page image

CHAPTER II - Performance By and Within 21st Century Organizations: Accelerating Innovation

The opening session of the Roundtable explored strategies for enhancing learning by embedding educational experiences in the texture of work rather than delivering education in separate formal “training” programs. Participants also considered some of the ways established companies are attempting to move toward new operational paradigms and identified some of the often powerful forces that exist within large organizations that actively work against large-scale change and how they might be overcome.

In a guided tour of the contemporary workscape, John Seely Brown, Independent Co-Chair of the Deloitte Center for the Edge, cited several interesting developments in work environments and in learning technology that hold promise of helping to expand opportunities for operating exponentially. He began by introducing two new collaborative spaces that help to accelerate learning by encouraging serendipity.

Hacker Dojo in Mountain View, California, describes itself as a “tech hub that is one part working space, one part events venue and one part maker space.” Housed in a 16,000 square foot former stained glass factory, the Dojo offers desk space and office space as well as high-speed Internet connections, a communal kitchen, a library, an electronics lab, and social spaces including a gaming lounge equipped with ping pong and pool tables. The facility also maintains a full schedule of activities that includes classes and workshops (e.g., laptop music production, product management, computer languages), talks (“how to negotiate the best salary if you are a girl,” “leading and managing innovation”), and meet-ups on a variety of topics (virtual reality, cognitive computing, big data, UX design). Founded in 2009, Hacker Dojo is the place where Pinterest was first conceived, and it served as the West Coast headquarters for the developers of the Pebble smartwatch. More recently, the Dojo was in the news as being the first location in the state of California to have a Bitcoin ATM. Hacker Dojo provides an environment that combines work, play and learning in order to encourage exploration and innovation. For many young people, the Dojo is a great place for them to learn the practical skills that they did not learn in college.

While Hacker Dojo offers individual memberships, RocketSpace has been designed to provide a home for fast-growing start-ups. The collaborative co-working space located in a 75,000 square foot, four-story building in San Francisco has an explicit set of criteria that companies have to meet to gain entry: they must be in the tech or new media business, have already secured some outside funding, and have fewer than 30 employees. The criteria are intended to attract fast-growing companies with good chances for success. According to founder Duncan Logan, RocketSpace wants to be known as a “hit factory,” and its track record suggests that it is succeeding: companies that got started at RocketSpace include Zappos, Spotify, Leap Motion and Uber. In fact, the facility’s web page boasts that its alumni include a dozen billion-dollar unicorns and that 1.5 resident companies secure outside funding every week.

Like Hacker Dojo, RocketSpace’s environment has been designed to “shape serendipity” among its residents. As one report on the facility noted, “setting a high quality bar means your company is rubbing shoulders with other successful companies, and your A-level employees are meeting and chatting with other A-level employees.” RocketSpace also offers educational programs (Startup Fundraising 101, How to Go from 100 to 3,500 Employees in Three Years), events and a support network. And in addition to providing a “curated community” for start-ups, RocketSpace also welcomes large corporations, encouraging them to visit and take up residence in order to spot promising tech concepts and develop relationships with start-ups that can lead to commercialization trials or even partnerships or joint ventures.

Taken together, Hacker Dojo and RocketSpace embody a range of strategies that are intended to spur creativity for individuals, small start-ups and even large enterprises. As examples of enriched environments, they can be seen as templates for workspaces that accelerate learning and innovation. In other words, they represent enablers of exponential organizations.

From Artificial Intelligence to Intelligence Augmentation
Technology itself is also having profound influence on the nature of work. Previous Roundtables have explored the institutional implications of technologies ranging from cloud computing and pervasive mobile connectivity, which are lowering the barriers to entry for new competitors and changing how and where work gets done, to social networks and telepresence that are enabling new forms of remote collaboration.

In his new book, Machines of Loving Grace, John Markoff tells the story of the long-standing rivalry between two diametrically opposed ways of using digital technology: artificial intelligence (AI) which strives to replicate specific functions of the human mind in software, and intelligence augmentation (IA) which attempts to create software tools designed to help people to work smarter and more effectively.1 Both strands go back to the middle of the last century and both represent ambitious attempts to push the power of computing beyond processing data to connect computers more directly to the realm of human cognition.

The field of artificial intelligence traces back to the 1950s, when the term was coined by the computer scientist John McCarthy to describe efforts to replicate the human thought processes in software. In its early years, the field seemed to make rapid progress, creating programs that could play checkers, solve certain kinds of mathematical problems, and even carry on plausible conversations with a human being. Looming in the distance was the prospect of building intelligent robots capable of autonomous action. But then progress slowed: while it was possible to build “expert systems” that could mimic the thought processes of a professional in a particular area of knowledge, creating such systems were highly labor intensive and remained restricted to narrow domains. And building machines that could perform many functions that are easy for humans—like understanding spoken language or recognizing a particular face—proved to be much more difficult than originally assumed.

