Technology Strategy: Big Data and AI
Term: XXXX
Professor: Michael Impink Meeting time: XXXX
Email: XXXX Format: In-person
Office: XXXX Classroom: XXXX
COURSE DESCRIPTION
Cloud computing has empowered firms to collect, store, and compute large amounts of data, enabling the
mainstream development and use of artificial intelligence (AI1) in organizations. The development of AI products
and the use of big data to train the algorithms underlying these products has been advancing rapidly in recent
years, measured in terms of the resources devoted to it and its outputs.2 These increases will likely continue for
decades, driven by large investments from private firms and governments.
But what does all this mean for managers of firms?
This course will provide you with the strategy and management frameworks needed to manage data resources,
the programmers using data to develop AI, and the deployment of AI in your organization. We will examine how
to implement internally developed or licensed AI products and how to use data resources, algorithms, and AI as
competitive tools in the marketplace. The cases and materials will provide you with an understanding of how
cloud technologies sparked the mass development of AI, enabling nascent startups to compete against tech giants.
You will also learn about the potential risk of using data in business activities, including the importance of
managing government regulators and other stakeholders.
Familiarity with any new technology is important for managers so that competitive strategies can be aligned with
new opportunities and can guard against potential risks from the technology. But new technologies don't appear
in a vacuum. They are commercialized against a backdrop of macroeconomic trends and cultural norms, and their
adoption depends on many firm and industry-specific factors. We will start the course at the industry-level
(Section 1) and then zoom in on the manager’s perspective at the firm-level (Section 2).
COURSE REQUIREMENTS
Readings
There will be a relatively heavy reading load for the class, including cases, academic research papers, popular
press, and others. All cases and materials will be available to you through XXXXXX. You are expected to prepare
for each class by completing the assigned reading and thinking carefully through any preparation questions (in
addition to completing any assignments). The best managers are those that can read and assimilate lots of material,
so think of this as good training for your future career.
Grading
Apart from teaching you the ability to think sharply, graduate education helps your professional development by
requiring you to speak and write in a logical and compelling manner and broadeningyour professional networks.
I have created assignments and a grading system that reinforces these benefits of graduate education. However,
you are responsible for a large portion of your own development. You are required not only to attend classes but
also to carefully prepare for each class and contribute actively to classroom discussions. Requests for make-up
assignments or classes will not be entertained.
1
AI loosely describes a range of advanced technologies that incorporate a human-like learning ability, including machine
learning (ML), autonomous machines, computer vision, language processing, and neural networks.
2
For instance, The Economist estimated that AI-related mergers and acquisitions were 26 times larger in 2017 than two
years earlier. [Link]
work-could-become-faireror-more
1. Class attendance and participation 30%
2. Individual case write-ups (20% X 2) 40%
3. Group case write up and presentation 30%
Class attendance and participation (20%)
Attendance is required and part of the participation grade. Late arrival or early exit will hamper your ability to
learn the material, impact your grade through missed participation opportunities, and disrupt the flow of the class.
In the case of unavoidable absences,please notify me ahead of time.
You may use laptops in class to access slides and take notes, but not for anything unrelated to class. It is evident
when you are not paying attention to your classmates and me. Also, it can be distracting for others sitting behind
or beside you.
You should be prepared to share your ideas and critically interpret the perspectives presented by others. Please
carefully review the readings and case preparation questions for every class session. Quantity of participation, if
lacking in quality, will not be rewarded. The chief criterion for this part of the grade is the student's contribution
to the learning of others.
I reserve the right to cold call, not to intimidate or create a hostile environment, but to move along our discussion.
If, for some reason, you are unprepared for any section, let me know before class, and I will not cold call you
during that session. I have listed questions under each session to guide your thinking while reading while
preparing for each class. Please come prepared to answer these questions in class.
Accordingly, quality participation has one or more of the following characteristics:
• Courtesy: Respecting and observing the rules of the game and the rights and dignity of your classmates;
limiting participation to a fair share.
• Coherence: Providing careful analysis and turning vague ideas into precise statements; integrating
experiences and learning material with classroom discussion.
