January-August 2003

Newsletter of the Acceleration Studies Foundation, a nonprofit corporation.

Part 1  Part 2   Part 3   Parts 4-5   Parts 6-7

3. Universal Dialog: Science and Technology
(esp. Bio. Sciences, Cog. Sciences, Engineering, Computer Science, Complexity, and Physical Sciences))

13. Prime Examples of Pure Surprise!
Advances in the Four Accelerating Meta-Trends of…
     14. Intelligence (Science and Systems Theory; I.A.: Intelligence Amplification; A.I.: Autonomous Intelligence)
     15. Interdependence (Developmental Ethics, Accelerating Compassion, Globalization, Social Justice)
     16. Immunity (Biological, Cultural, and Technological Immune Systems)
     17. Matter, Energy, Space, and Time (MEST) Compression or Efficiency (Miniaturization, Efficiency, Computational
            Acceleration, Information Exponentiation)
18. Visionary and Vaporware Sci-Tech (Prototype or Hype? You Decide)
19. Tools for Modern Living

Universal Dialog: Science and Technology
(esp. Bio. Sciences, Cog. Sciences, Engineering, Computer Science, Complexity, and Physical Sciences)

13. Prime Examples of Pure Surprise!

Apparent Strong Limitations of 21st Century Biotechnology: Clues to the Electronic Future of Local Intelligence, John Smart, Acceleration Watch, 2003
This short essay was inspired by one of the many intellectual dividends of our human genome project, our recent discovery of the amazing genetic similarity of all human beings. Apparently, there is more genetic diversity between randomly chosen members of one troop of baboons than between randomly chosen members of the entire human species. Wow. I read a review of the following paper, "Features of Evolution and Expansion of Modern Humans, Inferred from Genomewide Microsatellite Markers," Zhivotovsky et. al, May 2003, AJHG, that offers a catastrophic explanation for this similarity, and realized that they had an implausible explanation for the data. As most biologists are neo-Darwinists, not yet operating within the paradigm of evolutionary development, they will frequently miss developmentalist explanations, such as the alternative explanation I propose: path dependency as the most parsimonious explanation for human homogeneity. This essay also makes more general points about the strong limitations on making changes in complex, highly differentiated biological systems like human beings. These are constraints that most futurists, and even many biologists do not yet fully appreciate. Let me know your thoughts!

Asimo Assimilation / Robotic Nation.

Very Cute Quicktime Video: Robodex 2003 (dancing miniature robots)

This first picture above was an ad that ran in a major American magazine (the name escapes me) on Jan 1, 2003. Amazingly effective! That might have been the first U.S. announcement that Bicentennial Man is coming to town. Expect this to be a reality within a generation. Asimov would likely have been impressed (no relation, we're told) with the cute little techno-humanoid that evokes his name. At the first debut in Europe this July, ASIMO demonstrated early versions of new "intelligence technologies": "the ability to recognise people, objects and gestures, calculate distances and the direction of movement of several objects. These give ASIMO a range of abilities including approaching persons, following them, moving in the direction they indicate and even recognising their faces so as to address them by name. ASIMO can also identify voices and respond to certain instructions." Also from the Honda Site: "ASIMO's size, shape and weight are very deliberate. ASIMO weighs just 115 pounds (52 kilograms), light enough to be easily picked up and moved, if necessary. And at 4 feet (120 centimeters) tall, ASIMO is not only the perfect size for switching the lights on and off, working at a table, or opening and closing doors, but ASIMO's camera eyes are also at the same level as an adult sitting in a chair or in a bed, for easy communication."

For an insightful but dystopian series of articles on the future of robotics, read Marshall Brain's Robotic Nation essays. These give you an excellent sense of just how fast our planet's technological intelligence is developing. But be aware that Brain (How Stuff Works) makes some slightly off-kilter conclusions from this data, proposing that "robots will take your job" circa 2030-2050. To a significant degree he's right, but with a lot less disruptiveness than he's thinking. Let's be accurate: robots have already taken the job of the first world workforce, and we like it that way. Virtually all of that developed world paycheck that we take home each week isn't paid for by our own sweat and mental creativity, but as a direct result of the accelerating productivity of the technologies we've put in motion in our manufacturing, services, and information sector industries.

