The Place of Management in an AI Curriculum
Academic degrees in Artificial Intelligence focus too often on Data Science per se than its transformative impact in organizations and markets.
Most academic programs specializing in Artificial Intelligence today focus on understanding data preparation, statistics, modeling, and algorithms while leaving AI’s impact on business to a footnote. This technocentric vision of Data Science may help explain both why businesses will invest $383 billion in AI this year, and why the largest majority of these AI projects will produce no real-world applications.
As artificial intelligence gains a foothold in all aspects of business, what do students need to learn about Management and AI in engineering and business schools? In this three-part series, we will examine and illustrate challenges ranging from aligning AI with corporate strategy, to developing co-intelligence between human and machine agents, to AI’s paradoxical role in innovation. In this first contribution, let’s focus on the need to understand how AI can influence strategy, the concept of “AI readiness” and AI project management.
AI and Corporate Strategy
Although several authors have explored the notion of digital strategy[i], a more important question is how artificial intelligence influences the way we think about business. We can begin with an organization’s business model, i.e. how AI will modify how a company segments its market, develops processes and networks to address customer needs, and measures the results. The application of AI to business process improvement also needs to be addressed, focusing on how machine learning can improve an organization’s understanding of customer behavior, the cost of producing products or services, and the pertinence of the metrics used to evaluate the processes themselves.
The data needed to improve AI models is often produced by the organization’s suppliers, business partners, and customers requiring management to rethink the boundaries between the organization and its ecosystem. Because the data needed to train AI models is largely unstructured, management needs to take a new look at how data is captured, interpreted and shared. Finally, AI transforms the nature of management itself: Managerial talent will increasingly be defined around judgment related skills: mentoring, motivating and defining acceptable data practices.[ii] Finding the right balance between human and machine intelligence has never been more important to corporate success.
AI’s impact on banking and finance provides several cases in point. The progressive introduction of AI has diversified the underlining business models of both retail and investment banking from full-service organizations to infrastructure providers, service aggregators, and platform players.[iii] Machine learning can be leveraged in a variety of ways in improving core banking processes like fraud management depending on whether the organizations are trying to improve the process, to understand consumer behavior, or to develop new service offerings.[iv] The needs of Data Science are sorely testing the limits of the industry core IT systems, pushing management to seek more open and more agile data stores with Fintech providers. Finally, digital transformation is redefining managerial skill sets, for transforming data into action requires managers at ease both behind their screens and in front of their customers.
AI Readiness
Artificial Intelligence isn’t a plug-and-play technology, it must be applied, fine-tuned and evaluated in the context of distinct industrial logics and corporate cultures. AI Readiness refers to the extent to which organizations are prepared to leverage the potential benefits of artificial intelligence. Managers need to buy into a problem-solving mentality of constantly exploring the context of their business to understand the roots of their business challenge, qualifying the data at hand, choosing the algorithms that address the challenges, and transforming the data into action. Corporate culture needs to reward risk-taking rather than risk avoidance. Corporate policies need to integrate incentives that encourage employees to develop new products, services, and markets rather than protecting existing market share. Stakeholders must take a hard look at the “production boundaries” that separate “productive” activities from “unproductive” tasks to account for a large number of AI experiments that will not directly contribute to the bottom line. Finally, the metrics used to qualify performance will need to be enlarged to reflect not just the stockholder’s view of efficiency, but the potential effectiveness, innovation, and even ROI (public value) for the ecosystem as a whole.
Returning to our examples in banking and finance, AI readiness represents a managerial project that extends far beyond the mechanics of machine learning itself. How can individual managers be trained to look beyond bank processes to envision viable solutions to the challenges of state regulation, competition from new market entrants, and evolving customer demands? How can the evaluation metrics be refined to encourage retail banking to find niche areas of expertise and develop their unique brands around “Banking as a Service” (BaaS)? How will “productive activities” be defined and evaluated when algorithms replace human decision-makers in consumer analytics, process automation or risk management? The result of AI Readiness isn’t a binary “go/no-go” decision concerning a pilot program, but a series of granular insights that provide actionable recommendations around AI to foster organizational performance.[v]
AI Project Management
The gap between current investments in AI and actual returns underlines the fundamental importance of project management in any AI curriculum. Industry analysis of current AI projects demonstrates that upper management often lacks a sufficient understanding of what AI/ML entails, business users are not clear about to expect from AI experiments, and AI projects are characterized by time and cost overruns and the absence of vendor support.[vi] Managing AI pilots as projects can help align Data Science experiments with corporate strategy, produce operational models that are both directly applicable to the business, and demonstrate how their investments impact the bottom line.
Considerable insight can be gained in studying the differences between an AI experiment and an AI project through the optics of project initiation, planning, execution, evaluation, and closure. The challenges of evaluating Readiness, Risk, and Rewards[vii], developing collaborative intelligence[viii], and amplifying the network effects between cross-functional Data Science teams, the organization and its ecosystem are all fundamental levers that will contribute to successful project management. Artificial Intelligence doesn’t operate in a vacuum; the interplay between human, organizational, and technical resources transforms data and algorithms into products and services.
In banking and finance, the evolution of Open Banking provides ample food for thought here. Standards like CDR, CMA, PSD2, and promise to enhance competition between financial institutions, provide consumers with more personalized products, and increase service innovation. Open Banking frameworks impose legal guidelines concerning third-party access, informed consumer consent, data security and dispute resolution of personal financial data. These guidelines offer little help to management in understanding how to prepare their organizations for more open and fluid markets, how to adequately assess the risks of modifying current processes and practices, and where and how AI can play a role. Putting together an AI team, transforming AI experiments into projects that can lead to competitive products and services, and evaluating the results are managerial challenges that will determine AI’s contribution to the business. In the world of banking and finance, project management is a fundamental component of any AI training.
What have we learned?
Academic degrees in Artificial Intelligence today need to help students look beyond data and algorithms and explore how Data Science transforms organizations and markets. The place of management in Artificial Intelligence merits more than a footnote, for management is fundamental in aligning the promise of AI with each organization’s capabilities, strategy, capabilities, resources, and culture. In this contribution, we have addressed three themes, AI and Corporate Strategy, AI Readiness, and AI Project Management that can enhance any Data Science program. In our next contribution in the series, we will look at the importance of fostering Collaborative Intelligence, improving Managerial Decision-Making, and the implications of Digital Transformation.
Lee Schlenker is the Principal in the BAI http://baieurope.com and Professor of Business Analytics and Community Management. His LinkedIn profile can be viewed at www.linkedin.com/in/leeschlenker. You can follow the BAI on Twitter at https://twitter.com/DSign4Analytics
[i] See, for example, Gupta, S. (2018), Understanding Digital Strategy, HBR [ii] Fountaine, T. et al. (2019), Building the AI-Powered Organization, HBR [iii] Robinson, B. (2016), 4 banking business models for the digital age, [iv] Schlenker, L. (2020), Transforming Data into Action, Towards Data Science [v] Grasser, M. (2019), Six Steps to an Intelligent AI Strategy, Medium [vi] Heath, N. (2019), Managing AI and ML in the enterprise 2019, TechRepublic [vii] Shacklett, M. (2019), How to manage an AI project’s rewards, risks, and readiness, TechRepubic [viii] Wilson, H.J. (2018), Collaborative Intelligence: Humans and AI Are Joining Forces, HBR
