CHAPTER 3

A New Perspective of Change for the Artificial Intelligence Age

Krishna Raj Bhandari

Introduction

With the accelerated pace of technological change triggered by artificial intelligence and cloud computing, the existing theories of change in developing and implementing strategy have fallen short. Not only are competitive landscapes changing, but products are also transforming the value chain. New business models are created and empowered by digitalization. In this new era, products are enabling the revolution by unlocking new value and transforming both companies and competition. The emergence of products and services enabled by digitalization moved several product and software firms from startup stages to mainstream organizations. However, models of change need to be augmented with relative mass flourishing, real option reasoning, and the CEO’s attention. Also, leveraging contingencies with a positive mind-set is essential. This chapter proposes that relative mass flourishing, real option reasoning, and the CEO’s attention together with leveraging contingencies are critical resources in the face of change to make the conversion process successful.

Human race is at the cusp of revolutionary change enabled by artificial intelligence (AI). This wave of change creates both opportunity and threat for the 21st century organizations. The opportunity is obvious as it enables new growth avenues, allows new business models, and creates competitive advantage. However, threat arises due to the pace of change and hastily done adaptation without proper thought process or solely based on hype. This is a grave concern in an AI economy since past hype is a present reality. Many firms are unprepared. The complexities are many: AI still lacks governance, is emerging too fast, and is unlike anything seen before.

According to Schildt (2016) in the age of AI, big data and complex algorithms are the cornerstone of software and can help do complex tasks such as driving cars. For sure, automation is one way of optimizing and building open systems. Algorithms pave the way for new approaches to organize work. Following Schildt (2016), this chapter touches upon optimizing-oriented and open-ended systems that benefit from big data and algorithmic management, which triggers a wave of change and transformation. In optimizing-oriented systems, the goal is to have algorithmic management of human work enabled by numerical data. However, in open-ended systems, the goal is to search for response to a wide range of managerial questions enabled by textual data or visualizations so that a new definition of tasks and resource allocation is possible. The wave of “computer assisted transparency” created by algorithm-processing conversations is possible but has advantages and disadvantages. Thus, nurturing advantages and minimizing disadvantages should be the goal of change management in the AI age.

According to Phelps (2013), the prosperity era in the history of ­capitalism was possible due to mass flourishing, a concept for mass innovation anchored in creativity, individualism, vitality, and self-expression. Though the history of capitalism illustrates this phenomenon, in the age of AI, the rationales seem different. Apple’s app store-based business model captures the mass flourishing of developer community combined with Apple’s corporatism, which resulted in a new level of dynamism in the mobile industry. The future of AI will enable such dynamism, which the author calls relative mass flourishing. The operational definition of the relative mass flourishing is the ratio of level of mass flourishing divided by the total sum effect of level of mass flourishing and level of corporatism. In this notion, the benefits arise from balancing the level of mass flourishing with a bit of corporatism.

According to Brynjolfsson (2013) the new machine age is “digital,” “exponential,” and “combinatorial.” It drives abundance because the digitalization drives “almost” zero-marginal cost production and distribution. The innovation becomes “combinatorial” as previous products or platforms enable next generation of innovation. Even on its own, these trends are powerful. When they are combined, a wave of change emerges. The implications of this wave of change to business and society are so dramatic that it demands attention of the best minds that would be able to align and adapt to this new normal. Additional complication arises from the fact that the lessons acquired from the traditional software development are not easily transportable to machine-learning solutions. The best path is to race together with AI not against it. A key question is how should companies embrace change when little experience or history exists and uncertainty is high?

In the face of change and high uncertainty, the power of experimentation and simulation is one of the answers. Experimentation and simulation are excellent ways to feel one’s way through when the rate of change is too fast to comprehend. However, for it to be effective and truly become part of the DNA of a firm, experimenting and simulations must become part of an operating philosophy of an organization. Hence, it should be led by the board and CEO with real option reasoning. Being a CEO will be challenging under these circumstances of massive change, as they would have to preserve the tradition while experimenting with the new. A method exists with which CEOs can do both. Such a method is presented in this chapter. There is a need for experimentation, learning, failing forward, and leveraging contingencies where CEO’s attention and decision are critical resources.

