BULKY DOCUMENTS
`
`(Exceeds 3 0 0 page s)
`
`Proceeding/ Serial No:
`
`ff M o?/000
`
`Filed:
`
`Q/ »~Qé ——2Qpf
`
`Title:
`
`Oh
`
`’
`
`F
`
`(E
`
`E
`
`
`
`
`
`Part
`
`01
`
`of
`
`3
`
`

`
`5
`
`4
`
`THE VALLEY CULTURE
`
`ost profitable maker of inte-
`Intel is the world’s pre-eminent and m
`power everything electronic.
`grated circuits, those silicon chips that
`ICVCIILICS
`The company employs twenty—five thousand people and had
`in 1993 of $8.78 billion, along with one of the highest earnings per
`The heart of its value lies in its research
`employee ratio in the world.
`3.5 billion
`hich were expected to reach $
`and development projects, w
`1, is fond of proclaiming
`in 1994. Andrew Grove, the President of Inte
`h proi-H
`that his people routinely “bet the company” on huge researc
`“bet
`ects. In wilder moments, he expands this to declare that they
`millions on science fiction.”
`Intel today functions in some ways almost like the old Bell Labo-
`A
`mmitrnent to research qualifies
`seems nothing like the old Bell Labs
`for the place is saturated with Silicon Valley
`when you pay a visit,
`nonbureaucratic, wildly diverse. There
`style——informal, fast—moving,
`ilitary ethic and strict hierarchi-
`is little similarity to the traditional m
`d, pre-divestiture
`cal structure that prevailed at, and indeed epitomize
`AT8cT.When you drive up to Intel’s headquarters, a sprawling five-story
`building of sky—blue glass that looms over the suburban strip of Route
`101 running through Santa Clara north from San Jose, the first thing
`you notice is that there are no reserved parking places in the vast lots:
`it’s first come, first served, no matter who you are. Inside, too, the
`nonhierarchical Valley ethic prevails, for everyone, from the Chair-
`man to support staff, sits in cubicles, which are available only in
`zes. The outer rim of walls is not blocked off by enclosed
`several si
`s above the cubicles, enabling natu-
`offices as in most buildings; it rise
`ntel’s fluid
`for an open, spacious feeling. I
`tail light to fill every floor
`f space, which will be discussed more fully in
`and egalitarian use 0
`for it reflects much about how the
`Chapter 8,
`is worthy of note,
`company is structured.
`
`.gt.‘
`
`
`
`

`
`
`
`The organizational chart is very flat, with only six layers for
`thirty-five thousand people—down from thirteen layers in the early
`eighties. Virtually_ no positions are solely supervisory. Lines of com-
`mand are multiple and diffuse, with much overlapping of depart-
`and everyone reporting to multiple bosses, performing multi-
`ments,
`ing enables the
`ple jobs, and sitting on multiple councils. This overlapp
`company to run as leanly as possible, for people are constantly being
`shifted to where they are needed, which saves having to overhire in
`times of expansion. Reorganization is continual and evolving, a way
`of life; there is no such thing as a static job. Carlene Ellis, presently
`Vice President for Information Technology, points out: “I’ve been
`here twelve years, and in that time I’ve had seven different jobs. Four-
`teen,
`really, since each job was reorged once, which completely
`changed how it was done and who for.”
`One aspect of Intel’s culture that made its practice of constant
`5 work is its emphasis on nonpositional power. The
`reorganization
`f webs was pointed out
`role of nonpositional power in the formation 0
`by Ted Jenkins, an engineer with a philosophical bent who has been
`with Intel since its founding. “Intel’s great strength,” he declares, “lies
`in the way the company allows resources to flow to wherever there’s a
`problem. I’ve thought a lot about why, in so many other companies,
`this just doesn’t happen, and I think it’s because in most organizations
`resources tend to accumu1ate——-they get stuck wherever someone is in
`a position of great power. So what you‘ end up with in most compa-
`nies is a few powerful people who have more resources than they
`actually need, while everybody else has to try to make do with less.
`It’s static, irrational, and inefficient.”
`Jenkins attributes this situation to the fact that traditional hierar-
`chical organizations are structured specifically to validate and exalt
`positional power. “In most organizations, it’s easy to figure out who’s
`powerful: you just look at where someone stands on the org chart.
`The only kind of power that really matters is the power of position.
`That makes it very difficult for other kinds of power to develop. The
`people at the top hold on tight, so no one else can establish any kind
`of alternate power base. And it’s this absence of other power bases
`
`THE WEB or INCLUSION
`
`

