Vous êtes sur la page 1sur 6

THE HYPERCUBE: Organizing Intelligence in a Complex World*

by Lewis J. Perelman
One form of the distinction between information and knowledge is that knowledge is information about how to connect information. Not uncommonly, the discovery of a pattern connecting previously unrelated bits of information occurs through serendipity. The churning of stuff through one's mental inbox generates random juxtapositions that can suddenly suggest new meanings to anyone ready to notice them. Key discoveries in science and technology often happen this way. Consider these: An accidentally contaminated experiment led to the discovery of penicillin A drum, designed to record Morse code, when spun too fast, suggested to Edison the means to record sound A waste product from the illuminating oil industry, gasoline, was exploited by other inventors to transform modern transportation

And so on. James Burkewhose profession is a peculiar amalgam of historian, scholar, journalist, and showmanhas made a lucrative career out of reporting such oddly fortuitous connections. I mention all this mainly as background to a personal experience of this sort. A few years ago, in my quest to decipher the mason-barnstormer (M and B) clash (see sidebar), I interviewed the thirty-something founder of a hot young company in the booming field of data warehousing and data mining. That evening, I happened to read an article on the future direction of supercomputing technology. One thought led to another, and the following quasi-random notes are what evolved. In trying to peg the young CEO and his firm, I asked my usual question: How do you recruit the right kind of talent for such a new and fast-changing field? The split between Ms and Bs is roughly, but not necessarily, generational. And despite his youth, my interviewees answer had a distinctly M class tone: "We recruit from the top five percent of the top two percent of (the) university programs." He had enough B-class tendencies himself to recognize that there were people with valuable talent who lacked the usual academic credentials, and even noted a couple he had found in his company who hed reassigned to key engineering work. But to him, searching for the best talent in the best schools was a matter of actuarial efficiency: MIT has already filtered them out, he explained. In the course of our far-flung conversation, this cyberpreneur also had mentioned that massively parallel computing technology gives a boost to his firms software tools, which are designed to glean valuable gold nuggets from huge data mines. The comment triggered a twinge of irony that I could not quite grasp until later that night when I read a Science magazine article on supercomputing that had drifted fortuitously to the top of my ever-sprawling pile of reading matter.

2001, Lewis J. Perelman. Parts of this essay were originally published in Knowledge Inc. (1996, 1997) and later in a Cutter Consortium Business Intelligence Executive Update (2001).

Perelman

Hypercube

With the U.S. governments encouragement, the article reported, the next great leap in computing power would aim at a thousand-fold improvement in power to petaflops1000 trillion operations per second. The approaches to this daunting task are arrayed between two polar strategies. One strategy is to use a limited number of super-fast, superconducting processors. The other aims to employ a large number (millions) of relatively simple processors with the useful attribute of containing their own memory. So the first design extends what George Gilder characterizes in his book The Microcosm as the high-and-fast Cray-type supercomputer architecture, while the latter projects the sharing of the processing load among a population of low-and-slow chips of the sort that drive a PC. The designs of computers around the massively parallel (MP) integration of numerous cheap, puny microchips have had several names and brands, but an early one I've always fancied was hypercube reflecting the connection of processors in several more than three directions or dimensions. Metaphorically, the superprocessor is class Ma distinctly centralized, elitist design. The MP architecture represents the B-class alternativea decentralized, egalitarian, even literally vulgar community which, like a termite mound or beehive, achieves a brilliant intelligence through the networked relationships of a horde of rather dull constituents. You can see the same polarization of vision in other areas of advancing technology, too. In the field of robotics, the quest for human-equivalent, super-intelligent C3PO-type androids is complemented now by another track aiming for a larger, more diverse population of insectequivalent, relatively cheap mobile gadgets that are clever at doing one or two tasks really well. Similarlyas an alternative to computer mega-modelsgenetic algorithm and synthetic evolution systems pit a large population of simple-minded virtual creatures in a competitive churn that evolves rapidly to spawn digital beings that cannily solve problemsfrom finding effective drugs to picking winning stocksin ways that are often incomprehensible to their human parents. Therein lies what struck me as the paradox in the young CEOs preferences in computer architecture and organizational architecture. He saw the power that could be attained from harnessing together the capabilities of a collection of mundane computer chips, but preferred to build the architecture of his business organization around human superprocessors. This may seem to make sense if you look at an organization in isolation. But in a market space where many other enterprises are practicing the same credentialist employment paradigm, you get the same foolish diseconomies as appear in some professional sportswhere greed-driven expansion and boundless bidding for superstar talent leads to a degradation of team performance as well as ultimate insolvency. I saw a compounding irony in the story of this young company: The selling point of its data mining technology is the ability to find golden needles in giga-haystacks of statistical straw. Yet it had not occurred to the CEO that finding the right kind of human resources among the burgeoning billions of the earths population was an analogous problem to which his companys own tools might offer a practical solution. Not only practical, but potentially very lucrative, because so many other knowledge-age organizations face the same need. But I don't mean to pick on one manager or one company. The contrast, and growing mismatch, in our strategies for techical and human architectures is commonplace, practically ubiquitous.

