Three-sentence Summary

Bell Labs had its share of personality clashes, organizational challenges and strategic missteps. Many attributed the success of the Labs to a combination of timing, environment and curious and determined folks, but not geniuses. Ultimately what made Bell Labs’ operating model successful was a clear, shared vision.

… in any company’s greatest achievements, one might, with clarity of hindsight, locate the beginnings of its own demise [p. 186].

Introduction

Prior to reading this book, I held an unrealistic view about “super-human” researchers and that Bell Labs was some sort of utopia. As I read the book, I began to realize that my prior view was far from correct and that both the “super-humans”, were just normal — definitely curious and determined — people, often with humble upbringings and that Bell Labs had it share of organizational struggles.

Ultimately what made Bell Labs’ operating model successful was a clear, shared vision to improve global communications. Mervin Kelly was known for challenging innovations to either make communications, “better, or cheaper, or both” [p. 179].

Where I work today uses a similar operating model to that of Bell Labs — i.e., well-defined, high-level organizational goals, loose working constraints and focuses on incremental improvements. Not only is the operating model similar, but I am similarly surrounded by many domain experts.

Things that once made Bell Labs’ work environment so exceptional — such as doing “unfettered research” — later became, in general, inferior methods for sustaining a successful company. Andrew Odlyzko, a manager in the mathematics department at the “new” Bell Labs, wrote and circulated a paper titled, The decline of unfettered research, where he described why doing “unfettered research” is no longer a logical or necessary investment for a company [p. 334]. It takes too long for a breakthrough to pay off as a commercial innovation — if it ever does. The base of science is now so broad that a company can profit significantly by pursuing incremental improvements on existing technologies. Odlyzko wrote this paper around the time that Netscape had gone public, demonstrating that a lot of money could be made from building upon existing technologies, rather than discovering something entirely new.

Innovation and Execution

We usually imagine that invention occurs in a flash, with a eureka moment that leads a lone inventor toward a startling epiphany. In truth, large leaps forward in technology rarely have a precise point of origin.

At the start, forces that precede an invention merely begin to align, often imperceptibly, as a group of people and ideas converge, until over the course of months or years (or decades) they gain clarity and momentum…

Luck seems to matter, and so does timing, for it tends to be the case that the right answers, the right people, the right place — perhaps all three — require serendipitous encounter with the right problem [p. 51].

John Pierce, known for his work on pulse-code modulation (PCM) and naming the transistor, acknowledged the challenge of pursuing, or execution of, ideas. He once said,

You see, out of the fourteen people in Bell Laboratories, only one is in the Research Department, and that’s because pursuing and idea takes, I presume, fourteen times as much effort as having it [p.348].

Derek Sivers has a great short post on this, titled, Ideas are just a multiplier of execution.

Managers at Bell Labs managed ideas, rather than people [p. 348]. Pierce didn’t let people get in the way of his pursuit of an idea. He didn’t compromise because it would make someone feel good.

According to Mervin Kelly,

The purpose of innovation is sometimes defined as new technology. But the point of innovation isn’t really technology itself. The point of innovation is what new technology can do ‘better, or cheaper, or both’ [p. 343].

Pierce, expanded on Kelly’s “rule”, saying,

New gadgets or new technologies are important only when they really make good new things possible or good old things cheaper or better [p. 343].

According to Jack Morton, inventor of the microwave tube and who later became VP of Bell Labs, innovation is …

… not a simple action but a ’total process’ of interrelated parts. It is not just the discovery of a new phenomena, nor the development of a new product or manufacturing technique, nor the creation of a new market [p. 108].

Brock McMillan, research mathematician and later VP for military development, said,

… there weren’t many other departments [than Claude Shannon’s] where people just sat and thought [p.132].

Shockley pointed out that by the middle of the twentieth century, the process of innovation in electronics had progressed to the point that a vast amount of multi-disciplinary expertise was needed to bring any project to fruition and that it was increasingly unlikely that the multi-disciplinary knowledge would be known by a single person [p. 134].

It’s from individuals that ideas originate, but from groups where innovation responsibilities are transferred [p. 134].

There were times for innovative and clever solutions and there were times where simplicity and reliability trumped all else. One project where it was made especially clear by Mervin Kelly that using the “least innovative” components was desired was in the undersea cable project. One reason for the desire to keep it simple and use technology that was well understood was because it was extremely costly to make changes to the undersea cable. So to generalize this point, keep it simple where potential costs are high and innovate where costs are low.