`
`
`
`
`
`56
`
`that permits resources to get stuck.” Position, however, is a relatively
`crude way of measuring power, as Jenkins points out. It cannot begin
`to reflect the subtleties of alignment in a company such as Intel.
`“Here, there’s no single way of being powerful. The power of position
`is just one aspect. And I wouldn’t say it’s the most important aspect at
`all.”
`
`Enumerating alternate sources of power, Jenkins names first the
`power of expertise, of knowing and possessing crucial skills. “A com-
`pany like Intel very naturally emphasizes the power of expertise be-
`cause eighty percent of the people we hire are engineers.” As in a
`partnership of professionals——accountants or attomeys-—people are
`chosen because they exhibit specific skills, rather than because they
`have potential as all-purpose managers. And since skill is so highly
`regarded, anyone who exhibits an unusual level of skill tends to ac-
`cumulate power. “We place a very high value on the power of ex-
`pertise, and the way the company is run tends to increase it,” says
`Jenkins. “We have this intense level of training that never stops,
`which of course broadens and deepens expertise. Also, because we
`are continually undergoing reorgs, people move around in the com-
`-pany quite a bit, which also tends to increase the scope of people’s
`expertise.”
`Moving people around a lot also helps develop another alternate
`source of power: the power of personal relationships or connections.
`“Because people are always being shifted, they develop a lot of con-
`nections in the company as they go along. They know a lot of people,
`because they’ve had a chance to work with so many of them——they
`have personal friendships, and they also have a pretty good idea of
`what the people they know might be capable of doing. This makes for
`a very networked organization, with lots of informal lines of commu-
`nication, lots of links that would never appear on any organizational
`chart. This is absolutely invaluable when you’re trying to put together
`a team, because people know all kinds of unexpected places to look
`for various kinds of skills. You can put together a very creative team
`that gives people a chance to develop new talents—which of course
`then also increases their expertise.”
`
`I-.
`
`ii
`
`
`
`
`
`
`
`____,n._:=_.‘___..._,__,......__=,_,._......,.........-...._...._..
`
`
`
`

`
`
`
`THE WEB or [NCLUSION
`
`Ted Jenkins also mentions the importance of the power of per-
`sonal authority as crucial in developing alternate centers of power. In
`any organization, he points out, you have people who, by virtue of
`. their personalities, their natural leadership skills, and the trust that
`they inspire, wield a power greater than their official position would
`indicate. Yet in traditional hierarchies, such people are often viewed
`as a threat, disruptive links in the functioning of the chain of com-
`mand. “In a company like Intel, such people really thrive,” notes
`Jenkins. “They make a big difference in terms of where resources
`flow. And again, the power of personal authority is really emphasized
`as people get shifted around, because their experience is broadened
`and more people come into their orbit.”
`Jenkins believes the primary reasons Intel has been able to en-
`courage alternate centers of power are its penchant for constant reor-
`ganizations and the unusually low turnover rate among people who
`work for the company. “In this organization, people don’t leave.
`They're paid fairly, they have a great profit-sharing plan, and they get
`a lot of opportunity to move around——it’s always something different.
`In companies where people are constantly leaving, it’s. hard for alter-
`nate power centers to develop.
`“It’s especially hard to develop the power of personal connec-
`tions if you’re not in an organization for very long, and without
`personal connections, it’s hard to put strong teams together. Everyone
`knows that companies have two organizational charts-—the formal
`one, listing everyone’s position, from the president on down; and the
`informal one that shows the actual routes of connection that allow
`things to really get done. It’s the informal power structure that always
`determines how effective an organization will be. And stability is es-
`sential in order to build a strong informal structure. Also, reorgs don’t
`do much but disrupt a company if people are always coming and
`going. I think your emphasis is always going to be on positional
`power if people don’t stay with the organization.”
`
`
`
`..'-¢»‘Q
`
`

