Thursday, January 18, 2007

Early Indications January 2007: A Different Kind of Prediction Letter

"No man is an island, entire of itself; every man is a piece of the continent, a part of the main."

Before we can talk about systems, we need to start smaller. A standalone technological artifact can be called a device, which is technically "a machine used to perform one or more relatively simple tasks." Electronic calculators and electric drills are common examples, illustrating that devices are generally more complex than simple tools, such as a hammer or pencil.

A set of devices sharing communications rules and facilities can combine into a platform, in which the devices gain in utility by being able to work together. A hard drive, display, processor, and peripherals like cameras and printers constitute the personal computer platform. Like a device, some platforms are used by individuals, but others (such as computer-controlled machine tools) constitute commercial infrastructure.

When platforms are connected by communications standards and technologies, they become systems. An airline network is a system in that runways, pilot training, technically precise language (as used in the cockpit and in communications with control towers), refuelling inventories, jetbridge heights, catering cart dimensions, and countless other aspects are coordinated in order that passengers can get where they want to go at minimal cost and a high degree of safety. Perhaps the most salient fact of systems is their complexity, which often scales nonlinearly with attributes such as size, population, and so forth.

Because systems are big, expensive, and complex, they must be governed in special ways. Private ownership of the railroads created a backlash that still curtails the power of airlines. The power that AT&T derived from owning the U.S. telephone system, literally down to the kitchen handset, eventually led to its being broken apart by regulators. Because systems often appear to be natural monopolies, they tend to be either government-owned or highly regulated. Even so, systems often provide economic opportunity on a very large scale: Motorola could not sell over 50 million Razr wireless handsets at a high premium without existing networks for the devices to connect to.

Systems typically support and depend on other systems. An airline relies on banks for working capital, on hotel chains for traveler (and crew) accommodations, and on government for certifications of various sorts. If the government doesn't provide a certificate of airworthiness for a new plane, no bank will lend the money and an outside insurer won't underwrite the liability. A personal computer without a telecommunications system and an electric power grid is essentially useless. At a deeper level, a keyboard without a language and syntax of expressions, both grammatical and mechanical, is quickly reduced in value. These dependencies among systems, and the resulting complexities, drive our 2007 predictions.

Recall that the invention of a gasoline-powered automobile took over a half-century to generate related systems of gas stations, factories and factory jobs, financing companies, limited-access highways, suburbs, enthusiast and consumer rating magazines, race tracks, drive-in restaurants and bank windows, safety regulators and regulations, environmental impacts, insurance companies, vacation destinations, and dozens of other artifacts that marked the United States of the 1950s and afterward. Not long ago one out of every seven U.S. jobs was connected to the automobile. Systems of systems - in this case, mechanical, financial, governmental, legal, and economic - take a long time to develop but, once running, exhibit "flywheel effects" in that they possess considerable momentum and continue to spin for a long time even after the source of driving energy is removed. Systems also resist change.

The so-called ICT (information and communications technologies) sector has long been powerful, but in the recent past it has begun exhibiting the characteriststics of a system of systems. A key development in this transition is the development of third-party funding, often through advertising. Just as television broadcasts are "free," so too are more and more technology benefits funded through online advertising revenue. The Internet of 1997 was a system, no doubt, a "network of networks," but we now have human, economic, technical, and political systems all riding on the same technology roadbed. Brides and grooms, job-hunters, election-seekers, terror networks, stock-market investors and manipulators, stamp collectors, wildlife biologists, role-players, and any other identifiable population now has a significant digitally connected component. In addition, so-called Web 2.0 business methods and models are using those human networks as economic engines.

This preamble brings us to the 2007 predictions. It's getting difficult to find economic actors that don't need to be actively involved in technologically-constituted systems. The Internet flourished without substantial government regulation until a large number of constituencies sought redress and/or protection, most in the past several years. Online gambling, voice over IP services, "net neutrality," alternative broadband providers, targeted advertising, sexual material, maliciously edited public profiles on Wikipedia and elsewhere, unauthorized use of copyrighted digital content - the list goes on and on, but in each instance we see a collision between systems based on old and new models of regulation, remuneration, protection, privacy, and so forth. At base, we are having to redefine some of the core systems that make the world work: money, contracts, civil rights and civic responsibilities, identity, possession, and others.

This year, I believe that several of these collisions will reach new heights of unexpectedness, expense, and impact. Some candidates follow:

-Last year an agency of the U.S. government lost 26 million names with personally identifiable information on a single laptop. Monetary losses from online fraud, extortion, and related exercises were not reported, but are estimated to range in the hundreds of millions of dollars worldwide. This year, look for a still grander failure of data protection, either in one highly visible episode or a cumulative increase.

-YouTube and related content distribution mechanisms will push the envelope too hard, with a high-profile episode of unauthorized copy distribution prompting legislation, litigation, and potentially business failure.

-Some new activity - whether job referrals, recipe swapping, rotisserie baseball, geneology, Christian evangelism, or something similarly below radar - will break through using a Google-like monetization model and approach the growth rate we saw for video in 2006.