…too much, too soon [p. 210]. Do what is simple first, then and only then, move on the more complex solutions with wider scopes. Iteration of a decent solution is often preferred to a high-cost deployment of a superior solution. This actually relates to a recent post that I read, titled Managing technical quality in a codebase, by Will Larson. I summarized my takeaways in this thread on Twitter.

In part, I think we often jump to what we believe to be the “superior” solution — which might be using some new technology, etc. — because we believe it to be easy to spin something entirely new up, rather than understand the complexities of what exist to day and to build upon that. The “superior” solution will necessarily bring its own set of complexities. Really what we should be comparing is not the potential benefits, but the list of complexities.

“Innovations” are rarely a singular, novel, breakthrough, but rather iterations over time of things that already exist.

Telstar was not one invention but rather a synchronous use of sixteen inventions patented at the Labs over the course of twenty-five years [p. 222].

Relatedly, inventions that seemed useless — or minimally impactful — upon their invention later became critical components in larger future inventions [p. 222]. This happened at least a few times for inventions at the Labs, where it wasn’t until much later that inventions’ impacts were understood and realized.

With information theory, Shannon had never had any intention of changing the world — it just worked out that way. [Shannon] had pursued the work not because he perceived it would be useful…[but] because it intrigued him [p. 318].

Shannon once told an interviewer,

I think you impute a little more practical purpose to my thinking than actually exists. My mind wonders around… I’m thinking, ‘What if it were like this?’ or ‘Is there an interesting problem of this type?’ [p. 319]."

Their solutions were constantly identifying new problems that needed to be solved [p. 47]. One of the problems that they needed to solve was to build the tooling to measure and do things such as billing, etc.

The formal purpose of the new solid-state group was not so much to build something as it was to understand it [p. 90]

One major challenge that Bell Labs faced was helping the public understand and realize the benefits of its inventions, such as the transistor. The public clearly didn’t understand the important of the transistor, exemplified by a four-paragraph write-up on page 46 of The New York Times. Marketing their inventions was something that Bell Labs, being part of the monopoly that it was, didn’t have to do all that well because everything from discovery to development was done in-house.

The People and Place

Many of the now recognizable names are of people who had humble beginnings. Many were from the mid-west — not that I think this fact matters all that much — some grew up alongside latter-day pioneers, such as Brattain, where he spent much of his childhood around heavy machinery, seemingly far from a life surrounded by bleeding-edge technology. Brattain had a professor — at his small college — who recognized his talents and introduced Brattain to an advanced physics course [p. 316]. Brattain ended his career at his alma mater, where he wanted to have the opportunity to do the same that was done for him to others. I think this demonstrates that it’s from the help of peers and mentors, that we have the potential to become “great”; and recognizing when we need to be the person helping the next generation.

In the early twentieth century, there were few reputable schools in the United States where one could do post-graduate studies in physics. Often one would have to go abroad and study in Europe, until Robert Millikan — known initially for his oil-drop experiment, which proved the existence of the electron and the value of its charge — helped the University of Chicago become a reputable place to study. The University of Chicago became a primary recruiting ground for AT&T. Millikan was married in 1902 and his best man was Frank Jewett, a man who lived in the same boardinghouse as Millikan as the two were in their PhD programs. Many of the researchers who helped Bell Labs later become the research laboratory that it was in the mid-twentieth century studied under Millikan at the University of Chicago; Harold Arnold, Harvey Fletcher, Mervin Kelly (president of Bell Labs from 1951 to 1959), Clinton Davisson and probably more that I’m forgetting to name.

It was intentional that the new home of the Labs in Murray Hill, NJ forced people from various departments to be in each other’s way [p. 37]. The hallways were long and the cafeteria was centralized.

An instigator is… someone able to wrest excellence out of people who might otherwise fall short.

An instigator is different from a genius, but just as uncommon [p. 196].

Claude Shannon, later in life, while teaching at MIT, acknowledged that he had his doubts about keeping up with the field that he founded [p. 317]. Admittedly, the field that Shannon founded, namely Information Theory, ballooned into something massive and I don’t think anyone could have kept up with it. But more generally, I think the important bit to take from this statement is that growth outpacing our ability to have am exhaustive understanding, is a good sign. This is precisely what forces us to have to organize and become something bigger than just ourselves and what keeps us humble — even as someone who founded a field that resulted in one of the largest societal shifts in history and is the foundation of many things today.