Perelman

Hypercube

Which brings me to the notion of the human hypercube. If we can organize PC chips, minibots, and A-life creatures with all the intelligence of an amoeba into systems that can solve problems and lead to brilliant discoveries, what productive power could be unleashed if we did something of the sort with networks of human brains? Yes, I know that teamwork is a near-obsession in management gurudom these days. But as MITs Michael Schrage rightly points out, team is a facile organizational form that commonly has little relationship to the subtle process of collaboration. The anthropological observations of communities of practice as they naturally occur in business and other social groups may help meet the call of Schrage and others for organizational processes, practices, policies and incentives that are at least friendly to traditional forms of human collaboration. Work on multiple intelligences, learning styles, and other aspects of the cognitive sciences, behavioral traits, and performance assessment are fine-tuning our perception of the diversity of performance capabilities people can contribute to a collaborative process. Telematic networks loaded with groupware, whether proprietary or Web-based, may further expand the opportunities for collaboration. Ditto electronic performance/decision/knowledge support systems. But all that is not the same as inventing new and greatly more prolific forms of human parallel processing with the same fervor, focus, and funding that we see being invested in the cyberworld. It is abundantly possible this is going on outside the range of my limited knowledgeif so, Id love to know about it. Yet the impression I get (from what my Web surfing and text churning has turned up) is that we are collectively pretty much in the dark ages in our knowledge and design of human hyperworking. For instance, there are these items in the desktop moraine my serendipitous study has dredged up so far: In his book, A History of Reading, Alberto Manguel makes the surprising (to me, at least) revelation that, for all its hoary tenure and profound social impact, the basic process of reading remains largely uncharted by science. Mysteriously, we continue to read without a satisfactory definition of what it is we are doing, Manguel says. We know that reading is not a process that can be explained through a mechanical model. We know that it takes place in certain defined areas of the brain, but we also know that these areas are not the only ones to participate. We know that the process of reading, like that of thinking, depends on our ability to decipher and make use of language... but, Manguel senses, researchers seem reluctant to pursue that too far for fear that they may discover that language itself may be nothing more than an arbitrary absurdity. Whoa. When we don't even know how reading works, how much productivity can we expect to gain even from paperless digital documentation, content and groupware, much less collaborative writing, editing or problem-solving, all of which, like the Web generally, still are heavily rooted in text of one sort or another? Another article, by Lisa Alcalay Krug in Forbes ASAP, reported on the growing Hatred (the pieces title) between CEOs and CIOs. About two-thirds of managers surveyed said they would like to scrap their whole IS departments and start from scratch. Meanwhile 54 percent of IS techies told the surveyors that top management doesn't have a clue about ITs problems and challenges. Which leads me to wonder, where is the natural corporate

Perelman

Hypercube

consitituency that understands what the heck human hypercube might mean, much less what to do about it? One, final discovery in my search for direction on the human hypercube was an insightful essay I surfed up on the Web by Philippe Baumard, a researcher of strategic management at the University of Paris. The paper, on the paradigm shift From InfoWar to Knowledge Warfare, starts with an observation by General F.M. Francks that Vietnam saw the first battlefield use of computers, such as the van-filling Univac 1005, which digested reams of data from reconnaissance photos and an array of starlight, infrared, and radar sensorsall giving a battlefield map of unprecedented detail. Meanwhile, Francks noted, the Vietnamese were digging tunnels and hiding in plain sight among the peasantry. All this a reminder of the capacity of those David Halberstam tagged the Best and the Brightest to, in the words of Sherlock Holmes, see everything and observe nothing. Granted, recent battlefield successes in the Persian Gulf in 1991 and in Afghanistan now show that our military has made great strides in the implementation of tactical intelligence. But the rubble field that sprawls where the World Trade Center once stood is a desperate monument to the shortcomings in the capacity of our overall strategic intelligence to protect our society from horrific threats. From this and similar experiences in economic and commercial as well as military spheres, Baumard argues persuasively for national knowledge strategies focused on the advancement of knowing, instead of the accumulation of knowledge. And future development of national intelligence should target the improvement of interpretational and sense-making skills, instead of the utopian quest for universal access to mere knowledge as a commodity. Good advice as far as it goes. But, despite a passing reference to interconnectivity and interoperability, Baumards prescription tracks the obsession of a long line of distinguished Gallic thinkers, from Voltaire and Rousseau to Zola and Piaget, with reengineering of education to produce a new breed of better thinkers. The flaw in the reach for social intelligence through academic excellence was graphically demonstrated by the young CEO I mentionedMichael Saylor. He and the best and brightest he hired at his company, Microstrategy, manipulated accounts to create the illusion of profits where there were actual losses. When the deception was exposed some years later, over 90 percent of the companys market valuation, and of Saylors personal fortune, were evaporated when its stock price crashed. That tack of seeking organizational intelligence through better diplomas veers away from what is really the central issue: the organizational ecology that filters out or neuters over 90 percent of the creative and critical talents present in the current, normal population. The productive potential of hypertechnology and hypermedia is bounded by our progress toward human hypermation. end Sidebar

Masons vs. Barnstormers


For the first couple of years that I'd been talking about the culture split in the modern workforce, I sometimes apologetically hung the labels hip and square on the two emerging classes (falling back on Beat terms that were already archaic in the sixties), for lack of a better idea.