• Cleverness: Quick and deep thinking; ability to abstract and infer; contributing ideas that you would not
have come across; providing new ways of looking at the world.
• Courage: Willingness to take risks by asking good questions that potentially espouse unpopular ideas or
challenging authority.
• Character: Special personal attributes, such as a sense of humor or a supportive personality, are a service
to your classmates.
Individual case write-ups for “WhatsApp” and “CloudWars” (40%)
Two individual short assignments (each count for 20%) must be completed and uploaded to the course website
before class on their due date. Write-ups should be no longer than TWO typed pages (single-spaced, 12-point
font, one-inch margins). Figures and Tables can be used to supplement the text and do not count toward the page
limit but are seldom needed. To answer these questions, you should pull from the course readings, lectures, and
prior materials to answer the questions. You will not need to integrate outside resources (i.e., Wikipedia,
ChatGPT, or other unassigned readings ). However, you must cite any external materials used to attribute them
(including ChatGPT) correctly.
Answer all the questions listed for the class when the case will be discussed
Group case write-up and presentation for “GROW” (30%)
On the last day of class, you will present around five slides (ten minutes) to provide a business recommendation
for the GROW case. This presentation will be based on your response to the question (2 pages, following the
same guidelines for the other cases), which will be due before the final class. Please clearly list the names of your
group members on the first page of the assignment. The write-up will count for 20% of your grade, and the
presentation will count for 10% of your course grade (totaling 30%). You do not need to turn in your slides, just
the write-up.
2
CLASS OVERVIEW
Date
XXX Class 1 Overview of AI: Current macro-economic and cultural trends
Section 1 Industry-level
XXX Class 2 Part 1: Data and algorithms as competitive resources
XXX Part 2: The rise of superstar firms, cloud computing and digital
platforms
Class 3 Competing as the high-tech startup and as the big tech firm
Section 2 Firm-level
Class 4 Part 1: Managing AI product development (algorithmic bias)
Part 2: Managerial decision-making from AI (human bias)
Class 5 Managing customers’ data (privacy)
Class 6 Part 1: Managing potential AI regulation
Part 2: Group Presentations of GROW case recommendation
Class 1
Overview of AI: Current macro-economic and cultural trends
This class will lay the groundwork for understanding what people mean when they say “machine learning” or
“AI.” These rapid advancements apply not just to AI, but also to robotics, sensors, and the connection of them all
via digitization (also known as “Industry 4.0”). These advancements have started to manifest themselves in
various applications, including AI beating humans at complex strategy games,3 the creation of chatbots and virtual
assistants such as Alexa and Siri,4 and Amazon’s new cashier-less and cash-less grocery stores.5 We will then
discuss AI as a driving force in future macroeconomic progress.
Reading: AI and the economy, by Furman and Seamans (2019).
[Link]
Questions:
• How might the declining labor force participation rate in a highly developed country affect the adoption
of AI?
• Which firms or subgroups of the population would benefit?
• Which would be negatively affected?
Section 1: Industry-level
Class 2
Part 1: Data and algorithms as competitive resources
Data fuel the algorithms that support AI and are important in digital industries. Prior strategic frameworks, such
as the Resources-based view, have discussed strategic resources that lead to competitive advantage as being
scarce and difficult to imitate or replicate. In the class, we compare and contrast how information and data, as a
specific type of information, with more “traditional” resources. Then we explore how certain aspects of data make
it more or less valuable in certain industries and verticals.
3
[Link]
4
[Link]
5
[Link]
3
Part 2: The rise of superstar firms, cloud computing and digital platforms
This part provides a macro-level overview of how competition in the economy has changed amid digitization.
We will assess the emergence of superstar firms like Amazon and Google and explore what accounts for their
rise to dominance. Specifically, we will discuss the role of network effects in creating winner-take-all dynamics.
We will close with an overview of policies that have emerged to account for these new competitive dynamics.
Readings:
• Case: Cloud Wars Go Global: How Amazon, Microsoft, Google, and Alibaba Compete in Web
Services. [Link]
• Network effects and market power: What have we learned in the last decade, by Tucker (2018).