`
`5 8
`
`IN THE HEART or THE VALLEY
`
`It seems fitting that Intel should provide an opportunity for watching
`how webs form and function, for the organization lies at the very
`epicenter of the ever evolving web of enterprise that is Silicon Valley.
`Intel was born in 1968, one of the first companies spun off from
`Fairchild Semiconductor, which was itself the original prototype of a
`Valley start—up. Fairchild came into being when all the engineers who
`worked for William Shockley walked out one day to protest Shock-
`ley’s management style. Shockley had been the head of the Bell Labs
`team that invented the transistor, but in the early sixties, he had
`returned to his hometown of Palo Alto to found the world’s most
`advanced semiconductor company. Much later, he would gain notori-
`ety as a proponent of racist evolutionary theories.
`The eight engineers who abandoned him to found Fairchild were
`thus imbued from the very start with a notion of business as a colle-
`gial enterprise, one in which expertise should be valued over posi-
`tional power, and the independence and individual talent of people
`throughout the ranks should be honored and recognized for what it
`was: the true strength, the reserve power, of an organization. The
`Fairchild founders went on to spin off scores of other ventures in the
`years ahead, thus seeding the Valley for decades to come with talent
`that was stamped with an antiauthoritarian and egalitarian ethic,
`while also driven by an entrepreneurial spirit.
`Intel was co-founded by the legendary Robert Noyce, a former
`Shockley protégé and himself a member of the original Fairchild
`group. Noyce was the co-inventor of the original integrated circuit; it
`was he who figured out that transistors could be lithographically
`etched onto silicon chips, then miniaturized many times over, thus
`permitting dozens of transistors and resistors to exist on a single chip.
`This invention made possible the drastic shrinking in size of com-
`puters, laying the foundation for the personal computer industry. At
`his death in 1990, Noyce was Chairman of Sematech, a consortium
`
`1:
`
`
`
`

`
`
`
`THE W513 OF INCLUSION
`
`seeking to build American competitiveness in high technology by
`pooling research on costly ventures such as supercolliders and super-
`conductive ceramics. The collaborative and integrated nature of this
`effort was reflective of the legacy Noyce left behind at Intel.
`Intel’s other cofounder—and still
`its Chairman—was Gordon
`
`Moore. While at Fairchild, he had formulated what would become
`known in high—tech circles as “Moore’s Law,” which holds that the
`number of transistors on a chip will double every eighteen months,
`which has the effect of continually reducing prices. As a chip loses
`value, it in consequence drives down the price of any product of
`which it is a component, spelling financial disaster for even a success-
`ful product if it is not upgraded or replaced regularly over time. This
`quirk means that high—tech value is always defined in terms of how
`new a product is, rather than by how much it costs to produce.
`Moore’s Law was the first formal recognition of a phenomenon that
`has had tremendous implications for high—tech development, for it
`made clear that economies of scale would no longer apply in the post-
`industrial world.
`This reversal of a basic premise of industrial economics has oc-
`curred because, whereas high—tech development costs are high, pro-
`duction costs continually become cheaper, while the cost of the natu-
`ral resources that comprise the products is essentially negligible. A
`corollary to Moore’s Law therefore holds that, if values can no longer
`be wrung out of mass production, they must be created by continually
`upgrading, improving, and modifying an existing product. The impact
`of this truth cannot be overestimated. It means that high—tech organi-
`zations, in order to survive, must incorporate continual change as part
`of their daily process: they barely have a month to sit back and basic
`in success. For no high—technology product can be considered success-
`ful in and of itself; each must be understood as part of the learning
`
`curve for what will he produced next.
`Under the leadership of Noyce and Moore, Intel patented the
`first microprocessor, which mimics the mainframe’s central process-
`ing unit by integrating logic, memory, and communication chips. The
`microprocessor made it possible for small computers to handle busi-
`
` .
`
`:—4:.~..:,
`
`

