singularity U Days 5 & 6: Magical Mystery Tour

After four days in a drab classroom absorbing insights and information from the frontiers of science and technology, the Singularity University Executive Program busts out for a succession of field trips to real, live businesses. At their best, these places are harnessing rapidly evolving technology to a revenue model in anticipation of tectonic social and economic changes.

TechShop, an open-access DIY machine shop and fab studio in Menlo Park, California. The business model couldn’t be simpler: selling memberships like a health club. For $99 a month (plus a little more for training), you get access to lathes, mills, routers, sewing machines, welding stations, paint booths, cutters, 3D printers, and anything else you might need to produce a physical product of just about any spec.

TechShop takes advantage of the falling price of fabrication — still industrial-grade, but less so as the price of a $200,000 computer-controlled lathe has fallen to $40,000. As for members, they’re perfectly positioned to serve the long-tail market for niche goods, selling through the Internet’s global channel. One member, Roy Sandberg, runs a telepresence robotics company for $250/month plus materials, providing remote-control home care in the Netherlands.

Sandberg’s operation is competition for AnyBots, a manufacturer housed in the Y Combinator complex in Mountain View, California. AnyBots’ debut product will be a remote-controlled robot ($10,000). Singularity U students get a virtual tour of the company’s offices, watching a projection screen as an operator drives a bot from room to room. The bot is dependent on the operator, but it’s smart enough to roll straight down a hallway even if it get steered into the wall.

“It does what you want it to do, not what you tell it to do,” a tech tells me. When the contraption suddenly trundles around the corner into the room where we’re sitting, the crowd spontaneously emits a collective “Awww!” This thing is super cute. It looks like skinny Segway with a nerdy face, stereoscopic cameras peering like two eyes, a screen perched atop its head like a cap. This design is a shoo-in for day care centers; the corporate market might require a redesign.

If so, Ideo would be perfect for the job. Sprawling over a block in dowtown Palo Alto, the designers behind Ford’s hybrid vehicle’s dashboard and Kraft Foods’ supply chain practice a holistic approach that encompasses the entire product/process ecosystem. Outlining the company’s approach, Director of Technology Dave Blakely recalls a contract with NASA to make space suits less constricting.

“They’re always going to be stiff,” he says, “so the improvement was to let astronauts spend less time in them” by designing an enclosed lunar rover. Blakely sketches out a three-stage approach: Immerse yourself in the user’s world view, visulalize the solution, and prototype as quickly and simply as possible. Then repeat, and repeat again. The same steps apply, he says, whether you’re designing hardware or government policy.

Halcyon Molecular has taken holistic design to heart: Not only the product but the financial strategy, business model, and even the corporate offices are cleverly integrated to push genomics to the next level. It’s biotech innovator as garage startup — literally, with an electron microscope, on loan from the US Energy Department, among the other hardware packed into a four-car space. Led by PayPal cofounder Luke Nosek, the staff lives and works in a McMansion in Los Altos Hills, California, complete with swimming pool, jacuzzi, and an airy living room full of beanbag chairs.

Twenty-something molecular biologists stare intently at their screens, oblivious to the student brigade. The Singularity U visit is a preview — everyone signed an NDA before descending on a buffet of California cuisine, so you won’t find details here. Suffice it to say, as does pulse2.com, “Halcyon plans to sequence complete human genomes in less than 10 minutes at a cost of $100.”

After a final hors d’oeuvre and long look at the stream rising from the Halcyon swimming pool, the students pile back into the bus to imagine how they might anticipate and facilitate the world to come.

Singularity U Day 6: How to Build a Molecular Machine

A Singularity University student taped a picture of Ralph Merkle to a dorm room wall, executive director Salim Ismail tells me. Merkle makes an unlikely pinup, but I get it. He’s a great speaker, engaging students with unusual enthusiasm, clarity, and humor. It doesn’t hurt that what he’s explaining is one of the weirdest, scariest, most promising technologies on the horizon: molecular machines.

There’s some question about whether the concept of a minuscule machine assembled one atom at a time is even physically viable, but Merkle makes it palpable. In today’s session, he details a few gizmos designed by his collegue Robert Frietas, a research associate at the Institute for Molecular Manufacturing. Each one is a complex biomedical device designed to float freely in the human bloodstream, fixing problems that make doctors tear their hair out.