Perelman

Hypercube

But one day, while rearranging my office, I stumbled on a quote from the visionary Theodore Sturgeon whose imagery suggested what I feel are more apt terms. In a book review, Sturgeon offered this pivotal insight: the universal quest for stability takes two formsthat of the pyramid builder, who stakes out his territory, measures it, and plans course after course toward a definite goal, or apex; and that of the gull, whose stability is dynamic, who must be ever in motion to remain stable, and who has equipped and trained himself to tilt and recover, to adjust to direction and temperature and density and whatever else he may encounter. This is very close to the division in workforce, organization, and management we see developing through the transformation from an industrial to a knowledge economy. So, translating Sturgeons metaphor into human terms, I proposed to call one workforce culture class M, for masonic, and the other class B, for barnstormer. Pyramid builders after all are literally masons. In fact, the fraternity of Freemasons traces its spiritual roots to the builders of ancient Egypt. But as a creature of the 18th-century dawn of the industrial era, the formal Masonic order is also based on the human pyramid of the organization chart, and its rituals of social exclusion, induction, and pyramid-climbing advancement by degrees are typical of the culture of many industrial organizations. It occurred to me that, in contrast to the M class builders and climbers of social pyramids, the human analogy to the seagulls are the type of people who studied, analyzed, and ultimately recreated in metal and fabric the fluid flexibility of the gull's flight: the barnstormers who pioneered the aviation industry. The barnstormersfrom Wright to Curtis to Lindbergh and a slew of otherswere both individualists and collaborators, but not bureaucrats. They lacked credentials, standards, rank, or position. They developed aeronautical engineering by building aircraft, and they mastered flight by flying. Class B clearly is not limited only to the aviation industry and has played a visible economic role at least since the time of the American colonies. Arguably it was the dominant paradigm in America at least until the middle of the nineteenth century. But the explosive growth and sprawl of giant industrial bureaucracies in the course of the twentieth century, accelerated particularly by the Second World War, edified class M and pushed class B down and out to the economys margin. Now the pendulum is swinging with gathering momentum the other way. While weve been talking about the postindustrial economy for a full generation, the impact of what Alvin Toffler called the Third Wave only became broadly tangible and visible around 1990, marked by the simultaneous collapse of the Soviet empire, the obliteration of a Second-Wave by a Third-Wave military force in the Persian Gulf, and a macroeconomic upheaval (downheaval?) which departed from historical patterns. The recession that followed the end of the Cold War spawned a wave of mergers, downsizing, outsourcing, and reengineering that generally dislocated the previously unscathed M class: whitecollar, college-degreed, managerial and technical professionals. The combination of venture capital, structural innovation, and labor shortage that marked the succeeding 10-year boom gave unprecedented opportunity, power, and wealth to the B Class. The same technological, organizational, and market forces that give nightmares to the class M workforce are a barnstormers dream. But for the past decade, the inherent conflict between the two cultures was mollified by the balm of growth. In fact, as the image of normal corporate success and culture shifted toward the B-typeand its pattern of overt informality, fun,

Perelman

Hypercube

inventiveness, youth-focus, fluid if not fickle relationships, micro-entrepreneurialism, and go-go optimism bordering on arrogancea fair number of M-Class stalwarts defected to the B-side, often with some fanfare. The combination of economic recession and the strategic upheaval following the events of September 11 have thrown us suddenly into a new new economy whose form and direction are as yet unfathomed. In the short run, there has been some backlash in the direction of more conservative dress, behavior, and management. But the ink, once spilled, does not go back in the bottle without leaving its mark. Already it is evident that the imperatives of both economic recovery and strategic security demand an accelerated, albeit adapted implementation of the structural transformations launched during the late expansion: more frictionless collaboration across internal and external (B2B, B2G, G2G) boundaries, more intensive data-gathering, data-mining, and application of intelligence to operations, more adaptation of management to complex distributed systems, more flexible management of human capital, and so on. In this new churn of power and competition for control, direction, and resources, the inherent conflict between the two cultures inevitably will be in play. Neglected, it may boil over. Managed creatively, it can proceed to a new synthesis of organizational, and inter-organizational designs. Finding a way to handle this transformation of work, workers, and organization will remain a firstorder challenge as we chart new paths toward prosperity and security in dangerous times. XXX

Perelman

Hypercube

Vous aimerez peut-être aussi