`
`
`
`60
`
`ness functions, which enabled desktop machines to move beyond the
`hobbyist phase. All subsequent advances beyond the mainframe have
`been based upon this invention, the microprocessor being nothing less
`than the computer’s brain. Having patented this essential device in the
`early years of its existence, Intel moved quickly to capitalize on its
`production, its greatest coup being its agreement to produce chips for
`the IBM PC.
`By the mid-late eighties, Intel had established the standard with
`its 286 microprocessor, known throughout the industry as “the 286.”
`It comprised the guts not only of the IBM PC and all its clones, but
`those of scores of other computers——more than 100 million machines
`around the globe. The success of the 286 provided Intel with the huge
`reserves of cash it needed to fund the costly research that enabled the
`continual upgrading and improvement of its whole range of products.
`For given the ruthless exigencies of Moore’s Law, which decrees that
`any high—tech innovation loses value the moment it hits the market,
`the company had no choice but to make enormous investments in
`technology, continually developing new products, new versions, new
`upgrades, and putting them out on the market.
`
`THE THREAT or DISASTER
`
`By late 1988, the time had arrived for the company to introduce the
`next generation of microprocessor, the successor chip to the 286. The
`atmosphere at Intel as it prepared for the release is recalled by people
`in the company as euphoric: after years of work, the 386 was finally
`ready to hit the market. Everyone was sure that Intel had come up
`with another winner, for the’ 386 microprocessor was far from being a
`
`mere upgrade.
`This represented a major step forward into the future, a revolu-
`tion in what desktop computers could do. The new chip had 32-bit
`capacity instead of the standard 16-bit——a tremendous increase in
`terms of speed, capacity, and power. Developing it had been a great
`technical success, in the tradition pioneered by Robert Noyce. Every-
`
` .
`
`_ ‘_;>_¢,
`
`

`
`THE W513 or INCLUSION
`
`one at Intel—the engineers, the marketing department, and the great
`international sales force~—was convinced that once the 386 got out
`
`there, it would sweep the field.
`Instead, a few months after its introduction, the chip was threat-
`ening the company with the specter of almost total failure. The prob-
`lem had nothing to do with the quality or attributes of the 386. Intel’s
`customers, it seemed, were attached to the 286, which was viewed as
`the workhorse of the personal computer industry. The 386 was re-
`garded as being useful only if you needed tremendous power; it was
`also considered very costly. The resistance from the market was of
`course potentially disastrous for a high-tech company like Intel, and
`pointed up with particular poignancy the paradox inherent at the
`heart of Moore’s Law. This is that failure can result from any too
`
`successful product, since its very success will discourage its buyers
`from wanting to replace it.
`When a new product meets market resistance, the producer will
`very often try to find a way to cut its price, and this is what Intel at
`first sought to do. The company’s executive committee sent the engi-
`neers back to the drawing board to develop a scaled—down model of
`the 386, a version that would have most of its power and could
`perform most of its functions, but could be priced not much higher
`than the old 286. An engineer named Dennis Carter, then in his late
`thirties, was put in charge of the effort; be had been working as
`Andrew Grove’s Special Assistant for Technical Affairs.
`“We came up with a chip we called the 386SX,” he says. “For
`what we were trying to do, it was absolutely the perfect product—
`scaled down, but still a 32-bit chip. The idea was just to introduce
`people to 32-bitness, get them hooked on that, make 16-bit obsolete
`forever. We were so excited and proud that we’d got it right. So we
`put the new chip out there, and then all stood back like good little
`engineers and waited for it to sell. But we still couldn’t move it! No
`one wanted the thing—no one.”
`.
`Dennis Carter felt that it was incumbent upon him to figure out a
`solution—or at least to get an inkling where the problem might lie.
`Intel simplyhad too much invested in 32-bit technology to accept the
`
`
`
`

