Thursday, November 19, 2009

Joel Garreau on radical evolution

Yesterday I heard Joel Garreau speak again at ASU, as part of a workshop on Plausibility put on by the Consortium for Science, Policy, and Outcomes (CSPO). I previously posted a summary of his talk back in August on the future of cities. This talk was based on his book, Radical Evolution: The Promise and Peril of Enhancing Our Minds, Our Bodies--and What It Means to Be Human.

Garreau was introduced by Paul Berman, Dean of the Sandra Day O'Connor School of Law at ASU, who also announced that Garreau will be joining the law school faculty beginning this spring, as the Lincoln Professor for Law, Culture, and Values.

He began by saying that we're at a turning point in history [has there ever been a time when we haven't thought that, though?], and he's going to present some possible scenarios for the next 2, 3, 5, 10, or 20 years, and that his book is a roadmap. The main feature of this turning point is that rather than transforming our environment, we'll be increasingly transforming ourselves, and we're the first species to take control of its own evolution, and it's happening now.

At some point in the not-too-distant future, he said, your kid may come home from school in tears about how he can't compete with the other kids who are more intelligent, more athletic, more attractive, more attentive, and so forth--because you haven't invested in the human enhancement technologies coming on the market. Your possible reactions will be to suck it up [somebody's still gotta do the dirty jobs in society?], remortgage the house again to make your kid competitive, or try to get the enhanced kids thrown out of school. What you can't do is ignore it.

He then asked people to raise their hands who could remember when things were still prevalent:
  • The Sony Walkman
  • When computer screens were black and white. (An audience member said "green and black!")
  • Rotary dial phones
  • Mimeograph machines and the smell of their fluid
  • Polio
This shows, he said, that we're going through a period of exponential change.

His talk then had a small amount of overlap with his previous talk, in his explanation of Moore's Law--that we've had 32 doublings of computer firepower since 1959, so that $1 of computing power is about 2 billion times more than it was then, and an iPhone has more computing power than all of NORAD had in 1965. Such doublings change our expectations of the future, so that the last 20 years isn't a guide to the next 20, but to the next 8; the last 50 years is a guide to the next 14. He pulled out a handkerchief and said this is essentially the sort of display we'll have in the future for reading a book or newspaper.

He then followed Ray Kurzweil in presenting some data points to argue that exponential change has been going on since the beginning of life on earth (see P.Z. Myers' "Singularly Silly Singularity" for a critique):

It took 400 million years (My) to go from organisms to mammals, and
  • 150My to monkeys
  • 30My to chimps
  • 16My to bipedality
  • 4My to cave paintings
  • 10,000 years to first settlements
  • 4,000 years to first writing
At this point, culture comes into the picture, which causes even more rapid change (a point also made by Daniel Dennett in his talk at ASU last February).
  • 4,000 years to Roman empire
  • 1800 years to industrial revolution
  • 100 years to first flight
  • 66 years to landing on the moon
And now we're in the information age, which Garreau identified as a third kind of evolution, engineered or radical evolution, where we're in control. [It seems to me that such planned changes are subject to the limits of human minds, unless we can build either AI or enhancement technologies that improve our minds, and I think the evidence for that possibility really has yet to be demonstrated--I see it as possible, but I place no bets on its probability and think there are reasons for skepticism.]

Garreau spent a year at DARPA (the Defense Advanced Research Projects Agency), the organization that invented the Internet (then the ARPANet), which is now in the business of creating better humans, better war fighters. [DARPA was also a subject of yesterday's Law, Science, and Technology class. It's a highly funded organization that doesn't accept grant proposals, rather, it seeks out people that it thinks are qualified to give funding to for its projects. It has become rather more secretive as a result of embarrassment about its Total Information Awareness and terrorism futures ideas that got negative press in 2003.]

Via DARPA, Garreau learned about their project at Duke University with an owl monkey named Belle, that he described as a monkey that can control physical objects at long distances with her mind. Belle was trained to play a video game with a joystick, initially for a juice reward and then because she enjoyed it. They then drilled a hole in her head and attached fine electrodes (single-unit recording electrodes like the sort used to discover mirror neurons), identified the active regions of her brain when she operated the joystick, and then disconnected the joystick. She became proficient and playing the game with the direct control of her brain. They then connected the system to a robotic arm at MIT which duplicated the movements of her arm with the joystick.