The emergence of “smart, connected products” has transformed the competition. The products have become complex systems that combine hardware and software with cloud connectivity empowered by improvements in processing power and device miniaturization (Porter and Heppelmann 2014). Not only is the competitive landscape changing, but products are transforming the value chain and business models are growingly enabled by digitalization. Thus, digitalization is redefining industries and forcing companies to rethink everything they do throughout the value chain and even in the value networks. The changes are not only happening on the activities stream of the organization but there are profound implications for how work is organized such as cross-functional collaboration and creation of new functions. Though the change wave has just started, CEOs are required to focus their attention on this new emerging paradigm (Porter and Heppelmann 2015). A note of caution is that “smart, connected products” are not synonymous with the ­Internet of things (IoT). In the new era, products themselves are enabling the revolution by unlocking new value and transforming both companies and competition. Earlier waves of digitalization change were focused on ­value-chain automation and integration. The current paradigm is different as it embeds into products, creating added value and changing the locus of competition (Porter and Heppelman 2015).

The emergence of new wave of products and services enabled by the digitalization wave brought proven product and software development practice from the startup world to the mainstream organizations. A new movement called “Lean Startup” (Ries 2011) has been in the forefront of product development in the software industry. The core of this movement is to develop new products through “minimum viable” ¹ development approach where a “build-measure-learn” loop is executed in a fast pace enabling experimentation and learning to avoid product failures in the long run. However, experimentation must be augmented by simulation to make sense in the algorithm-as-service business models in the age of AI.

In an environment where competitive dynamics are changing constantly, change is the new constant. Strategic management literature highlights dynamic capabilities in order to “sense,” “seize,” and “orchestrate” resources (Teece et al. 1997) and to gain a sustainable competitive advantage. However, as critically evaluated by Andreeva and Ritala (2016) there is a vacuum in understanding where these capabilities originate and how their dynamism lasts for a long time. Building further from the lean startup thinking, this chapter introduces the solution for the dilemma of origination of dynamic capabilities, called as “Lean Capability.” Therefore, reasoning with the dynamic capability-based view (Teece et al. 1997; Teece 2007; Teece 2014), it is argued that “lean capability” anchored in the hypothesis-driven development where relative mass flourishing, CEO’s attention, and real option reasoning as core elements are the new perspectives of change in the AI age. However, in this conceptualization “lean capability” goes beyond learning loops of lean startup thinking and encompasses Schildt’s (2016) optimizing-oriented and open-ended systems where algorithm-as-service-based business models are possible for sustainable competitive advantage mainly driven by real option reasoning and attention as valuable, rare, inimitable, and nonsubstitutable resources (Barney 1991).

A New Perspective of Change

Combining hypothesis driven development (Taylor 2011; O’Reilly 2013) and the lean-startup thinking (Ries 2011) and arguments from Humble, Molesky, and O’Reilly (2014), the following new perspective of change for the AI age is proposed. In this framework, relative mass flourishing, CEO’s attention (Ocasio 1997), and real option reasoning based on real option theory (ROT) (Trigeorgis and Reuer 2017) as the central elements and leveraging contingencies (Sarasvathy 2001) as a lever are the major contributions of this chapter. This chapter departs from all existing models that do not explicitly discuss the role of CEO’s attention ­(Ocasio 1997) nor make real option reasoning as their reasoning instrument under the conditions of uncertainty or leverage contingencies for making a change project successful. The loop follows steps that are constantly moderated by rare resources—relative mass flourishing, CEO’s attention, and real option reasoning.

McAfee, Brynjolfsson, Davenport, Patil, and Barton (2012) see big data as the next management revolution. Their argument is that big data enables better decisions by enabling evidence-based decision making but not on the intuition. This means today’s concepts of decision making will change. For example, Google and Amazon benefitted from this revolution already but potential for other companies is equally good. However, if CEOs and senior managers do not adapt to the evidence-based decision making based on real option reasoning to build experimentation, learning, and optimizing culture, the new transformation will be challenging. The role of data scientists in generating patterns and meaningful decision-making evidence is becoming crucial for the 21st century organizations. This change should reflect in the entire organization and a new meaning of “judgment” should be developed and contemplated.

Kolbjørnsrud, Kolbjørnsrud, Amico, Amico, Thomas, and Thomas (2017) argued that AI is becoming a cornerstone in transforming the nature of work and in creating a symbiosis of relationships among human beings and machines in organizations. However, managers are not ready. AI will change the reporting culture and nature of work itself. This transformation is even larger than the industrial revolution. Focusing only on managers to adjust to this new reality with their attention and real option reasoning-based thinking in the decision-making is not sufficient. To be successful in such a revolutionary change, top-level executives must empower society through relative mass flourishing.