`
`
`
`- possibility of its defeat. The company’s position was particularly pre-
`carious because it had just emerged from a major recession that had
`devastated the computer industry in the mid-1980s. Considering the
`
`problem from a variety of angles, Carter tried to imagine just where
`the hitch might lie. The market had rejected an excellent product: so
`might not the problem lie in who constituted the market, in just who
`Intel defined as its customer base?
`
`Intel had always defined its primary customers as computer man-
`ufacturers, companies like IBM and Dell and Compaq that bought
`microprocessors for use in the assembly of their products. These
`“original equipment manufacturers,” or OEMs as they’re called in
`Silicon Valley, have traditionally comprised the major market for mi-'
`croprocessor chips. Intel had other, secondary customers as well:
`manufacturers who sold circuitboards to OEMs, and distributing
`
`agents who acted as microprocessor brokers for both circuitboard
`makers and computer manufacturers. But the OEMs had always been
`Intel’s main customers, and it was in order to serve them that the
`
`company deployed and trained its huge sales force.»
`Certainly, Intel had never considered the people who bought
`computers to be its customers. Nor did it envision itself as selling to
`MIS professionals, those technology buyers who since the advent of
`the IBM PC had purchased desktop units for large companies. The
`consumer market——“end users” in Valley jargon—were viewed simply
`as customers of Intel’s customers, encountered necessarily at one re-
`
`move. After all, Intel was a technology company, not a maker of
`consumer products. Since their invention, microprocessors had al-
`ways been sold as components, internal parts used in the assembly of
`a larger product. It was considered inconceivable to think in terms of
`selling them to the public, comparable to the manufacturer of auto-
`mobile steering columns trying to sell its products directly to drivers
`instead of to GM or Honda.
`
`And yet, as Dennis Carter considered it,
`the steering column
`analogy was not really accurate, for the microprocessor chip is hardly
`just another component in a complex product. It is, rather, the essen-
`tial element, the piece that powers the product of which it is a compo-
`
`
`
`

`
`
`THE WEB or INCLUSION
`
`nent; the rest is just plastic shell, and an operating system that inter-
`prets the chip’s commands. Thus the microprocessor is what the
`customer is really buying when he or she purchases a computer-even
`if it had never been considered in that way. And so it began to occur
`to Dennis Carter that perhaps Intel was being stymied in its efforts to
`sell the 386 because the company misperceived who the customers for
`it actually were.
`“What we began to realize was that 32-bit architecture was great
`for the computer user; since you can run a lot of very powerful and
`useful programs off it. But the computer manufacturers were doing
`just fine selling products based upon the 286—which not incidentally
`was also becoming cheaper for them to buy. They were doing very
`well on it, so they didn’t feel any incentive to spend a lot of extra
`money just to make their products more useful to the buyer. When we
`considered that, we began to wonder if we weren’t making a funda-
`mental mistake, counting on the OEMS always to represent our inter-
`ests in the market. Here was an obvious case of their interests not
`necessarily coinciding with ours. But since we’d always regarded them
`as our real customers, that took a while to see.”
`Still, Intel had to start selling its 386 chips. Moore's Law decreed
`that the company must move the technology on to the next stage, or
`be stuck selling a product that was losing profitability with each pass-
`ing day. Adding to the sense of urgency was the fact that Intel’s engi-
`neers believed that the PC manufacturers who were their customers
`Would themselves become obsolete, unless they made the step beyond
`the 16-bit chip in a timely fashion. If they did not, the market would
`simply move to other kinds of technology. “It began to dawn on us,”
`says Dennis Carter, “that our only real hope was to appeal directly to
`PC buyers. If we could convince them of the benefits of 32-bit
`strength, they might start to demand boxes that ran on the chip. And
`if that happened, the buyers could in essence pull the market for us,
`forcing manufacturers to start using the 386.”
`The notion of reaching out to end users was revolutionary for a
`manufacturer of computer components, and would entail a complete
`realignment of how Intel positioned itself in the market. As it was, the
`
`
`
`