Why did they do this? Garreau said there's an official reason and a real reason. The official reason is that an F-35 jet fighter is difficult to control with a joystick, and wouldn't it be better to control it with your mind, and send information sensed by the equipment directly into the mind? The real reason is that the DARPA defense sciences office is run by Michael Goldblatt, whose daughter Gina Marie (who recently graduated from the University of Arizona) has cerebral palsy and is supposed to spend the rest of her life in a wheelchair. If machines can be controlled with the mind, machines in her legs could be controlled with her mind, and there's the possibility that she could walk.

Belle first moved the robotic arm 9 years ago, Garreau said, and this Christmas you'll be able to buy the first toy mind-machine interface from Mattel at Walmart for about $100. It's just a cheap EEG device and not much of a game--it lets you levitate a ping pong ball with your mind--but there's obviously more to come.

Garreau said that Matthew Nagel was the first person to send emails using his thoughts (back in 2006), and DARPA is interested in moving this technology out to people who want to control robots. [This, by the way, is the subject of the recent film "Sleep Dealer," which postulates a future in which labor is outsourced to robots operated by Mexicans, so that they can do work in the U.S. without immigrating.]

This exposure to DARPA was how Garreau got interested in these topics, which he called the GRIN technologies--Genetics, Robotics, Information science, and Nanotechnology, which he identified as technologies enabled by Moore's Law.

He showed a slide of Barry Bonds, and said that steroids are sort of a primitive first-generation human enhancement, and noted that the first uses of human enhancement tend to occur in sports and the military, areas where you have the most competition.

Garreau went over a few examples of each of the GRIN technologies that already exist or are likely on the way.

Genetics
Dolly the cloned sheep. "Manipulating and understanding life at the most primitive and basic level."

"Within three years, memory pills, originally aimed at Alzheimer's patients, will then move out to the needy well, like 78 million baby boomers who can't remember where they left their car, then out to the merely ambitious." He said there's already a $36.5 billion grey market for drugs like Ritalin and Provigil (midafonil), and asked, "Are our elite schools already filling up with the enhanced?" [There's some evidence, however, that the enhancement of cognitive function (as opposed to staying awake) is minimal for people who already operate at high ability, with the greatest enhancement effect for those who don't--i.e., it may have something of an egalitarian equalizing effect.]

He said DARPA is looking at ways to end the need for sleep--whales and dolphins don't sleep, or they'd drown, but they do something like sleeping with one half of the brain at a time.

DARPA is also looking at ways to turn off hunger signals. Special forces troops burn 12,000 calories per day, but can't carry huge amounts of food. The body carries extra calories in fat that are ordinarily inaccessible unless you're starving, at which point they get burned. If that switch to start burning fat could be turned on and off at will, that could be handy for military use. He observed that DARPA says "the civilian implications of this have not eluded us."

Sirtris Pharmaceuticals, started by David Sinclair of the Harvard Medical School, aims to have a drug to reverse aging based on resveratrol, an ingredient from grapes found in red wine. [Though Quackwatch offers some skepticism.]

Garreau looks forward to cures for obesity and addiction. He mentioned Craig Venter's plan to create an organism that "eats CO2 and poops gasoline" by the end of this year, that will simultaneously "end [the problems in] the Middle East and climate change." [That seems overly optimistic to me, but ExxonMobil has given Venter $600 million for this project.]

He said there are people at ASU in the hunt, trying to create life forms like this as well. [Though for some reason ASU doesn't participate in the iGEM synthetic biology competition.]

Robotics
Garreau showed a photo of a Predator drone, and said, "Ten years ago, flying robots were science fiction, now it's the only game in town for the Air Force." He said this is the first year that more Air Force personnel were being trained to operate drones than to be pilots. 2002 was the first year that a robot killed a human being, when a Predator drone launched a Hellfire missile to kill al Qaeda members in an SUV in Yemen. He said, "while there's still a human in the loop, philosophical discussions about homicidal robots could be seen as overly fine if you were one of the guys in the SUV."

"We're acquiring the superpowers of the 1930s comic book superheroes," he said, and went on to talk about a Berkeley exoskeleton that allows you to carry a 180-pound pack like it weighs four pounds, like Iron Man's suit. He asked the engineers who built it, "Could you leap over a tall building in a single bound?" They answered, "yes, but landing is still a problem."

Functional MRI (fMRI) is being used at the University of Pennsylvania to try to determine when people are lying. Garreau: "Then you're like the Shadow who knows what evil lurks in the hearts of men."