More and more, we receive our paychecks in order to be educated about all the exponentiating creative options we have within our particular clusters of technologies and resource bases (we call these "companies"), and then to try to figure out in a surprisingly short period of time (we call this "opportunity") how to exercise our technological (and to an ever-declining degree, biological) creative options cleverly enough so that someone else playing with machines in similar-looking concrete boxes (we call these "buildings") on the other side of the city doesn't take our jobs. If they do, that is generally healthy too (we call that "competition", or "creative destruction") just as long as too many of us in the "workforce" don't get displaced too quickly by the incessant waves of technological innovation. No, in 2050 we'll be "working" (read: learning, creating, and playing) just as hard as today. In other words, young hominid adults will continue to put in biologically unhealthy hours at "work" in order to climb the current social status ladder, because that is what we hominids do. And they'll be surfing the crest of an even more more powerful wave of semi-autonomous robots. The more things change, the more some things stay the same.

Nevertheless, soon after 2050, in my own current rough estimation, computational autonomy will finally become such a powerful local force that it will run away, in a very short time frame, to the emergence of systems that are able to exceed biological complexity in every domain we care about. But even then don't expect robots to take human jobs—they really wouldn't be interested. In fact, relatively soon thereafter it seems likely that continuing to learn, create, and play as a biological human is just not going to be the most interesting game in town.

Genetic Programming IV: Routine Human-Competitive Machine Intelligence, John Koza, 2003
John Koza's latest tour de force, describing his results with genetic programming, one of the leading evolutionary software methods. From the book description: "Genetic programming (GP) is a method for automatically creating computer programs. It starts from a high-level statement of what needs to be done and uses the Darwinian principle of natural selection to breed a population of improving programs over many generations. Genetic Programming IV: Routine Human-Competitive Machine Intelligence presents the application of GP to a wide variety of problems involving automated synthesis of controllers, circuits, antennas, genetic networks, and metabolic pathways. The book describes fifteen instances where GP has created an entity that either infringes or duplicates the functionality of a previously patented 20th-century invention, six instances where it has done the same with respect to post-2000 patented inventions, two instances where GP has created a patentable new invention, and thirteen other human-competitive results. The book additionally establishes: GP now delivers routine human-competitive machine intelligence. GP is an automated invention machine. GP can create general solutions to problems in the form of parameterized topologies. GP has delivered qualitatively more substantial results in synchrony with the relentless iteration of Moore's Law." We are honored to have John Koza as a speaker at ACC2003.

Motion Sensor Nears Quantum Limit, Technology Review, 8.12.03
Single Electron Transistor (SET) plus a nano-cantilever can now measure movements a thousandth of a nanometer in dimension. These systems are now within a factor of 100 of empirically testing Heisenberg's Uncertainty Principle. Computational closure in motion sensing, at the bottom end of the universal material domain, is now within sight! We have only begun to consider the implications of this development for the future trajectory of computation. Thanks to Alex Lightman.

14. Intelligence Advances (Science and Systems Theory; I.A.: Intelligence Amplification; A.I.: Autonomous Intelligence)

Robotic “William” and “Mary” Cruise Virginia Campus, USA Today, 8.14.03
Student hobbyist robots are falling under the $1,000 price range (laptop not included). We are witnessing a profusion of tinkering with these systems, a very fortuitous development. Over the next 20 years, our robots will learn to see, to smell (Cyranose), to touch, to balance (Segway), and to talk (computer telephony, CUI) among other new capabilities. We are beginning to tell the story of the revolutionary effects of their ability to talk. But as they gain in their sensorymotor capacities as well, industrial robots will also become increasingly powerful, and must begin to have more distributed, personal, consumer applications in less controlled environments, perhaps first as intelligent appliances. There are plenty of industrial examples of the efficiencies resulting from primitive robotic sensorimotor skills, but there will come a time (after 2010?) when more distributed applications must become available. To my knowledge no one has envisioned that near-term story clearly yet, but I think someone will soon.