`
`.4-«an
`
`64
`
`company had no links to computer buyers and no routes of contact
`for reaching out to them, since its sales staff was entirely focused on
`manufacturers. Intel had no relationships it could leverage, no credi-
`
`bility with the general public, no history as a consumer products
`company; it didn’t even have an advertising agency at the time. Nor
`did Dennis Carter, as the President’s Special Assistant for Technical
`
`Affairs, have the authority to undertake marketing microprocessors to
`
`the general public.
`Going directly to the public would entail the kind of total re-
`orientation that most large companies would be unable even to con-
`ceive of putting into practice. One thinks of the Detroit auto makers’
`initial response when the market began to reject its cars: they tried to
`cut costs so that they could reduce their prices, but for a long time
`continued making and marketing cars in the same old way. Reconcep-
`
`tualizing to whom a company sells means re—envisioning why the
`organization is in business, which calls into question its very reason
`for existence. Doing so in turn undermines the prestige of the chain of
`command, unsettles established turf, and exposes an organization to
`
`incalculable risk. However,
`
`the very crisis that Intel had passed
`
`through during the mid-1980s recession made possible the kind of
`
`radical repositioning that Dennis Carter had begun to believe might
`
`be the only Way to save the 386SX. By radically opening up the
`company and creating a flexible structure that let resources flow
`toward problems, Intel had set the stage for webs to form at the
`company’s roots.
`
`STOP DOING WHAT'S STUPID
`
`The problem facing most companies, according to Carlene Ellis, In-
`tel’s Vice President for Organization and a member of the sixteen-
`
`person Executive Staff, can be summed up fairly simply: "You just
`have to find a way to stop doing what’s stupid.” That is, you must
`constantly question the utility of what is driving the organization-
`not only in terms of the present, but also in regard to what lies ahead.
`
`
`
`

`
`
`
`ganizations are far too concerned with maintaining
`' Most people in or
`the status quo to be able to do this——in particular, the people at the
`come invested in doing what’s been done, in defending
`top. People be
`its need to be done, because it justifies what they have been doing.
`This prevents them from questioning the basic underlying assump-
`tions that lie at the root of their enterprise. They don’t ask, should it
`be done at all?
`That Intel persistently asks questions is due to an earlier baptism
`by fire. “You have to go back to 1985 if you want to understand why
`we were able to handle the problems around the 386,” Carlene Ellis
`points out. “Intel had always been this growing, successful company.
`We were known for having a growing, stable employee base——we
`e stayed. And believe me, we benefited enor-
`were a place where peopl
`ple
`mously from having such a low turnover, because only when peo
`really know their way around can they figure out how to get stuff
`done here. So most of us assumed this was just the nature of our
`company. We took our stability for granted, without really having
`earned it.”
`in early 1985, the whole industry “suddenly just went
`Then,
`through the floor. Everyone got caught with a massive overstock of
`microprocessors, including us. It was a terrible crisis, we just couldn’t
`sell our stuff, and to deal with it we had to let thousands of employees
`go. The company was just torn apart—great people thrown out on the
`street, very bitter. Those left behind were as scared as those who left.
`They couldn’t concentrate on their work. The motivation was just
`gone. Fear was up as people waited for the next hit. Here everyone
`had all worked so hard to build up trust over the years, and now we
`were watching it be destroyed overnight. Seeing what happened
`scared us to death.”
`As head of CIS at the time, Carlene Ellis was in a particular
`position to feel the pain. She was sensitive to it because of what she
`had witnessed growing up in a small town in the South. “I remember
`like yesterday how NASA suddenly cut thousands of people from the
`space program. Huntsville, Alabama, just fell apart—the town, peo-
`ple’s families. You’d see all these incredibly talented engineers walking
`
`THE WEB or [NCLUSION
`
`