Cochlear implants to give hearing to people for whom hearing aids do nothing, connecting directly to the auditory nerve. Ocular implants to allow the blind to have some vision. Brain implants to improve memory and cognition. Garreau asked, "If you could buy an implant that would allow you to be fluent in Mandarin Chinese, would you do it?" About half the room raised their hands. [I didn't hear a price or safety information, so didn't raise my hand.]

Information
He showed a photo of a camera phone and said, "Fifteen years ago, a machine like this that can fit in your pocket, with a camera, GPS, and MP3 player, and can send email, was science fiction. Now it's a bottom-of-the-line $30 Nokia."

He asked, "Does anyone remember when music players were three big boxes that you put on your bookshelves? Now they're jewelry. Soon they'll be earrings, then implants."

Close behind, he said, are universal translators. "Google has pretty good universal translation on the web, and see it as moving out to their Droid phones." He observed that Sergey Brin was talking in 2004 about having all of the world's information directly attached to your brain, or having a version of Google on a chip implanted in your brain. [I won't get one unless they address network security issues...]

Nanotechnology
Garreau said, "Imagine anything you want, one atom or molecule at a time. Diamonds, molecularly accurate T-bone steaks." He said this is the least developed of the four GRIN technologies, "so you can say anything you want about it, it might be true." It's estimated to become a $1 trillion/year market in the next 10 years. There may be nanobots you can inject into your bloodstream by the thousands to monitor for things about to go wrong [see this video for the scenario I think he's describing], hunter-killers that kill cancer cells. "When you control matter at a fundamental level, you get a feedback loop between the [four] technologies."

At this point, Garreau said he's really not all that interested in the "boys and their toys" so much as he is the implications--"where does this take culture and society and values?" He presented three possible scenarios, emphasizing that he's not making predictions. He called his three scenarios Heaven, Hell, and Prevail.

Heaven
He showed a chart of an exponential curve going up (presumably something like technological capacity on the y axis and time on the x axis).

He said that at the NIH Institute on Aging, there's a bet that the first person to live to 150 is already alive today. He mentioned Ray Kurzweil, said that he pops 250 pills a day and is convinced that he's immortal, and is "not entirely nuts." [I am very skeptical that 250 pills a day is remotely sensible or useful.]

For the last 160 years, human life expectancy has increased at about 1/4 of a year every year. He asked us to imagine that that rate improves to one year per year, or more--at that point, "if you have a good medical plan you're effectively immortal." [I questioned this in the Q&A, below.]

Hell
He showed a chart that was an x-axis mirror of the Heaven one, and described this as a case where technology "gets into the hands of madmen or fools." He described the Australian mousepox incident, where researchers in Australia found a way to genetically alter mousepox so that it becomes 100% fatal, destroying the immune system, so that there's no possible vaccine or prevention. This was published in a paper available to anyone, and the same thing could be done to smallpox to wipe out human beings with no defense. He said the optimistic version is something that wipes out all human life; the pessimistic version is something that wipes out all life on earth. [In my law school class, we discussed this same topic yesterday in more detail, along with a similar U.S. paper that showed how to reconstruct the polio virus.]

The problem with both of these scenarios for Garreau is that they are both "techno-deterministic," assuming that technology is in control and we're "just along for the ride."

Prevail
He showed a chart that showed a line going in a wacky, twisty pattern. The y-axis may have been technological capacity of some sort, but the x-axis in this case couldn't have been time, unless there's time travel involved.

Garreau said, if you were in the Dark Ages, surrounding by marauding hordes and plagues, you'd think there wasn't a good future. But in 1450 came the printing press--"a new way of storing, sharing, collecting, and distributing information," which led to the Renaissance, enlightenment, science, democracy, etc. [Some of those things were rediscoveries of advancements previously made, as Richard Carrier has pointed out. And the up-and-down of this chart and example of the Dark Ages seems to be in tension, if not in conflict, with his earlier exponential curve, though perhaps it's just a matter of scale. At the very least, however, they are reason to doubt continued growth in the short term, as is our current economic climate.]