But the field continued to make progress and has benefitted from the availability of vastly more powerful computing resources. In recent years, computers have developed the capacity to play chess at a very high level, to scan complex documents and create cogent summaries, and even to drive a car without human assistance. As AI has grown more powerful, it has raised the real prospect of machines taking over an increasing range of functions from people. Thus, workers need to compete not only against workers all over the world who may do their jobs as well or better than they do and for less money, but increasingly will need to compete against machines who may have the ability to outperform them and can work non-stop.

The parallel quest to augment human intelligence has received less attention than AI, but it has an equally long history that traces back to the pioneering work of Douglas Englebart whose goal was to create tools that would help individuals and teams grappling with highly challenging problems. Englebart invented the computer mouse and built interactive systems that were ahead of their time, but he failed to move beyond building experimental prototypes to develop a practical product. Like AI, progress in the field of IA was much slower than initially expected. Early attempts at creating “virtual personal assistants” such as Microsoft’s Bob (in 1995 to help users of its Windows OS) were notable failures. But like AI, the steady increases in networking and computer power has made applications of IA possible that were previously not practical. One familiar example is the emergence of a new wave of virtual personal assistants such as Apple’s Siri or Microsoft’s Cortana that enable users to control multiple functions of their smartphones through voice commands.2

According to John Seely Brown, some of the most promising current applications of technology use the power of AI to complement human intelligence rather than replace it. To illustrate, he cited the changing role of computers in the world of competitive chess—a classic challenge to the developers of AI programs. In 1996, Gary Kasparov, the then reigning world chess champion, defeated IBM’s chess-playing program, Deep Blue in a six-game match (4 games to 2). But in a rematch one year later, Deep Blue prevailed (3 ½ to 2 ½ games), the first time a machine beat a human grand master in a formal tournament.

The match was widely seen as a landmark in the evolution of AI that demonstrated that a machine was in fact capable of excelling in a challenging area where human intelligence seemed to be required. When IBM’s Watson defeated the best human players at Jeopardy! in 2011, it provided another confirmation of the potential power of AI to outperform humans. According to IBM, the triumph of Watson ushered in the age of what it has called “cognitive computing.”

But John Seely Brown noted that there is an alternative to pitting people against machines, and that is the prospect of working with machines. Back in the late 1990s, Gary Kasparov proposed what he called Freestyle Chess (also known as Advanced Chess or Centaur Chess) in which individuals or teams able to make use of any chess-playing computer programs played against each other. Kasparov believed that combining men and machines would result in chess matches played at levels never before reached by either human or machine players alone. Freestyle tournaments have taken place regularly since then, with perhaps the most interesting match being the PAL/CSS Freestyle Chess Tournament that took place online in 2005. As the tournament’s organizers noted, “The use of computers is not just allowed, it is encouraged.” Remarkably, the winner of the tournament was a team named ZackS, which was led by two young unranked amateur players, Zack Stephen and Steve Cramton, who made use of three different computer programs to defeat teams that included established grandmasters. What enabled Stephen and Cramton to prevail was not their stronger chess skills but rather their greater ability to combine the power of a computer with their own intelligence.

Another example of the power of AI to amplify rather than replace human capabilities comes from the domain of military training. With funding from DARPA, a small company called Acuitus is attempting to “revolutionize training” by using AI to provide students with highly personalized programs that allow them to learn and to practice skills anyplace and any time under the guidance of a computer-based tutor. In addition, the company is “creating an online community of instructors and learners modeled after the online communities formed around multi-player games that are so familiar to this next generation of soldiers.” According to Brown, the evaluation of an early trial found that computer-based mentoring outperformed both conventional classroom instruction and training by fleet experts by two standard deviations while substantially reducing overall training time.

In fact, we have entered a time in which a wide and expanding range of technologies are available to leverage human capabilities and enhance the value of human workers. Data is everywhere, and new possibilities exist for sensing and reacting to changes in the environment. An unprecedented amount of data—and the tools to analyze it—are available not only about “things” but about people and their behavior. In this “era of social physics” the ability to optimize work environments as well as marketing strategies can bring businesses to a new level of performance. These tools, when mobilized effectively, make it possible to operate exponentially.

Reports from the Field
Can established hierarchical organizations move from a linear method of operating to a more dynamic and highly leveraged exponential operation? How can an organization shift its core strategy from attempting to achieve economies of scale to pursuing scalable learning? John Seely Brown suggested that the biggest obstacle to overcome in achieving such a transformation is the challenge of “unlearning and reframing”—letting go of the very principles that have been responsible for past success in order to embrace a new, less familiar (and therefore less trusted) set operating principles in pursuit of goals that previously would have seemed impossible if not improbable.

Roundtable participants described recent efforts they have been involved with to innovate more rapidly or to sustain rapid growth:

  • Casey Carl, Chief Strategy and Innovation Officer at Target, agreed that the need to unlearn operational routines represents the largest source of inertia that must be overcome in order to do things differently. Ironically, people who have been the most successful, who have accumulated the most institutional knowledge, are often the hardest to change.