15. Interdependence Advances (Developmental Ethics, Accelerating Compassion, Globalization, Social Justice)

Customer-Owned Networks (Future of VoIP and Wireless), Clay Shirky, 1.7.03
Clay Shirky does a great job here of outlining the nature of the transition we are heading for in telephone and wireless communications. Tomorrow's dominant communications networks are being built out day by day by us, the customers, buying (at present) Cisco ATA converters for Vonage.com and other voice-over-IP telephone providers, or 802.11b cards and ports (and their mesh-network successors) for wireless access. In each case, Bob Metcalf's network effect (what we saw with fax machines, and with MP3 downloads) has become the new economics.

A certain amount of development on the central infrastructure is necessary to create a minimum quality (minimum sound quality with MP3, minimum resolution with fax machines, minimum voice quality with VoIP) and then the main capital investment occurs at the edges of the network, invested by the individual users, not by the centralized companies (Fed Ex, Telco's, etc.). This kind of economics wasn't possible before, because our distributed computational systems weren't smart enough to do this, until recently. The more powerful computational systems become, the more we will move toward a more biologically inspired system with centralized and relatively "stupid" central infrastructure (David Isenberg's stupid network model) where most computation occurs at the peripheries of the network, between its massively distributed intelligent nodes. In other words, like human social/linguistic/swarm computation (the relatively "stupid" network), vs. individual human consciousness (the intelligent node). Telco's have given us a stupid network for a long time. It's about time we added some interdependent intelligent nodes in our technological evolutionary development.

Get ready for the telcos to start buying these VoIP companies, as they are forced to move to this new, far lower priced business model for basic services. Premium services, of course, like broadband and intelligent interfaces to your data (think Wildfire and her proto-CUI successors) can be expected to keep the surviving telcos cash-flush and happy all the way to the singularity. Thanks to Johann Gevers.

SocialText, Social Software Solutions
Social software is an accelerating new category of tools that includes online communities (enhanced BB's, forums with chat), Wikis and Blogs, and an always-evolving set of internet-enhanced collaborative tools. These systems emphasize usability and productivity: what people really use and what actually gets done, versus what the company org charts and strategic plans say. They've got some simple and powerfully self-organizing systems in use, including a conference Wiki for Ester Dyson's PC Forum. We are honored to have Ross Mayfield, CEO of Socialtext, as a speaker at ACC2003.

(Statistical Machine Translation Advances) "From Uzbek to Klingon, The Machine Cracks the Code," Christopher Farah, NYT, 7.31.2003
A brief article on the steady advances in automatic machine translation over the last four years. In combination with increasingly larger language databases (the web) and increasingly powerful processors, a simple statistical tool like the n-gram "allows scientists to develop machine translation systems for a wide number of obscure languages at a pace that experts once thought impossible." The universal translator inches closer, tying us all, including all the dead cultures and forgotten languages (Don’t remember the Etruscans? Your children will!) into one tightly knit, incredibly resilient weave in coming decades. I can't wait!

16. Immunity Advances (Biological, Cultural, and Technological Immune Systems)

Depleted Uranium Weapons: The Whys and Wherefores (PDF), Andre Gsponer, Jan 2003
A helpful study showing that one of the few remaining valid uses for depleted uranium, armor piercing rounds, is being steadily outcompeted by tungsten alloys. DU now has only a 10% advantage. As we move steadily toward a nuclear-minimized world, the bad PR of using these "nuclear" materials in modern arsenals just isn't worth it anymore, if we can get around it any other way. Thanks to Wayne Radinsky.

Applicant Live Scan Centers Quietly Expand (DOJ's Electronic Fingerprinting System)

Back in 1997, California's Department of Justice decided to move beyond manual fingerprinting cards into an electronic scanning system that allows agencies much more rapid file sharing and background checks. Here's a PDF on the technology, called Applicant Live Scan. An increasing number of employers are using this system during hiring, to check criminal backgrounds.

Citizen privacy and confidentiality is of course part of the process, with civil and criminal prosecution allowed for breach, and such reports cannot be retained after the hiring decision is made. Some of the privacy info can be found here. This PDF shows that the number of Live Scan centers in California have quietly expanded from 120 in 2000 to about 300 today.

Live Scan centers are presently primarily at Police and Sheriff's Departments, Education and Public Safety Departments, and Sylvan Identix locations. That will change. It would be nice to see this rollout greatly accelerated across the country, with the input (and expected opposition) of all the privacy groups, of course. Unfortunately, the DOJ still accepts manual fingerprint cards, which is a backpedaling from their original position. This seems to be a small but important step toward better social immunity, but still too focused on fear rather than pride.

