As AI evolves and new business applications emerge, so do the leadership threats and opportunities. This section discusses some important ones, including initiatives to protect privacy and security, the impact of automation on the job market, and models that can be deployed to help ensure AI is deployed for public good.
Transform the Organization
In order to enable human-AI collaboration and reap the benefits of digital transformation, leaders need to build symbiotic human-AI organizations through mutually beneficial relationships. They need to build awareness across all levels of their organizations so that AI solutions can augment rather than displace human capital. The effective collaboration between humans and machines will help leaders enhance sensemaking, visioning, relating, and inventing skills in themselves and others without losing the irreplaceable context that humans apply to ensure insights are appropriately actioned. And, “as leaders look at how to incorporate AI into their organizations, they’ll have to manage expectations as AI is introduced, invest in bringing teams together and perfecting processes, and refine their own leadership abilities” (De Cremer, 2021).
For example, a healthcare organization designed a new medical coding system that required nurses to work as trainers for the AI system. Drawing on the nurses’ desire to apply their medical knowledge in new ways, the nurses felt in control of their work, were motivated to learn new skills, and did not fear being displaced. (Wilson & Daugherty, 2019).
Leaders need to focus on enabling human and AI collaboration and helping people recognize the benefits of using AI solutions to enhance their capabilities. “In order to create a symbiotic AI workforce, organizations will need to use human-centered AI processes that motivate workers, retrain them in the context of their workflow, and shift the focus from automation to collaboration between humans and machines” (Wilson & Daugherty, 2019). As discussed in Module 3, collaboration is the foundation of collaborative leadership and a key element of AI success. Further, collaboration is built into the very fabric of xTEAMS and is manifested through the extensive coordination between core, operational, and outer-net members.
An example of an innovative solution resulting from a process of co-creation that emerged within an xTEAM is the Audi Robotic Telepresence (ART). The xTeam composed of technicians, mechanics, and AI technologists created a system based on telepresence robots that helps train technicians in diagnostics and repair. The system allows expert technicians to remotely control a robot that sees, hears, and rolls next to a technician on-site. The AI tools embedded in the ART system analyze the communication and improve the collaboration between the expert mechanic and the robotically embodied technician (Daugherty & Wilson, 2018). Co-creating the ART system helped establish a sense of trust among technicians and reduce their fear of being replaced by machines.
Build External Collaborations
The ability of AI technologies to solve problems and support humans in decision-making is “only as good as the data they have access to, and the most valuable data may exist beyond the borders of one’s own organization” (Kiron, 2017). Given the need to feed ML systems large amounts of data, companies need to open the door to external collaborations, looking for “competitive advantage in strategic alliances that depend on data sharing” (Kiron, 2017). Recall that the ability to forge and leverage external connections is also an important quality of xTEAMs. In fact, xTEAM members are chosen for their networks and their ability to access top decision-makers and specialized expertise both inside and outside the organization. xTEAM leaders expect team members to engage in external outreach to stakeholders and bring new ideas to the team from Day 1. This helps teams understand how the outside world is changing, what needs to be done, and how to shape the strategic direction of the organization in response.
When BMW, Daimler, and Volkswagen formed an alliance to create HERE, a digital mapping company based in Berlin, the goal was to create a real-time platform that used data collected from cars and trucks to track and monitor driving conditions such as traffic congestion, estimated commute times, and weather (Mohanty & Vyas, 2018). This external outreach and collaboration provided a sufficiently robust data platform and thus gave the individual companies the opportunity to create a system that none of them, alone, could have built (Kiron, 2017).
Address Security and Privacy Threats
One of the main reasons cyber security has become a US$75 billion per year industry is that threats from sophisticated algorithms have scaled exponentially, and all companies must invest in preparedness strategies. Cyber crime, a criminal application of AI that is often classified as “evil AI,”has become automated, and protecting sensitive corporate information is now the responsibility of all senior leaders, not just the IT department.
As data about workers and consumers is used widely to train ML systems, leaders need to invest in initiatives to protect individuals’ privacy and conduct frequent audits to understand what data is used and with whom it has been shared. This is more easily and effectively accomplished by xTEAMS that have clear internal processes around data governance, data privacy, and data protection.
Because data privacy is a pressing issue for consumers and regulators, privacy-preserving AI methods like federated learning will continue to gain momentum. Federated learning is “an ML technique that trains an algorithm across multiple decentralized devices or servers holding local data samples, without actually exchanging the data” (Li et al, 2020).