1. Relative mass flourishing, real option reasoning, and CEO’s attention
2. Observe
3. Hypothesize
4. Design experiment
5. Identify the metric
6. Conduct the experiment
7. Evaluate or learn
8. Decide
9. Pivot
10. Leveraging contingencies

In the following section, these phases of hypothesis regarding new theory of change as shown in Figure 3.2 are described.

Relative Mass Flourishing, Real Option Reasoning, and CEO’s Attention

As discussed earlier the main thrust of this new perspective of change is to bring back the relative mass flourishing, real option reasoning, and CEO’s attention at the center of discussion (see Figure 3.1 for a detailed real option reasoning approach and Figure 3.2 inside the hypothesis-driven learning, optimizing loop). Balancing the need for a new business model where mass flourishing is possible and at the same time an optimum level of corporatism is utilized would be the CEO’s agenda as Apple did through App Store. However, without the attention by the CEO and in the absence of real option reasoning, change projects aiming for mass flourishing will fail. These three elements become the DNA of companies in the AI age to run fast experiments to learn, adapt, and optimize. Also, as evident from the model, the current reality is that information technology is not only a function anymore. It pervades and changes the business model and hence changes the business strategy completely.

Figure 3.1 Real option reasoning (Adapted from Trigeorgis and Reuer (2017, p. 53) inspired by Trigeorgis and Baldi (2013))

Figure 3.2 A new theory of change (Inspired by Phelps (2013); Ries (2011); and Humble et al. (2014)): Relative Mass Flourishing, Real Option Reasoning, and CEO’s Attention

The creation and implementation of strategy are normally top-down but in current reality bottom-up approaches are needed. Thus, the CEO’s attention is even crucial so that he does not miss the boat of change. In real option reasoning, being competitive and flexible at the same time is a trade-off and CEO’s role in brokering this balance is crucial as shown in quadrant II in Figure 3.1. The focus of the organization should go beyond learning loops to optimizing loops enabled by machine learning and algorithm-based programming.

Observe

The first step in the loop is to gather information, observe patterns, and make sense of the current operating environment and the business itself. Most of the time executives are not alert on what is happening. While observing, the CEO’s attention overload deters them from focusing on the right change loop. To avoid attention overload, multiple but quick loops must be done. The loop that validates hypothesis must be pursued and the others must be discarded. However, the goal is to enable relative mass flourishing with algorithm-as-service business model.

Hypothesize

Most of the change management processes are without proper hypothesis to test and learn from. Therefore, once the operating environment and business itself are observed in step “A,” a plausible hypothesis to test and learn from should be developed. The hypothesis development as other processes needs to be verified by the CEO, anchored in real option ­reasoning and furthering relative mass flourishing.

Design Experiment

This is the very minimum viable experiment with limited resources to test the hypothesis and learn from. Not only the organizational resources but also the CEO’s attention as a resource are scarce and need to be managed well. Therefore, rather than developing a full product or service, the idea is to build a prototype or a smaller version of the test item to test the hypothesis on the riskiest assumptions and understand the numbers behind the test.

Identify the Metric

The goal of the experiment is to measure progress. One needs to define actionable metrics (not the vanity metrics) for the hypothesis testing. If a right metric is not in place, possibility of learning will be missed. Therefore, it is very important to set the metrics in the design phase already.

Conduct the Experiment

Once the metrics are identified, the next step is to conduct the experiment. For sure, CEO’s attention is needed in this phase as well as it was needed for the other phases. The goal is to fail safe and fail fast. The culture in the organization must embrace these initiatives. Otherwise it will not succeed.

Evaluate/Learn

As mentioned in the start of the experiment, the goal is to have a validated learning as a measure of progress as suggested by Ries (2011). Therefore, after the experiment, there is a need to make sense of the metrics being tracked. This phase is crucial in making the next move.

Decide Through Real Option Reasoning

Though real option reasoning to balance competitive moves with flexibility is always there, this is a particular phase where, once the hypothesis is proven right, one must decide to go further in evangelizing the change program. If not, the next phase of “pivot” needs to be executed to go through the change loop once again. Leveraging contingencies are a rare resource in the face of uncertainty and at this stage pausing and digesting the numbers to nurture intuition are highly recommended. As the decision making under uncertainty is hard, the strategic management literature suggests that ROT (Trigeorgis and Reuer 2017) is a good solution as shown in Figure 3.1 earlier. Quadrant II is important because the strategic dilemma of competition versus cooperation is high and at the same time commitment versus flexibility dimension is also high. The nature of change in the AI age is uncertain and betting on only one option is not preferred.