From my perspective, I'd like to see Live Scan centers required to be stocked with information on community service and civics opportunities. Let's lead with vision, not a negative model. We don't yet require everyone in this country to take civics class, or to do a brief community-building stint in either AmeriCorps or Military service as part of their socialization. Perhaps it's time to rethink that, in a world of accelerating change? We'll see.

17. Matter, Energy, Space, and Time Compression or Efficiency Advances (Miniaturization, Efficiency, Computational Acceleration, Information Exponentiation)

The Frontier of Inner Space: UCLA's Institute for Cell Mimetic Space Exploration (CMISE) and
University of Michigan's New Program for Intracellular Exploration, Feb 2003
Here's an amazing career path for any who have a passion for the small. See UCLA's new NASA-funded program, co-Directed by Carlo Montemagno. Then take a look at U. Michigan's new $2M project to extract views of the inner workings of cells. In realtime. In 3D. Synthetic nanoprobes to measure moment-to-moment ionic activity. Sophisticated statistical modelling programs to interpret the explosion of new data we are collecting. Another big step toward beyond the genome, beyond even the proteome, toward building the metabolome, a realtime 3D model of what occurs inside a cell. Grappling with the incredibly complicated, incredibly important protein translation and messaging story. Understanding how noise is useful to emergence (stochastic resonance), and how microscopic evolutionary chaos, under specially tuned conditions, leads highly reliably to macroscopic developmental order.

A humble prediction: We are going to discover things in Inner Space that are stranger and more valuable than we presently imagine, and we will use these new insights to construct the next generation of biologically-inspired evolutionary developmental computation systems. Awesome!

Integrated Design Flow, The Magma Electronic Design Automation (EDA) Story

EDA is a breathtaking example of MEST compression/efficiency and autonomy. Today's layout simulation systems are allowing us to create increasingly complex, multi-million gate logic circuits and comprehensive systems on a chip (SoC's) much faster and cheaper, per circuit, than ever before.

Magma is a relatively new company in the EDA chip design space. They IPO'd in 2001 and have been running at a loss since inception, but are nearing breakeven. They've acquired about a dozen blue-chip customers, such NEC, TI, and Vitesse. Magma's innovation is EDA software that mixes both the "logical" (front end) and "physical" (back end) layers in the same data model. This approach, which those in the industry call "integrated design flow," requires extremely powerful computers because the whole layout has to be in memory concurrently—an amazing feat. The basic idea is to reduce the amount of time it takes to design a chip by predicting the timing more accurately, thus reducing the number of "iterations" that have to be done to get the layout of the logic gates. Their approach allows them to get these iterations down to as short a time as six weeks to spec out a five million gate chip. A six week development cycle for a new technological organism? That's impressive. What will the generation times be in 2013?

As futurist Wayne Radinsky reminds us, "layout simulation has become more critical because, as the feature size of semiconductors has gotten smaller, quantum effects make the timing harder to predict. Whether Magma will make it to profitability remains to be seen. The gorillas of the design automation industry, Cadence Design Systems, Synopsys, and Avant! which was recently acquired by Synopsys, have been making products that do similar things. Cadence has acquired Silicon Perspective and Plato Design Systems, and is in the process of acquiring Simplex Solutions. So the design automation industry appears to be finally consolidating, to some degree."

Hitachi's Mu-Chip. An RFID Small Enough to Implant in Paper Money, RFID Journal 3.14.03

You might not believe it if you didn't see it (or in this case, hardly see it). Tracking of anonymous currency now becomes possible, a major social immune systems advance.

18. Visionary and Vaporware Sci-Tech (Prototype or Hype? You Decide)

Face Transplants? The Media Circus Begins, Feb 2003
Apparently Peter Butler, an Irish plastic surgeon who has been aggressively pushing this issue for years now, is going to get his first experimental subject for face transplantation. First candidates are burn victims, facial cancer survivors, or others who have unsuccessfully undergone reconstructive surgery. The idea is that a recently deceased donor's entire face, and facial bones when helpful, are grafted onto the recipient. Someone leaked his plans for the first surgery to the press (first link above) and the indignant family of the prospective patient is now considering suing (second link). Maybe the rumored $250,000 for media rights will keep them from seeking redress in the courts.