`
`.«.
`
`66
`
`around doing nothing. I remember it gave me a feeling of terrible
`
`waste.” At Intel in the mid-eighties, Carlene Ellis feared reliving that
`
`experience, and became convinced that “we were watching something
`that we could never let happen again."
`
`Of course, layoffs had always been considered part of organiza-
`
`tional life, a given; inevitable because the nature of business cycles is
`
`to expand and contract. “But what we realized in about 1985," says
`Carlene Ellis, “was that while that might have been all right for indus-
`
`trial factories—though I doubt it——it’s not acceptable for a company
`like Intel. Whenever you have highly trained people, you’re just shoot-
`ing yourself in the foot when you have to fire them. They comprise the
`value of your company, so it makes no sense. Also, if you’re going to
`ask people to be creative, you have to provide an environment that
`inspires them. People have to be motivated if you want them to think,
`and constant fear destroys motivation.”
`
`The lesson of the late eighties for Carlene Ellis, then, was that
`“layoffs are a last resort. They destroy your foundation and it’s tough
`to rebuild. Once you make the commitment to try never to let people
`go, however, you have taken the first step toward reconceptualizing
`how your organization works.” With support from the Executive
`Staff, Carlene Ellis’s Human Resources department began in 1990 to
`formulate a plan so that “ideally, we would never have to go through
`layoffs again, no matter what the external situation. The fact that the
`economy was cyclical could not be used as an excuse.” In order to
`achieve the goal, however, “We had to completely change the way we
`were thinking. We had to start looking at everything from the stand-
`point of where we were going to be in the future; We had to know
`what kind of skills we were going to need. Not just for the next year
`
`or so, but way down the road as well.”
`The goal became never to hire permanent people whose skills
`would become obsolete, or who would not be needed if and when the
`industry hit a downturn. “Once you decide to start thinking of it like
`that, you have to really start listening to people——your administrators,
`analysts, planners, designers. You have to find out exactly what they
`need, in terms of skills and support. You have to challenge them to
`
`
`
`
`
`
`
`‘_______"_______,,...,..,..-._...._._...__........
`
`
`
`
`
`

`
`
`
`THE W513 or INCLUSION
`
`think about what their work will be like in the future, which of course
`gets them thinking about the company, and the industry, as a whole.”
`This in turn drastically alters the role played by Human Resources,
`which is no longer in the position of simply trying to find good people
`and then hiring and training them in an all-purpose way. Human
`Resources becomes instead more the partner of the business units,
`prodding them to think in an integrated and large—scale way about
`what they will be needing over the long term.
`Once people are hired, they must then be trained in a broader
`sense of skills, so they can perform a lot of functions and be shifted
`around with relative case. “You can’t afford to hire someone who can
`only function as a process engineer, for example. You have to train
`that person in design engineering as well. That means moving him
`around a lot, which therefore means he has to get‘ his training while
`he’s on the job. The whole trick to avoiding layoffs is that you have to
`be both lean and broadbased, having lots of utility people. When I
`think back to what I saw at NASA, I’m sure one of the problems was
`that everyone was trained very narrowly, for just one specialty.”
`Carlene Ellis makes clear that Intel’s culture of constant reorga-
`nizations and broadbased training is not only a strategy for ensuring a
`work force that is flexible and lean, but also a way of building in a
`bias for the pragmatic and tactically-based approach. With jobs as-
`sumed on a provisional basis, and training conceived of as ongoing,
`people get used to trying things out and learning from whatever
`works. Such an approach, she points out, facilitates innovation by
`encouraging people to take responsibility for projects as a whole,
`rather than focusing only upon the part that they have to play. The
`emphasis on strategy built from tactics thus encourages large-scale
`thinking among people throughout the organization, by spreading
`.responsihility and the opportunity to improvise around.
`As a former school athlete, Carlene Ellis is fond of observing that
`Intel “plays by sandlot rules”——the stress is on working together, on
`learning and improvising within the context of play. “When I played
`basketball in college,” she recalls, “I always hated the zone defense.
`You just stood there guarding, no matter where the ball was. Lots of
`
`_
`
`
`
`

`
`
`
`68
`
`companies play that way today: the market moves or disappears en-
`tirely, but they’re still standing there, guarding this empty spot. They
`do it because that’s what they’ve been taught to do, and no one has
`thought to give them permission to stop doing something because it’s
`stupid. But when you move beyond the zone, and start moving the
`ball, all you’ve got is this core of people, with everyone accountable
`and responsible, all over the ball. And that’: where the technology has
`moved us. We don’t have time to play the zones. It’s all one—on-one
`out there today.”
`
`WEAVING THE WEB
`
`Intel’s restructuring in the wake of the 1985 recession made it easier
`for internal webs to form by accentuating those aspects of the com-
`pany that permit and encourage alternate centers of power to develop.
`But the recession also paved the way for what was to follow by forc-
`ing Intel to reorient itself, to question the bedrock assumptions of its
`enterprise. The depth and level of questioning about such basic issues
`as layoffs prepared the way for Dennis Carter to question just who
`Intel’s customers really were, and in the process of doing so,
`to
`reinvent the market.
`
`In early 1989, Dennis Carter went before the