-Around the time of Gerald Ford's administration, a bumper sticker was distributed as speed limits on federally funded highways were reduced to save on petroleum consumption: "55: It's not just a good idea, it's the law." Science nerds quickly came out with their version: "186,000 miles per second: It's not just a good idea, it's the law." For all their amazing capabilities, communications and computing systems still can't cheat physics. 2007 will see the so-called virtual world continue to encounter the physical environment in important ways. A few examples suggest the breadth of the issue:

*Data centers are beginning to scale up to the size of factories and even foundries in their energy consumption.
*Intel continues to struggle with the heat output of microprocessors, apparently more than AMD does.
*The long-term effects of electromagnetic radiation (especially introduced to the brain through a hole in the protective skull) remain unknown.
*The digital bread crumbs left behind by automobiles, cell phones, PCs, credit cards, and identification technologies (toll passes, fingerprint readers, browser cookies) continue to accumulate on a massive scale. The potential for abuse grows apace.

-A different facet of the energy and transportations systems relates to automobiles. While Chevrolet just introduced a good-looking electric car, the Volt, that anticipates developments in battery technology, Tesla Motors will ship over 200 Roadsters at $100,000 apiece that out-accelerate a Porsche 911 and achieve the equivalent of 135 miles per gallon fuel efficiency. Many thousands of similar electric cars could be deployed on the current power grid once Tesla and other firms approach mass-market pricing driven by economies of scale. In contrast, ethanol technology is energy-inefficient (but getting better), and there is no infrastructure in place to store or move ethanol at the required scale: because alcohol conducts electricity while gasoline is a dielectric, for example, fuel pumps need to be redesigned. More critically, alcohol at 85% concentration attacks the metals and gaskets used in current fuel pipelines and tankers. E85 also delivers inferior gas mileage compared to gasoline. In short, innovation -- at the device level -- to improve electric cars produces a much lighter load on the current supporting systems than would a mass adoption of ethanol. How politics and markets react to rising oil prices, from a systems-of-systems perspective, will determine quite a bit about the shape of the next 10-20 years.

-After 9/11 and 7/7, Katrina and the other hurricanes of 2005, the Pacific tsunami, SARS, the 2003 power grid failure, and the Long Beach port strike, 2006 did not see a major disruption to the world's transportation and communication systems. Such good fortune cannot last indefinitely, yet readiness for the unexpected remains lower than it could be, especially after the many upheavals that followed the Y2K non-event. Some observers see video peer-to-peer networks, such as Joost from the founders of Skype, as potentially crippling the current Internet, for example. In the past month, try telling someone who had Norovirus that it's less severe than avian flu.

-Paradoxically, even as people and devices grow more connected, with access to more information, the need for intermediaries evolves rather than disappear. In the 1960s, most Americans could watch one of three network news shows and subscribe to one of two or possibly three daily newspapers. Retail selection was vast, compared to the 1930s or 40s, but still constrained to what merchandisers chose to present. Now, consumers can shop across vast inventories of food, fashion, entertainment, and other products (the so-called "long tail"), theoretically making each person his or her own merchandiser, video producer, or editor. Just as the late 1990s was the heyday of disintermediation predictions before brokerage firms, real estate agents, car dealers, and other middlemen responded, so now we will see how newspapers, magazines, book publishers, movie studios, universities, and other content businesses respond to the increase in direct access to many categories of content. The fact that the American Film Institute noted in its "2004 Moments of Significance" that "one of the best sources of news today is [Jon Stewart's] faux news show" would be inconceivable in the 1970s: Chevy Chase more authoritative than John Chancellor? Expect more, not fewer, of such puzzling inversions going forward.

In short, 2007 will continue the trend toward more complex types of change as information and communications technologies intertwine the world in more ways. When John Donne wrote the famous words in the headnote nearly 400 years ago, he anticipated a system of connection that remains incomprehensible in its scale, its impact, and its possibilities.

Early Indications December 2006: How'd We Do?

[distributed 12/13/06]

In January of this year, we published eight predictions. At this
point, the score is six hits, an incomplete, and a slight miss.
Overall, it was a pretty good year for technology and innovation, with
a variety of new wireless technologies getting closer to market, a
breakout year for web video, and an ambitious set of web services from
Amazon (including everything
from artificial intelligence to warehouse shelf space).

On the mis-step front, the litany of privacy breaches became longer
and louder, with a stolen laptop costing the U.S. Veteran's
Administration a projected $160 million dollars had it not been found.
In the midst of absorbing losses in the hundreds of millions of
dollars because of defective batteries, Sony launched an expensive,
powerful gaming platform that it couldn't supply in holiday
quantities. The "netroot" bloggers made a lot of noise but in the end
did not get their U.S. Senate candidate elected.

1) The second half of the year will be stronger than the first half in
the PC sector

Score: Hit
Even though Microsoft's Vista operating system has yet to generate any
PC sales, because consumers can't yet buy it and enterprises will have
to certify their existing applications on the new platform, this
prediction did in fact come to pass: at HP, the Personal Systems
group saw revenues for the three months ending October 31 rise 13%
over the quarter ending July 31. At Dell, only preliminary numbers
for the quarter ending November 3 have been reported, so it's
difficult to say with certainty what's going on.