Beware of AI Dependence
Leaders depend on AI algorithms to conduct critical business tasks. In turn, this dependence may create “new vulnerabilities, inefficiencies and, potentially, ineffective operations. Errors and biases may creep into algorithms over time undetected, multiplying cancer-like, until some business operations goes awry” (Kiron, 2017). Leaders and team members must invest in continuous monitoring of undetected biases, starting from a careful review of the underlying data, because biases are frequently the main source of the problem rather than the algorithm itself. A solution is to build human-AI teams where machines work to fully complement human work. A complementary solution that leverages the xTEAM tool of task coordination, whereby xTEAM members identify interdependencies among tasks and mechanisms to coordinate their work, is to ensure that everyone involved in the creation or deployment of AI at any stage is accountable for considering the system’s vulnerabilities (IBM, 2019).
An example of overcoming vulnerable systems through coordination is illustrated by a recent study where radiologists were asked to examine images of lymph node cells to determine if the cells contained cancer. The same images were then processed by AI algorithms. While the human pathologists had a 3.5% error rate, the AI solution had a 7.5% error rate. A combined approach combining AI and human input resulted in an error rate of 0.5%, representing an 85% reduction in errors (Wang et al., 2016).
Anticipate Technological Unemployment
While promising to enhance human performance in a variety of tasks, AI and automation advancements are also associated with a replacement threat to business, as low-skilled jobs may no longer be necessary once organizations fully implement AI systems. AI adoption is causing many industries to cut their workforce, with the telecom industry predicted to undergo the biggest workforce reduction by 2023 (Statista, 2020).
Technological unemployment is common during any major technological change, and the current Fourth Industrial Revolution is not immune. A recent study released by McKinsey Global Institute reports that roughly one-fifth of the global workforce will be impacted by the adoption of AI and automation, with the most significant impact in developed nations like the UK, German and US. By 2030, it is anticipated that robots will replace 800 million workers world-wide (Manyika et al., 2017).
Paradigm shifts create dynamic job markets, with new types of careers that can arise after an initial adjustment. Government leaders and the corporate world need to provide education and transition assistance, investing in programs to upskill and re-skill people, especially in industries that rely heavily on AI.
While autonomous vehicles may soon replace Uber drivers, and cleaning robots may diminish demand for building cleaning services, the threat of unemployment due to AI also applies to white-collar professions. For example, Japanese insurance company Fukoku Mutual Life is replacing human agents with AI to match customers with the right insurance plans (Kiron, 2017). Another industry that could face AI-related job loss is the legal sector. Startups like RAVN ACE and NexLP are automating tasks such as legal research and contract review.
xTEAM leaders who maintain an external orientation, keeping informed of changes in technologies, markets, politics, and competition, have the benefit of spotting and responding to these trends, needs, and opportunities more quickly than traditional, internally focused leaders.
Responsible AI and Machine Learning
To remain ethical from legal, regulatory, and public perspectives, leaders in both government and private organizations must invest more robustly in research on the societal implications of AI technologies. Like any other technology, AI has the potential to be used to promote good or bad causes. An informed debate on the uses of AI should involve interdisciplinary teams from diverse backgrounds that can build a multi-stakeholder perspective on how to deploy AI in ways that enrich society. Given the complexities of implementing and deploying AI in various industries, it is key for companies to build systems that protect people and their data, prevent data exposure, and remove unfairness in AI systems, which can reinforce biases and stereotypes or withhold information from individuals (Hao, 2021).
Multi-stakeholder collaboration such as that used in xTEAMS is key to creating and maintaining trust in AI, by promoting transparency and impartiality. This multi-stakeholder collaboration is a hallmark of both responsible AI and xTEAMS. In this model, responsibility and accountability for AI risk is expanded from traditional risk managers to individuals and teams working on AI at different stages (Heires, 2021).
In 2016, Microsoft released its AI-based chatbot Tay via Twitter. The bot was trained to generate responses based on interactions with users, but when users started posting offensive tweets toward the bot, Tay began making replies that reflected the malicious content. Within hours of the initial release, Tay was taken offline followed by an official apology for the bot’s controversial tweets. This is an example of AI that required more research in AI design and AI systems feed to ensure responsible interactions with people (Lee, 2016).