Pivot

If the decision in the previous step shows that the minimum viable experiment did not satisfy the need, then the next loop needs to be initiated with a new set of hypotheses. This is called pivot by Ries (2011). The “pivot” can take us to new sets of hypotheses and tests. Or this “pivot” becomes a new scalable solution for customer creation and company building (Blank 2013).

Leveraging Contingencies

Upon reflection from the effectuation theory (Sarasvathy 2001), in entrepreneurial opportunity creation the role of contingencies is central. Those who would like to manage well in the face of uncertainty, those entrepreneurial managers need to be able to leverage contingencies (both internal and external) to become successful. In articulating this phenomenon, while designing experiments or at the final decision point, reflecting on the role of contingencies is wildly important. This becomes a rare resource while operating under extreme uncertainty as is the attention.

Discussions: Revisiting Porter, Barney, Phelps, and Teece

As discussed earlier, relative mass flourishing is measured as the ratio between the level of mass flourishing divided by the total sum effect of level of mass flourishing and corporatism. In this conceptualization, mass flourishing at the expense of corporatism will not result into economies of scale nor economies of scope. This chapter goes a step further on the Phelps (2013) conceptualization of mass flourishing and creates an index on balancing mass flourishing with corporatism. Too little of it or too much of it does not lead to an optimum level of firm performance. One example is how Apple has developed a developer community with a business model anchored in the App Store. This is a form of mass flourishing and dynamism in the developer community but combined with the corporatism of Apple as an orchestrator. With the advent of AI, such business models could be scaled and nurtured well. However, as shown in Figure 3.3, at low level of AI the curve shows lower performance but at high level of AI the performance curve shifts to the next level. This simulation suggests that while balancing mass flourishing and corporatism are needed but at the same time level of AI needs to be assessed. However, the pattern of too much or too little of anything is detrimental to performance is valid even in the presence of AI or absence of it for that matter.

Figure 3.3 Impact of level of artificial intelligence on the curvilinear relationship between relative mass flourishing and firm performance

What would the key strategic management discourse offer about this phenomenon and new perspective of change? Looking from the lens of Porter, he is not ready to change his old economics-based thinking as demonstrated by the latest papers on the impact of information technology on the value chain and strategy. Similarly, Barney’s thinking has not evolved since 1991 except it has been labeled as tautological by some authors. Teece has ventured into theory of the multinationals in his 2014 publication. But looking from the dynamic capabilities standpoint (Teece et al. 1997; Teece 2007) one can feel that there are elements of sense, seize, configure emerging into new models albeit with a new purpose.

Thus, this chapter highlights the need for a new perspective of change as a process that managers and policy makers alike can take in building the 21st century organization. However, the addition to this debate are the core circle of relative mass flourishing, real option reasoning, CEO’s attention, and leveraging contingencies. CEOs have limited attention capacity and they need to sense, seize, and reconfigure the resources properly in a fast experimentation and learning cycles so that old planning school does not become a bottleneck for the implementation of change.

Based on the discussion above, there are seven key managerial implications:

• Focus on relative mass flourishing, real option reasoning, and CEO’s attention.
• Drive the experimentation culture, reward failure, and ­celebrate learning.
• Build validated learning as a tool for measuring progress rather than existing accounting metrics.
• Find early change agents, nurture them, and use them as ambassadors of change.
• Give managers ownership and control to create and manage projects.
• Leverage contingencies and return on luck.
• Follow open innovation and open strategizing.

Policy Implications

This chapter has followed in the footsteps of practitioners and academicians alike to assert that the new artificial wave driven by “digital,” “exponential,” and “combinatorial” trend is faced with dilemmas. On the one hand productivity is rising but on the other hand the employment is falling (Brynjolfsson 2013). This is the most alarming “great decoupling” in history. There is a race between people and machines. And the machines are winning that race. The discussions in this chapter, thus, further the notion suggested by Brynjolfsson (2013) that policy makers must think to race with the machines not against the machines. In this symbiosis of man and machine no supercomputer can beat teamwork.