The breakthrough that makes this procedure possible is that a lifelong harsh regimen of immunosuppressive drugs may no longer be necessary. At least that's the theory. Tolerance induction is a promising new approach in immunology, but it has had only limited success, and we don't yet understand how it works. (Gene therapy, another interesting new and equally poorly-understood approach, has had far less success). Grafting of donor bone marrow to the recipient's body, along with the transplant, helps retrain the recipient's immune system to not reject. The experiments are very new, and there could be a lot of problems with this process. Here's PDFs from 2001 and 2002 using mixed allogeneic chimerism in mice, the leading approach, but it may require some very aggressive interventions (cyclophosphamide, thymic irradiation). Here's one showing that selection of a population of graft-tolerant immune CD4 cells also seems feasible, at least in immunodeficient mice. So where's the beef?

Butler says the technical problems have been solved, that only the social debate is needed now. But he's being disingenuous. Sure, there have been a handful of short term successes in animal studies. But we don't understand how any of this really works yet, and why it doesn't work when it doesn't, or the long term issues. Labs like David Sach's at Harvard have been working on this for four years now, but the field is still very early in development. Let's hope this poor disfigured girl makes it through successfully. As long as she knows the risks and has Butler's support team behind her, perhaps it should be her decision vs. a social one, although that certainly remains debatable with very untested, high-risk biointerventions.

Will we see 60 year old Beverly Hills clients bidding on the faces of 20 year old accident victims by 2010? Head transplants in 2015? Don't bet on it. My prediction is that while this therapy may eventually turn out to be appropriate and socially acceptable for a few hundred people around the world a year, it will be quite carefully regulated, and turn out to be a lot more of a technical hassle than Mr. Butler is letting on. For some very interesting reason, rejigging complex adaptive systems to try to make them mix-and-match-modular is a lot less productive than working within their natural evolutionary developmental frameworks. And by far the most dynamically complex natural ev-devo framework available on this planet, as physicists like Eric Chaisson have proposed, is the technological one, not the biological. Expect very moderate progress in transplants, as newsworthy and helpful to humans as these issues are. Thanks to Peter Voss.

19. Tools for Modern Living

Scirus, Elsevier's Scientific Search Engine
"The most comprehensive science-specific search engine available on the Internet. Driven by the latest search engine technology, it enables scientists, students and anyone searching for scientific information to chart and pinpoint data, locate university sites and find reports and articles quickly and easily. You can:

The web continues to self-organize. Thanks, Elsevier!

Remote Control of Your PC
PC Magazine in July lists a number of programs (GotoMyPC (best), In Touch (OK), LapLink (OK) etc.) for this important ability, but all have the monthly or yearly subscription pricing model. No thanks! I'll wait impatiently for a good flat-fee commercial product, or for an open source solution. This is basic functionality that all computer users should have inexpensively, now that we've got broadband internet.  I'm sure that measurable collective productivity for the country as a whole is being lost without it. In my ideal world, we'd have a president that rails about digital productivity on national TV. Now that's leadership. How many times would remote control have solved an information access problem for you in the last year? The two starter features needed being access to your file trees and to Outlook/email client (contacts, email, etc.) from work/home/road. Why can't Microsoft work a flat-fee deal with one of these companies, or if necessary fund a third party company to make a cheap version of this? It would be one way to keep their OS ahead of the relentless march of the Linux machines. March march march along, penguin.

A Run-Proof (Almost) CD Player for Under $100, Sony D-SJ301
(Type "Sony D-SJ301" into http://froogle.google.com/ for the best latest prices).

Yay Sony!  They've nearly done it. This beauty will play your CD's, including homemade CD-R's, while running with it attached to your body, mostly without skips. Well, let's be 100% accurate. It will skip when you run down trails at high speeds. And it will skip when the battery gets very weak. And it may skip more frequently the older your unit gets. But otherwise, its perfect. I keep mine in a little pouch attached to my waist, but you could also string a small nylon waist belt through the tiny, nonfunctional clip on the back.