Take the respirocyte. “It’s like a tiny scuba tank,” Merkle says, a one-micron container filled with compressed oxygen. Essentially, it’s an artificial red blood cell, except that it would have 100 times a red blood cell’s carrying capacity. “Today your response to a heart attack would be, ‘I’ve got a heart att—’. With respirocytes in your bloodstream, your response is, ‘I’ve got a heart attack. I’ve got an hour or so to get myself into the hands of the emergency medical care system or I’m in big trouble.’”

You’d be in even bigger trouble if your immune system viewed nanobots as a threat, but Merkle doesn’t worry about that. His favorite material, diamond, is biochemically “pretty inert.” As for structures that can’t be produced out of carbon “you just have to design them so the immune system finds them uninteresting,” he says. “If you find a feature that excites the immune system, you just put some fuzz on it.”

Ultimately, he envisions fleets of fuzzy machines floating through the body, replacing DNA, repairing damage, delivering drugs, then being flushed out when their job is done. Once patients experience the benefits, they won’t be any more resistant to nanobots than they are to vaccines, fluorescent dyes, medicated stents, and the other futuristic preparations doctors inject into patients. How far in the future? With adequate funding and good luck, he says, 20 years; without, 40.

Singularity U Day 4: The Internet of Things

What comes after Web 2.0? For David Orban, the next phase is the Internet of Things: a digital lattice of interconnected objects — cars, handbags, sneakers, thermoses. Orban calls these objects spimes. Coined by sci-fi god Bruce Sterling, the term denotes a networked thingy that’s aware of its orientation in space and time. Your cellphone is a spime. A Roomba vacuum cleaner is another. Orban’s company WideTag is cranking out spime-ish gizmos and iPhone apps.

Location- and time-aware devices would be a lot more autonomous — they would take care of themselves rather than making you take care of them. Roomba already plugs itself in when it’s thirsty. A cell phone could go a lot farther toward making sure it had enough juice. Orban hopes the OpenSpime standard will lead the way to rapid proliferation. The danger is that connecting scads of spimes will max out the Internet’s capacity for connections. No problem: the next-gen Internet Protocol v6 is designed to accommodate 1,000 nodes for every person on earth.

If that sounds like overkill, consider claytronics, an initiative to make programmable matter. A lump of claytronics comprizes zillions of tiny programmable spheres, currently 1mm in diameter, eventually 1 micron. At that tiny scale, van der waals forces would bind the spheres into a putty that would hold any contour you impose on it. You’d be able to form the stuff by hand into any shape and determine the color, texture, and other characteristics. And, of course, a claytronic object will be able to change its form and character according to remote commands.

At Orban’s invitation, the class breaks into groups, each charged with conceiving its own new-breed spime. My group dreams up SPORE — Space Projects Offworld Resource Explorer — an autonomous botnet that roams the asteroid belt in search of mining opportunities. Other concepts: a P2P Teddy Bear and the Enlightenment spime cloud, which envelops the user in virtual reality.

The proposals get a little silly, but Orban reminds us that Vint Cerf himself, on Google’s official blog, wrote about delivery of soap to Internet-enabled washing machines. “If Vint Cerf can say something like,” he tells the class, “it’s okay to think crazy.”

Singularity U Day 4: Coping with Climate Change

It takes a lot of juice to power the contemporary lifestyle. How to keep the AC going without turning up the burner under Hothouse Earth?

That’s probably the wrong question, says Michel Gelobter, Environment & Energy Track Chair at Singularity University. A better one would be, How much more do we have to mess up the planet to get to a state where we’re not messing it up anymore?

The meat of Gelobter’s presentation is a primer on the laws of thermodynamics and how to use them to determine the best path to sustainable energy. He points out that Newton’s laws, for all their immutable truth, are nicely formulated for the sake of human utility. Take the first law, conservation of energy. If energy never disappears, why do we care about it? Because we need it to be available. At the moment, we’re transferring preserved sunlight into the atmosphere, where it causes problems that require further energy (or a gargantuan decrease in energy consumption) to fix. We need to engineer a more efficient system.