2) "Services" will become the corporate IT buzzword outside IT

Score: Hit
SAP recently announced that its core positioning will focus on what it
calls Enterprise SOA or ESA, with half of R&D spending committed to
creating up to 30,000 SOA-driven business processes. Accenture is
spending an announced$450 million on SOA, HP $500 million, and IBM a
total of $1 billion on SOA over the coming years. Oracle, BEA, Sun,
and most software vendors apart from Microsoft, which is not branding
services so aggressively, are joining the gold rush.

3)Google will launch a breakthrough business outside web advertising

Score: Hit
Counting only their distribution deal with BSkyB, the YouTube
acquisition, the launch of the core of an online office suite (Docs
and Spreadsheets), and the challenge to PayPal implied by Google
Checkout, it was a big year for Google. Going forward, the company is
joining Amazon, Yahoo, and other search firms in building enormous
data centers to support further expansion of so-called "cloud
computing." George Gilder's recent article on these data centers in
Wired, apart from its purple prose, is required reading, particularly
for its take on the electricity consumption issues:

4) HDTV will have collateral effects

Score: Hit
According to market researchers DisplaySearch, HDTV displays broke
through and accounted for over half of the North America TV market in
2006. Price drops continue, even in the face of such strong demand,
in part because the big producers see additional production capacity
coming on line in the near future. Verizon's bet on fiber to the
home, meanwhile, may be concerning the cable TV operators, whose
industry lab was reported this summer to have questioned the wisdom of
long-term investment in a coaxial infrastructure with strict, and low,
limits on HD traffic.

5) The relentless reinvention of business markets by the Internet and
digitization will continue

Score: Hit
"Who might be next? Television is my best guess." With YouTube and
BitTorrent both getting content distribution deals from major players,
and with cellular continuing its push toward video, the motion picture
and other video incumbents are confronting a dramatically different
landscape. A big story here was ESPN pulling its cellular phone
service after only about six months.

6) The quiet march of robot progress will continue

Score: Incomplete
The march was so quiet I couldn't hear it. There was no high-profile
story on par with 2005's DARPA challenge, which will be re-run in 2007
in simulated urban traffic, rather than last year's open albeit
obstacle-strewn desert environment.

7) Sensors and other location-awareness technologies will make the
news for an unexpected consequence

Score: Miss
RFID in the supply chain is finally providing suppliers (as opposed to
retailers like
WalMart) with a compelling cost justification: promotion
effectiveness. Think about a consumer products manufacturer (a
fictional example would be a battery company before Halloween): if I
deploy an expensive, time-sensitive end cap or other display, I want
to know if a chain's thousands of stores are in fact displaying my
promotional material and inventory. If they aren't on the sales
floor, customers could be confronted with stockouts and/or I must take
returns of seasonally-specific merchandise, such as Halloween
packaging in November. Apart from that realization, the toll bridges,
automated thermostats, automobile black boxes, and their kin
apparently worked well enough not to draw notice. RFID-equipped
passports are concerning industry observers who see how easy it is to
read them from a distance, and a company that sells tags for human
implantation (with readers being given away to emergency departments)
is raising fears, but neither was a major story.

8) The developing world will once again make headlines for innovation
and not just cheaper
production costs

Score: Hit
A Nobel Prize for microfinancing in Bangladesh, Brazil's leadership in
ethanol, and Korea's
launching a mobile WiMax service years ahead of the U.S. or Europe all
seem to count here. It's noteworthy that two of these three examples
involve organizational rather than primarily industrial innovation.

Also in January, we discussed six macro trends, and several of them
certainly made an impact:

Climate change: Just this week, a preprint of an article in
Geophysical Research Letters suggested that the Arctic could be open
water by 2040, maybe sooner.

Avian flu: Nothing on the epidemic front so far

Unstable energy prices: Big news here, especially mid-summer

The end of the bi-polar world: Russia, Venezuela, the Middle East,
North Korea and Darfur certainly proved that regional instability can
reach far and wide.

Decreased faith in government and authority: The U.S. midterm
elections, a wide lack of confidence in the United Nations, and the
privatization of large-scale humanitarian efforts by everyone from the
Gates Foundation to Rick Warren suggest this trend is continuing.

Increased evidence of class conflict: "Conflict" may be the wrong
word, because there was less overt class-related rhetoric this year
than in Katrina or the Paris riots. But the word "separation" still
applies, with significant implications for the middle class in
industrialized nations: in 2005, the average CEO of a U.S. company
with more than $1 billion in revenues made 262 times what the average
worker was paid, the second-highest multiple on record. The average
income of the top 126 hedge fund managers this year, according to
Barrons, was $363 million, up 45% over the previous year. At Goldman
Sachs, meanwhile, the mean average compensation for all 26,000
employees, from administrative assistants to top management, is
$622,000 in 2006.

On that note, happy holidays. Watch for the 2007 predictions in January.