References

Andreeva, T., and P. Ritala. 2016. “What are the Sources of Capability Dynamism? Reconceptualizing Dynamic Capabilities From the Perspective of Organizational Change.” Baltic Journal of Management 11, no. 3, pp. 238–59.

Barney, J. 1991. “Firm Resources and Sustained Competitive Advantage.” Journal of Management 17, no. 1, pp. 99–120.

Blank, S. 2013. “Why the Lean Start-up Changes Everything?” Harvard Business Review 91, no. 5, pp. 63–72.

Brynjolfsson, E. 2013. “The Key to Growth? Race with the Machines.” https://.ted.com/talks/erik_brynjolfsson_the_key_to_growth_race_em_with_em_the_machines#t-269127 (accessed March 09, 2017).

Higham, R. 2016. “The MVP is Dead. Long live the RAT.” https://hackernoon.com/the-mvp-is-dead-long-live-the-rat-233d5d16ab02#.x9gpchhne (accessed March 08, 2017).

Humble, J., J. Molesky, and B. O’Reilly. 2014. Lean Enterprise: How High Performance Organizations Innovate at Scale. Boston, MA: O’Reilly Media, Inc.

Kahneman, D. 2011. Thinking, Fast and Slow. Macmillan. London, UK

Kolbjørnsrud, V., V. Kolbjørnsrud, R. Amico, R. Amico, R.J. Thomas, and R.J. Thomas, 2017. “Partnering with AI: How Organizations Can Win Over Skeptical Managers.” Strategy & Leadership 45, no. 1, pp. 37–43.

McAfee, A., E. Brynjolfsson, T.H. Davenport, D.J. Patil, and D. Barton. 2012. “Big Data: The Management Revolution.” Harvard Business Review 90, no. 10, pp. 61–67.

O’Reilly. 2013. “How to Implement Hypothesis-driven Development.” https://barryoreilly.com/2013/10/21/how-to-implement-hypothesis-driven-development/ (accessed March 02, 2017).

Phelps, E.S. 2013. Mass Flourishing: How Grassroots Innovation Created Jobs, Challenge, and Change. Princeton, New Jersey, US: Princeton University Press.

Porter, M.E., and J.E. Heppelmann. 2014. “How Smart, Connected Products are Transforming Competition.” Harvard Business Review 92, no. 11, pp. 64–88.

Porter, M.E., and J.E Heppelmann. 2015. “How Smart, Connected Products are Transforming Companies.” Harvard Business Review 93, no. 10, pp. 96–114.

Power, B. 2014. “How GE Applies Lean Startup Practices.” Harvard Business Review (accessed April 23, 2014).

Power, B. 2014. “How Watson Changed IBM.” https://hbr.org/2014/08/how-watson-changed-ibm (accessed March 13, 2017).

Ries, E. 2011. The Lean Startup: How Today’s Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses. New York, NY: Crown Business.

Sarasvathy, S.D. 2001. “Causation and Effectuation: Toward a Theoretical Shift from Economic Inevitability to Entrepreneurial Contingency.” Academy of Management Review 26, no. 2, pp. 243–63.

Schildt, H. 2017. “Big Data and Organizational Design–the Brave New World of Algorithmic Management and Computer augmented Transparency.” Innovation 19, no. 1, pp. 1–8.

Taylor, J.L. 2011. “Hypothesis Driven Development.” http://drdobbs.com/architecture-and-design/hypothesis-driven-development/229000656 (accessed March 02, 2017).

Teece, D.J., G. Pisano, and V. Shuen. 1997. “Dynamic Capabilities and Strategic Management.” Strategic Management Journal, pp. 509–33.

Teece, D.J. 2007. “Explicating Dynamic Capabilities: The Nature and Microfoundations of (Sustainable) Enterprise Performance.” Strategic Management Journal 28, no. 13, pp. 1319–50.

Trigeorgis, L., and F. Baldi. 2013. “Patent Strategies: Fight or Cooperate?” In Real Options Annual Conference. http://realoptions.org/openconf2013/data/papers/45.pdf (accessed June 2014).

Trigeorgis, L., and J.J. Reuer. 2017. “Real Options Theory in Strategic Management.” Strategic Management Journal 38, no. 1, pp. 42–63.

¹ In this chapter, the word “product” is dropped from the “minimum viable ­product” (MVP) to make sense of seeking the right problem to solve rather than developing a product. This resonates very well with doing the “riskiest assumptions test” (RAT) (Higham 2016).