Building a culture of innovation is important because it gives organizations the ability to recognize market opportunities, react to changes, improve employee engagement and motivation, and foster competitive advantage. Understanding environmental pressures propels organizations toward change and innovation.
To create a culture of innovation, leaders are often reminded of the importance of setting clear goals, encouraging open communication and active listening, and assigning tasks based on an individual’s interests and expertise (Frohman, 1998). While all of this is important, there is one skill that leaders must practice that integrates and precedes all the others: sensemaking.
Sensemaking skills help leaders increase an organization’s adaptive capacity, increasing individuals’ and teams’ ability to both respond to and frame external changes. Some organizations embrace adaptive changes when faced with external pressures, such as declining sales, reputation pressures, or regulatory changes (Ancona, 2011). For example, some companies are slow at responding to racial equity and anti-discrimination issues. Only when negative press or lawsuits become a threat would these companies establish diversity task forces and set policies for ensuring the fair treatment of minority staff.
Previous success under another business model is another reason why many organizations may not adapt and respond to changes (Palmer et al., n.d.). There are several examples of companies whose success under a traditional model made it difficult for them to pivot to meet the challenges of digital transformation, starting with Kodak and Polaroid, which missed the rise of digital photography, to Blockbuster, which succumbed to more digital-savvy companies like Netflix.
Nimble leadership promotes the development of sensemaking skills to ensure that everyone in the organization has a clear understanding of the external forces at play, and how change can promote innovation and increase organizational resilience.
Summary
Section Time: 3 minutes
As AI evolves and new business applications emerge, so do leadership threats and opportunities. Opportunities include transforming the organization, building external collaborations, and ensuring responsible machine learning; challenges include security and privacy threats and technological unemployment.
As machines take over prediction tasks, leadership skill sets will need to adjust. Leaders will need to design systems to attract and retain individuals with transferable skills such as systems thinking, creativity, ethical decision-making, and teamwork. Leaders will also need to invest in new positions that require skills and training never needed before, such as AI trainers, AI explainers, and AI sustainers.
Leaders of companies that have implemented successful AI projects also share vital principles:
- Learn with AI.
- Create the right mindset.
- Foster human-machine co-creation.
- Map existing competencies and conduct gap-analysis.
- Integrate machine learning and organizational learning.
- Invest in people, teams, and technology.
AI, leaders, and organizations can co-evolve to take advantage of opportunities and overcome challenges along the way. The evolution of AI technologies is leading to a change in the role of AI itself. AI has been depicted as a coach and a recommender for leaders, augmenting human skills and reducing complexity. In addition, a promising area for the future of AI is the ability to recognize and display human emotions. The next generation of AI systems may be able to demonstrate emotional intelligence (EI). The concept of EI is a powerful tool to help individuals become more aware of their own emotions and empathize with others’ feelings.
Nevertheless, the future of AI is shaped by some of the current limitations of deep learning, especially in supervised learning, because of the dependence on vast amounts of data and the possibility of introducing biases. Furthermore, companies have an obligation towards employees, investors, and the larger society to ensure that they deploy AI for good and not harm.
To succeed in the current constantly changing and advancing environment, organizations must be transformed from traditional to nimble, which means that they create systems where innovative ideas go through a constant process of review and refinement. Nimble leaders must focus on sensemaking, a set of skills that helps them understand what is happening around them and reduce complexity. Architecting a plan for transformation helps organizations develop adaptive capacities and avoid the trap of self-complacency. John Kotter identified an “8-step process for leading change:
- Create a sense of urgency.
- Build a guiding coalition.
- Form a strategic vision and initiatives.
- Enlist a volunteer army.
- Enable action by removing barriers.
- Generate short-term wins.
- Sustain acceleration.
- Institute change” (Kotter, 2012).
This process should not be interpreted as linear. Effective leaders establish an iterative process of sensemaking, relating, visioning and inventing so that at every step they are guided by the maps that emerge from observations, experiences, and conversations.
Transforming from traditional to nimble is not without its own set of challenges. Leaders must handle employees’ resistance to change through education and communication, participation, and involvement. Leaders can also foster a nimble, innovative culture by building an innovation infrastructure that allows everyone to develop their entrepreneurial spirit. This culture of innovation will rely heavily on leadership support for risk-taking, open innovation, and organizational learning.