Garreau called the Prevail scenario more of a co-evolution scenario, where we face challenges hitting us in rapid succession, to which we quickly respond, which creates new challenges. He expressed skepticism of top-down organizations having any capacity to deal with such challenges, and instead suggested that bottom-up group behavior by humans not relying on leaders is where everything interesting will happen. He gave examples of eBay ("100 million people doing complex things without leaders"), YouTube ("no leaders there"), and Twitter ("I have no idea what it's good for, but if it flips out the Iranian government, I'm for it.") [These are all cases of bottom-up behavior facilitated by technologies that are operated by top-down corporations and subject to co-option by other top-down institutions in various ways. I'm not sure how good the YouTube example is considering that it is less profitable per dollar spent than Hulu--while some amateur content bubbles to the top and goes viral, there still seems to be more willingness to pay for professional content. Though it does get cheaper to produce professional content and there are amateurs that produce professional-quality content. And I'll probably offer to help him "get" Twitter.]

The Prevail scenario, he said, is "a bet on humans being surprising, coming together in unpredicted ways and being unpredictably clever."

He ended by asking, "Why have we been looking for intelligent life in the universe for decades with no success? I wonder if every intelligent species gets to the point where they start controlling their own destiny and what it means to be as good as they can get. What if everybody else has flunked. Let's not flunk. Thanks."

Q&A
I asked the first question, which was whether there is really so much grounds for optimism on extending human lifespan when our gains have increased the median lifespan but not made recent progress on the top end--the oldest woman in the world, Jeanne Calment, died at 122 in 1997 and no one else has reached that age. He answered that this was correct, that past improvements have come from nutrition, sanitation, reducing infant mortality, and so forth, but now that we spent $15 billion to sequence the first human genome and the cost of sequencing a complete human genome is approaching $1,000 and personalized medicine is coming along, he suspects we'll find the causes of aging and have the ability to reverse it through genetic engineering.

Prof. David Guston of CSPO asked "What's the relation between your Prevail scenario and the distribution of the success of the good stuff from GRIN technologies?" Looking at subgroups like males in post-Soviet Russia and adults in Africa, he said, things seem to be going in the wrong direction. Garreau answered that this is one of the nightmare scenarios--that humans split into multiple species, such as enhanced, naturals, and the rest. The enhanced are those that keep upgrading every six months. The naturals are those with access to enhancements that "choose not to indulge, like today's vegetarians who are so because of ethical or aesthetic reasons." The rest are those who don't have access to enhancements, and have envy for and despise those who do. "When you have more than one species competing for the same ecological niche," he said, "that ends up badly for somebody." But, he said, that's assuming a rich-get-richer, poor-get-poorer belief, "a hallmark of the industrial age." Suppose that instead of distributing scarcity, we are distributing abundance. He said that transplanted hearts haven't become cheap because they aren't abundant, but if we can create new organs in the body or in the lab in a manner that would benefit from mass production, it could become cheap. He pointed out that cell phones represent "the fastest update of technology in human history," going from zero to one phone for every two people in 26 years, and adapted to new uses in the developing world faster than in the developed world. He brought up the possibility of the developing world "leapfrogging" the developed world, "the way Europeans leapfrogged the Arab world a thousand years ago, when they were the leaders in science, math, and everything else." [I think this is a very interesting possibility--the lack of sunk costs in existing out-of-date infrastructure, the lack of stable, firmly established institutions are, I think, likely to make the developing world a chaotic experimental laboratory for emerging technologies.]

Prof. Gary Marchant of the Center for the Study of Law, Science, and Technology then said, "I'm worried about the bottom-up--it also gave us witch trials, Girls Gone Wild, and the Teabaggers." Garreau said his Prevail scenario shows "a shocking faith in human nature--a belief in millions of small miracles," but again said "I'm not predicting it, but I'm rooting for it."

Prof. Farzad Mahootian and Prof. Cynthia Selin of CSPO asked a pair of related questions about work on public deliberations and trying to extend decision-making to broader audiences, asking what Garreau thought about "DARPA driving this or being porous to any kind of public deliberation or extended decision-making?" Garreau responded that "The last thing in the world that I want to do is leave this up to DARPA. The Hell scenario could happen. Top-down hierarchical decision-making is too slow. Anyone waiting for the chairman of the House finance committe to save us is pathetic. Humans in general have been pulling ashes out of the fire by the skin of their teeth for quite a while; and Americans in particular have been at the forefront of change for 400 years and have cultural optimism about change." [I think these questions seemed to presuppose top-down thinking in a way that Garreau is challenging.]