Many of the folks struggled in their new environments after Bell Labs [p. 321].

The Culture

There was a very strong culture of learning, sharing knowledge and “idle curiosity [p. 184]”. For example, there were study groups the met weekly where they would work their way through a textbook one chapter at a time, sharing the role of lecturer over tea and cookies. One of these study groups was informally led by William Shockley, who was joined by Brattain, Fisk, Townes and Woolridge, among others to discuss topics in the field that would later become “solid-state physics”.

Mervin Kelly displayed extreme managerial bravery to demote long-time senior folks and promote younger folks with fresh ideas [p. 79].

Fair Share of Challenges

The Great Depression

During the Great Depression, researchers at the labs were forced to work fewer hours. Instead of being upset, the researchers viewed this as a good opportunity to enroll in academic courses.

Organizational

While World War I and II required priority shifts, wartime is also credited for many technological and organization improvements at Bell Labs. For example, research and supply of quartz crystal increased and as men went off to war, women and Jews filled their positions — bringing more diversity. Unfortunately, there was a strain of antisemitism that ran deep at AT&T. I think the takeaway here is that you should see how you can leverage what is seemingly a “bad” situation and align it with your priorities such that the situation works in your favor.

Shockley, the solid-state group’s manager, did the unforgivable act of implementing a competing design to that of the people he managed [p. 102]. Instead, a manager is expected to guide their team toward the superior solution, or support the solution that came about. Shockley had kept his design secret until a meeting where he could tell his group was approaching a similar design to his, at which point, he literally jumped from his seat and presented his design; the unforgivable act [p. 103].

Ian Ross noted that,

The great tragedy of Bill Shockley’s life was that he did almost nothing of scientific worth after leaving Bell Labs [p. 314].

Shockley was extremely racist and became obsessed with trying to prove that race was correlated to intelligence. Many of the folks that he previously worked with at Bell Labs, already weren’t fond of Shockley because of what he had done with the transistor and his inability to receive feedback.

Shockley had few visitors while on hospice [p. 315].

As Brattain, co-inventor of the transistor, put it, by Summer of 1951,

… one of the greatest research teams ever pulled together on a problem had largely collapsed.

Personal reflection

This book documented the highs and lows and the beginnings and ends of many folks’ lives from Bell Labs made me think about what truly is important in life. Being completely honest, in recent years, one of my primary focuses has been to try to identify a problem and build a solution that would garner domain or industry recognition.

I was thinking about my mom’s funeral — she passed away from cancer in November of 2017. There were so many people at her funeral — many said it was the largest funeral they had ever been to — who all had a story to tell. My mom was a registrar — i.e., she helped students and professors with class schedules and prerequisites, etc. — at my alma mater; it wasn’t a position where she would have ever received any sort of domain recognition, yet she touched so many lives. There was an overwhelming amount of support when she was sick from many of these folks. I especially now recognize that what I want is my mom’s “legacy” over Shockley’s. The simplest thing that I can do to continue my mom’s legacy is to just focus on being helpful and cheerful, everyday, to everyone.

Something that I need to work on to uphold this legacy is being able to hold conversations with people on topics that I may not be interested in, but the other person may be passionate about. And, when appropriate, ask about how someone is honestly doing and to get to know people that I work with better. I am so fearful about overstepping that I don’t step at all and as a result, I’m not building lasting, deep, relationships.

The Monopoly and Government Regulations

In order to keep the telephone monopoly intact, AT&T made a few agreements that may have ultimately lead to Bell Labs’ demise, one of which was that they agreed to not enter the computer or consumer electronics industry.

Keeping the monopoly intact brought about challenges when releasing and/or evangelizing (or not) their discoveries and inventions due to fears related to the government breaking up the monopoly, especially if these discoveries or inventions were arguably outside the domain of communications. One invention in particular that was challenging to release (or not) was the transistor.

Max Matthews, known primarily for his work on computer-generated music , outlined how Claude Shannon’s Information Theory broken down the barriers which previously made sense for a phone monopoly to exist [p. 185].

Western Electric, prior to the split, was AT&T’s telecom equipment manufacturer. After the split, Western Electric became Lucent. Lucent did fairly well financially after the split, during the telecom and dot com booms, but by 2000, the actual market demand for telecom equipment was realized and Lucent was hit hard. While Lucent was affected by lack of market demand, the real killer was that it was discovered that they had been inflating their profits by providing customers with financing options [p. 335].