The unit is designed to be held in the hand, of course, but that's irritating and dehumanizing. I find these technologies make daily exercise fun, or to be honest, at least tolerable, and daily exercise will keep you healthy and happy. Now will somebody please interface a CD player with a static memory buffer that holds the entire CD in RAM so it won't skip anymore, under any conditions? And while you are at it will you make a RAM-Delay Car Radio so we can push a button and hear the last few seconds or minutes of that song or information that we want to hear repeated? Designers: We overworked and underexercised first world humans are all SO TIRED of living in the Audio Stone Age!

Interactive, Intelligent Gym Equipment: TechnoGym, The Future of Exercise
Repetitive exercise can be terribly boring, particularly to first world multitaskers. Equipment manufacturers are finally realizing this, and designing a new crop of gym-based machines that have personal entertainment built in. Exertris has a line of interesting, first-generation exercise bikes where pedaling improves your ability to play your game of choice. Not networked, but a start. The Reebok CyberRider is another such interesting video game/bike combo system.

The Sports Club/Beverly Hills, opening Fall 2003, has promised individual television screens with each piece of cardio equipment, breaking a longstanding barrier. This will finally allow people to tune into their own information feeds of choice. Presumably you'll buy your own set of ear buds for $5 from the front counter, so people can exercise near each other without noise pollution. Finally, TechnoGym (shown here) has a new Excite line of TV-equipped treadmills, bikes, rowers, and other cardio equipment that are going to become come required equipment at all the top-line gyms during this decade. I'll be very interested in the studies, but I'd imagine that this will have a positive impact on length of exercise, which is a good thing.

Expect DVD-playing gym equipment as the next standard in personalization (Want to watch a 30 or 50 minute sitcom while exercising? A full length movie? The gym will have a library for you to choose from, touchscreen accessed from your gym equipment). Post 2010 I'd expect equipment that will accept an interface with our laptops, PDAs, and wearables, which by then will be starting to become personal entertainment and educational devices for some of us, once they are able to accept downloads of tens of hours (eventually, hundreds of hours) worth of specialty-interest video clips each month.

But perhaps most promising from a social standpoint, networked home gym equipment, allowing you to compete with others in realtime, and to be coached virtually through the screen, is another excellent innovation just waiting for bandwidth and display technologies to drop a bit more in price. Such home gym systems will work best when you are monitored for ability and matched with close competitors within your city, and when these are integrated with optional regular monthly live meetings, both at your local health club and out in some real world environment, such as a kayak race.

Tomorrow's Health Club and Amateur Sports Networks will give us the convenience of working out anytime with a virtual group and coach, as well as access to many more real world sports communities, populated with worthy competitors who become exercise friends, than ever existed before in your area. Such developments are win-win, non-zero sum emergences that seem inevitable, from my perspective. Bring it on!

Mini-B, A Simpler Scuba System.

Have you been waiting patiently for a simpler approach to scuba? Mini Breather International has done the first redesign of scuba gear in 50 years. Their Mini-B, pictured here, contains everything a sport diver needs in this small, lightweight backpack, including a backup regulator and independent emergency air source. Just strap it onto your wetsuit and you're ready to go exploring, Captain Nemo-style.

Now all we need to add to this is a clamshell wetsuit (3 minutes on and off) and an acrylic bell helmet with a mouth-accessible emergency air backup inside the bell, and the sport diver would be a much more commonly seen species on our coastlines.

Rebreathers, Tomorrow's Dive Technology Rebreathers filter out CO2, and feed you new O2 as your body consumes it. Bubble free, potentially lighter, and longer bottom time. The advanced units can dramatically reduce decompression times too. Dolphin and Azimuth are two Open Circuit sport diving rebreathers you can use, but both have drawbacks. The best unit available currently seems to be the Closed Circuit Inspiration, which dramatically reduces decompression times (only 12 minutes for a 30 meter, one hour dive. Nice). These systems all need special training, and are a lot of responsibility for upkeep, gauge monitoring, fresh cylinders, etc. The Inspiration looks like the least real-world hassle, and is the furthest along toward the unit we'd all want. Personally, I'm going to wait for at least another generation before trying them. Diving is a hazardous sport, and there are reports of people who have died using this equipment (in addition to ordinary scuba, of course). Unless you enjoy learning about and maintaining the technology, going with simpler systems is probably better until they have a lot of intelligence and self-diagnostics built into them.

Part 1  Part 2   Part 3   Parts 4-5   Parts 6-7