He said he had reported a few years ago about the maquiladoras in Mexico and called it a "revolution," to which he got responses from Mexicans saying, "we're not very fond of revolutions, it was very messy and we didn't like it," and asking him to use a different word. By contrast, he said, "Americans view revolutions fondly, and think they're cool, and look forward to it." [Though there's also a strange conservatism that looks fondly upon a nonexistent ideal past here, as well.] With respect to governance, he said he's interested in looking for alternate forms of governance because "Washington D.C. can't conceivably respond fast enough. We've got a long way to go and a short time to get there." [Quoting the 'Smokey and the Bandit' theme song.]

He went on to say, "I don't necessarily think that all wisdom is based here in America. Other places will come up with dramatically different governance." He talked about the possibility of India, which wants to get cheaper drugs out to the masses, taking an approach different from FDA-style regulation (he called the FDA "a hopelessly dysfunctional organization that takes forever to produce abysmal results"). "Let's say the people of India were willing to accept a few casualties to produce a faster, better, cheaper cure for malaria, on the Microsoft model--get a 'good enough' version, send it out and see how many computers die. Suppose you did that with drugs, and were willing to accept 10,000 or 100,000 casualties if the payoff was curing malaria once and for all among a billion people. That would be an interesting development." By contrast, he said, "The French are convinced they can do it the opposite way, with top-down governance. Glad to see somebody's trying that. I'll be amazed if it works." His view, he said, was "try everything, see what sticks, and fast." [This has historically been the intent of the U.S. federal system, to allow the individual states to experiment with different rules to see what works before or in lieu of federal rules. Large corporations that operate across states, however, which have extensive lobbying power, push for federal regulations to pre-empt state rules, so that they don't have to deal with the complexity.]

There were a few more questions, one of which was whether anyone besides DARPA was doing things like this. Garreau said certainly, and pointed to both conventional pharmaceutical companies and startups working to try to cure addiction and obesity, as well as do memory enhancement, like Eric Kandel's Memory Pharmaceuticals. He talked about an Israeli company that has built a robotic arm which provides touch feedback, with the goal of being able to replace whatever functionality someone has lost, including abilities like throwing a Major League fastball or playing the piano professionally.

Prof. Selin reported a conversation she had with people at the law school about enhancement and whether it would affect application procedures. They indicated that it wouldn't, that enhancement was no different to them than giving piano lessons to children or their having the benefit of a good upbringing. Garreau commented that his latest client is the NFL, and observed that body building has already divided into two leagues, the tested and the untested. The tested have to be free of drugs, untested is anything goes. He asked, "can you imagine this bifurcation in other sports? How far back do you want to back out technology to get to 'natural'? Imagine a shoeless football league." He noted that one person suggested that football minus technology is rugby. [This reminded me of the old Saturday Night Live skit about the "All Drug Olympics."]

All-in-all, it was an interesting talk that had some overlap with things that I'm very interested in pursuing in my program, especially regarding top-down vs. bottom-up organizational structures. Afterward, I spoke briefly with Garreau about how bottom-up skeptical organizations are proliferating and top-down skeptical organizations are trying to capitalize on it, and I wondered to what extent the new creations of bottom-up organizations tend to get co-opted and controlled by top-down organizations in the end. In that regard, I also talked to him a bit about Robert Neuwirth's work on "shadow cities" and the Kowloon Walled City, where new forms of regulatory order arise in jurisdictional no-man's-lands (I could also have mentioned pirate codes). Those cases fall between the cracks for geographical reasons, while the cases that are occurring with regard to GRIN technologies fall between the cracks for temporal reasons, but it seems to me there's still the possibility of the old-style institutions to catch up and take control.

UPDATE: As a postscript, I recently listened to the episode of the Philosophy Bites podcast on human enhancement with philosopher Allen Buchanan, who was at the University of Arizona when I went to grad school there. Good stuff.

3 comments:

Joshua said...

I'd be very curious what he meant by Dark Ages and what time period he is talking about specifically. By most uses of that term, 1450 is well past the end of the Dark Age. The term is actually coined in the 1300s by Petrarch. IIRC he seems to see it as ending or have ended. If there was an actual Dark Age it would have been from about 600 or 700 or so around 1100 at most. That puts the printing press 350 years after the end of the Dark Ages.

Jim Lippard said...

Joshua: I don't think he meant that 1450 was in or immediately succeeded the Dark Ages, only that the Dark Ages was a time prior to the invention of the printing press when future progress didn't appear likely. Since he specifically mentioned the plague, he was no doubt referring to the 14th century, the era of the Decameron.

c2 said...

Thanks for that great synopsis. Now I don't have to read the book!