    One of his main goals at Target has been to “infuse a digital mindset” in a company that has long been defined by brick and mortar operations. One key to getting new answers is to ask new questions. To do this, he brought together disparate IT teams from across the company and put them under a new leader who has been teaching them to ask different questions about what they can create. Another strategy has been to create a new cohort of change agents inside the company. He began by identifying a group of “unicorns,” front line people throughout the company who have been “doing things the right way”—not the way they are “supposed to be done,” but the way that actually works. This group of highly talented people had not been recognized for their contributions, but had generally felt stifled and frustrated by how slowly things changed. He thanked them for their contributions and gave them a chance to meet and support each other in their efforts to make things happen more quickly. His next step will be to invite each of the initial group of unicorns to “bring a friend” who is also struggling with change. Rather than attempting to follow a grand master plan, Carl described his strategy of “iterating through a transformation.” His goal is to encourage staff to find ways to “disrupt themselves,” rather than waiting to respond to disruption that comes from the outside.

  • Kaiser Permanente is attempting to achieve a major transformation. With annual revenues of $54 billion and 10 million members, Kaiser is one of the largest health care organizations in the country. It has long been a pioneer in providing integrated care by combining health insurance and health care services in a single organization. Although its structure is distinctive, the way in which it has defined and delivered medical care is relatively conventional: It employs nearly 18,000 physicians and 50,000 nurses who provide care through a network of 38 hospitals and 619 clinics in eight states and the District of Columbia.

    According to Vivian Tan, Vice President for Strategy and Transformation, like other doctors, Kaiser’s physicians have been trained “to identify medical problems and solve them.” But it is becoming increasingly clear that the definition of health is much broader than most physicians recognize and the determinants of well-being go beyond those factors that they normally deal with. Like Casey Carl at Target, Tan is attempting to change the fundamental mindset of her organization: instead of focusing on discovering “what is wrong with you,” the goal is to shift the conversation with patients to exploring the question of “what matters to you.”

    Changing the mindset of physicians and other professional providers is hard to do. But Tan identified four key factors that can drive change: First is mission, having a shared goal that can provide strong support for change if a new way of doing things can be clearly linked to it. Second is affiliation, which means that if someone in the organization is already functioning effectively in a new way, there are more likely to be followers. Interestingly, the biggest influencers are not the early adopters but the “early resistors” who start out opposing change but switch to supporting it. Third is mastery, which is manifested in the visible presence of a respected change advocate who can help bring about wide adoption of proven innovations. Fourth is autonomy, the fact that people do not want to be told what to do, but can be empowered to change. One good way to do this is to include real patients in the process. For example, Kaiser brought in a group of pregnant women and new mothers to interact with an internal team working on redesigning maternity care. Their input about their own experiences proved instrumental in shifting the mindset of the care team, defusing opposition to change and resulting in a better end product.

    Tan noted that Kaiser is engaged with pursuing two different types of innovation, one that involves physical changes in how services are delivered (e.g., new types of clinics or hospital rooms), the other that is digital (new uses for electronic health records or new forms of doctor-patient communication). Although both are important and Kaiser has made progress in both areas, it has been challenging to bring these types of innovation together.

The Digital Imperative: AT&T and GE Attempt to Reinvent Themselves

Even the largest corporations in America are recognizing the need to respond to the challenge posed by the rise of digital business and are attempting to remake themselves in some fundamental ways.

AT&T, which had revenues of $147 billion in 2015, is attempting to reinvent itself in order to compete more effectively, not just with other telecom companies but with technology companies like Amazon and Google. In the past three years, the company has invested more than $20 billion annually to build its digital businesses. AT&T Chairman and CEO Randall Stephenson has announced that he expects the company’s 280,000 employees to develop new skills such as digital networking and data science. To help them make the transition, AT&T has partnered with Udacity to offer a range of online courses and will provide employees up to $8,000 per year to pay for the courses.

In early 2016, General Electric announced that it was moving its headquarters from suburban Connecticut to Boston, a move which has been described as illustrating “how much old-line companies in nearly every industry have been force to rethink their business for the digital age.” With 360,000 employees and 2015 revenues of $117 billion, GE is the nation’s largest industrial company. Nonetheless, CEO Jeffrey Immelt has stated that he expects the company to be a “top 10 software company” by 2020. To help facilitate the transition, the company decided to relocate its headquarters to the Seaport District of Boston which is a center for computer, biomedical and pharmaceutical companies. And most of the people who will work at its new headquarters will not be traditional corporate staff, but rather “digital industrial product managers, designers and developers in disciplines like data analytics, life sciences and robotics.”

1 John Markoff, Machines of Loving Grace: The Quest for Common Ground Between Humans and Robots (Ecco, 2015).
2 The final chapter in Markoff’s book tells the story of the talented team that developed the technology that led to Siri, a process that took several decades.
Title Goes Here
Close [X]