Ultimately, he says, only three kinds of efficiency matter, all derived from thermodynamics, and they all have to do with theoretical maximum efficiency. (For the technically minded, the options are Thermal, Carnot, and Second Law efficiency.) For example, in theory, you can roughly double coal plant efficiently by burning it hotter — an opportunity that many businesses are trying to crack.

On the other hand, the closer you get to theoretical maximum, the harder it is to close the gap. So a smart strategy is to look for technologies that yield the greatest efficiency from the get-go. “That’s the beauty of the laws of thermodynamics,” Gelobter says. They give you a quick way to short-cut the political and economic complexities of the energy business and find out whether you’re operating from a solid foundation in terms of physics. With physics on your side, you’ve got a fighting chance.

Singularity U Day 3: Neuroscience

The greatest mysteries yield the biggest opportunities. And for Christopher deCharms, the human brain is the most mysterious thing of all.

A neuroscientist specializing in real-time brain imagery, deCharms suggests that the next wave of knowledge, technology and business will come from cracking the code that gives humans the capacity to perceive, think and act.

He flashes a slide on the screen listing a dozen things we don’t know: How does the brain make choices? Predictions? Plans? How does it produce an impression of identity, of experience? How does it adjust to change? How do we see, hear, touch, taste, smell? Why do we feel motivated one moment, depressed the next? Why do we sleep?

One thing that makes the answers so elusive is the staggering complexity of what the brain does. It’s such a thicket that scientists and philosophers can’t even reach consensus on a definition of consciousness. The quickest route to answers, deCharms says, is to break down the problem into manageable pieces. He differentiates between brain functions that involve high information density — say, reading and writing to the visual cortex — and those involving very small amounts of information, like moderating pain.

“Neurotechnology may benefit from questioning what kinds of low-information-content signals we can read and write before we try to upload and download consciousness,” he says.

Case in point: Deep brain stimulation. DeCharms shows video clips of Parkinsons patients moving involuntarily in a jerky, repetitive, exhausting dance. Their ability to control motion is so disrupted they speak in gasps. Switch on the electrodes reaching into the motor cortex, and suddenly they stand still and start talking about how good it feels.

The rest of deCharms’ presentation is devoted to groundbreaking research in brain cartography, perceptual function, neuronal physiology and several ways to mediate brain activity from drugs to biofeedback. Still, he’s circumspect about the prospect of rapid advance in practical developments.

“If this research follows the usual pattern, progress will take longer than we imagine — but when it happens, it will deliver more benefit than we can imagine,” he says. The bottom line is that the frontier has been breached and wave after wave of troops are flooding over the border, mapping the territory, reshaping it, bringing new capabilities, hopes and challenges.

The mystery won’t remain a mysterious for long.

Singularity U Day 3: The Eye of the Hurricane

The Singularity University routine is nonstop. Breakfast at 7:30 am — good, wholesome food, regrettably low on sugary and fatty goodness (presumably consonant with Ray Kurzweil's life-extension regime) — followed by a series of deep-dive lectures in delivered in mind-boggling 90-minute blocks.

Lunch at 12:30 pm isn’t a time to recharge; it’s an opportunity to deliver more information … The first day, it was presentations by grads of last summer’s nine-week program, detailing the businesses they’ve founded since (SU gets one percent) … Yesterday it was a close look at the Tesla electric car parked in the driveway, with a company rep on hand to answer questions (and presumably to take orders). Today, it’s a detailed demo of the SU spinoff that’s farthest along, the Gettaround car-sharing service. Then it’s back to the lecture room to get your brain stuffed anew.

Dinner at 7 pm is a bit more leisurely, but afterward come more lectures and demos. Last night, executive director Salim Ismail’s discourse on metaphysics lasted until 11:30 pm. Get some sleep, rinse, and repeat.

Salim tells me I’ll get a chance to recharge when the lecture portion of the program ends and days are filled with field trips to Silicon Valley businesses — but somehow I doubt it.

Singularity U Day 2: After Hours with Peter Diamandis


After dinner on day two of Singularity University, Peter Diamandis gives a fantastic presentation about the X-Prize and what it means. This is a guy who radiates energy, seriousness, and goodwill. He would have made a first-class motivational speaker, but he’s focused on substantial issues and favors leather jackets over sharkskin suits.

The man doesn’t think small: “My mission on this planet is to be an agent of transformation for the human race moving beyond planet earth,” he says. Which would seem pompous if he didn’t walk the walk. Over 90 minutes, he shows the assembled SU students how it’s done. “if you don’t do what you love,” he counsels, “do something else. Because all of it is hard.”

Diamandis’ approach is to imagine the world as he would like it to be, chart a course to that destination, and identify the first possible stop along the way. If the larger aim is to diversify humanity’s habitats in the universe, then a first logical step is the Ansari X-Prize for private space flight. (That’s not the first step, actually, but the third or fourth — but it makes the point.)

The scope of his accomplishment becomes obvious as he runs down the list of results: “We brought a new industry to market, made the front page of Google, generated 6 billion media impressions, and ended up with the winning vehicle hanging from the ceiling of the Smithsonian Air and Space Museum — right above Charles Lindburgh’s Spirit of St Louis. So cool.”

To have an impact, he says, prizes need to attract the right contestants, set a clear, doable goal, make a huge splash, and allow the contest organizer to retain media rights. Do it right, and you’ll attract capital to the problem you’re hoping to solve, pool brainpower worldwide, make heroes of the winners, and change the world’s perception of what is possible. And he’s ready to do it many times over, with competitions underway to revolutionize transportation, genomics, medicine, bionics, brain-computer interfacing, and undersea exploration. The latest prize, for a lunar lander design, was awarded only a couple of weeks ago.

It’s in the activities of another of his enterprises, Zero G, that the human dimension of his work shines brightest. A while back, Stephen Hawking asked for a ticket. “I was told I would kill this guy and ruin my company,” Diamandis says, but he wasn’t going to pass up the chance to introduce the world’s expert on gravity to zero gravity.

It took six months to get the FAA to drop its requirement that passengers be able-bodied, and when the wheelchair bound astrophysicist was finally aloft, he cajoled the flight crew into flying not one, not two, but eight parabolas. The scientist is paralyzed except for a few muscles in his face, Diamandis says. Yet in the photos Diamandis flashes on the screen behind him, his customer wears a child’s wide-eyed grin.

SU Exec Program Day 2: Biotech & Bioinformatics

Subsequent to writing up Ralph Merkle's initial lectures, I was so harried that I satisfied myself with sending posts to Wired's Epicenter blog and dispensed with trying to put up everything on Blogger. However, I'd like this blog to represent the complete Singularity U experience, so I'm re-posting the Wired posts here, in order (hopefully). We pick up the thread with Andrew Hessel:


Biotech is a hidebound industry. Elephantine budgets. Glacial development timelines. Stultifying regulatory oversight. Pitiful productivity. Andrew Hessell, once employed by a major biotech drug developer (unnamed) and now an evangelist of synthetic biology, observes that, while biotech and IT share important characteristics — both had similar inception dates and depend on information — they’ve followed completely different trajectories.

The reasons, he says, boil down to greed and fear. Where the captains of IT formed partnerships, forged standards, and opened their source code (however reluctantly), biotech moguls protected their turf tooth and claw, and still do.

Now biotech’s dark lords face a digital-age army, people raised with a networked mindset, bent on taking over the territory. And that, in Hessel’s view, makes biotech the next IT.

The price of gene sequencing is falling precipitously; the $1000 human genome is on the horizon. Inexpensive tools are becoming available, such as LavaAmp, a $10 gene amplifier, currently in prototype. Meanwhile, molecular biologists are beginning to understand how to engineer processes like photosynthesis and sugar metabolism.

MIT’s BioBricks program is recruiting hundreds of bright students every year, teaching them how to create synthetic organisms by snapping together DNA components like Legos. Student teams are engineering new industrial processes and programming VR biotech training. And the DIYbio movement is gaining momentum, poised to make end runs around industry and government. “This isn’t heavy programming,” Hessel says. “More like writing a script for Excel.”

The reward? Multibillion-dollar opportunities removing bottlenecks from the current system. People need drugs. They value health. They will pay for a new generation of medicines, diagnostics, protective measures, tests and measurements. Make biotech more like infotech, Hessel says, and it will happen.