[Link]
Videos:
• The rise of superstar firms, by VanReenan (2019). [Link]
• The rise of superstar firms and the fall of the labor share, by Autor (2020).
[Link]
Cloud wars case questions (answer all parts):
• What are the complementarities between these large technology firms' cloud and non-cloud businesses?
• Which aspects of their prior business model were most valuable in gaining competency in cloud services?
• Would you invest in a cloud services startup competing against Big Tech?
Class 3
Competing as the high-tech startup and as the big tech firm
This class focuses on platform business models that have emerged in the digital economy and play an important
role in how many data-centric firms compete. We will explore different types of platforms and their corresponding
ecosystems and discuss their connection with the network effect, how they scale, and how they are governed. We
then focus on a specific type of platform that has enabled the use of data and AI across industries: cloud platforms.
Digital technologies, ranging from using open source to hiring freelance workers to crowdfunding, and access to
pay-as-you-go cloud services reduce costs for entrepreneurs to enter the market. But they also increase pressure
on them to create value in new ways to compete with one another and incumbents. We will explore strategies
entrants may pursue—from multi-homing to finding a niche to integration—in order to succeed.
We will shift to applying digital competition fundamentals to the incumbent’s strategy. Specifically, we will
assess how established businesses—tech and traditional ones—can maintain a competitive advantage in the
digital economy by addressing multi-homing challenges, creating barriers to entry, and acquiring entrants. We
will conclude this session with a case discussion on Facebook’s acquisition of WhatsApp, where students will be
assigned roles as either Facebook (the acquirer) or WhatsApp (the acquiree).
Readings:
• Case: $19B 4 txt app WhatsApp...omg!
• Disruptive innovation (HBR), by Christensen, Raynor, and McDonald (2015).
[Link]
[Link].
• Strategy for start-ups, by Gans, Scott, and Stern (2018). [Link]
Videos:
• How the Platform Ecosystem Works, by Schilling (2019).
[Link]
• Video: The Explainer: How to be a disruptor, by HBR (2019).
[Link]
• Driving digital strategy, by Gupta (2018). [Link]
WhatsApp case assignment questions (answer all parts):
• If you are Facebook, should you acquire WhatsApp? Why or why not?
• If you are WhatsApp, should you agree to Facebook’s acquisition offer? Why or why not?
4
Section 2: Firm-level
Class 4
Part 1: Managing AI product development (algorithmic bias)
What makes an algorithm biased? Is the data used to train the algorithm or the programmers coding the algorithm?
As managers, how can we create an environment more conducive to developing fairer products ?We will discuss
some of the ethical issues surrounding the use of data in product development. By the end of this class, you will
understand the tradeoffs between more ethical production
Part 2: Managerial decision-making from AI (human bias)
Humans choose which AI to adopt and how to implement that AI product in their firm to accomplish their
objectives. However, humans are biased, and norms in their industry and competitive environment may impact
their adoption and use of certain technologies or types of AI. We discuss how managers, through increased
awareness, can overcome some of these biases to make fairer decisions.
Readings:
• What managers need to know about big data, by Provost and Euchner (2017).
[Link]
• Biased Programmers? Or Biased Data?, by Cowgill, Dell’Aqua et al. (2020)
[Link]
• How managers can enable ai talent in organizations, by Rock (2020).
[Link]
• The two friends who changed how we think about how we think, The New Yorker.
[Link]
how-we-think
Questions:
• What enables and constrains AI adoption at the firm level?
• How do differences in management practices affect the decision to adopt AI and its use within the firm?
• What enables and constrains AI adoption at the firm level?
• How do differences in management practices affect the decision to adopt AI and its use within the firm?
Class 5
Managing employees' and customers’ data (privacy)
The use of AI, specifically neural networks, requires enormous data sets which may include personally
identifiable information (PII). The use of this data may negatively impact privacy and may lead to biased (less
fair) outcomes for certain groups. There are numerous tradeoffs between the technological advancement of AI
and aspects of consumer welfare. For instance, increased usage of personal data may enable algorithms to run
faster or be more accurate; however, at the same time, they may harm consumers that would like their personal
information to remain unused and more protected.
Readings:
• Case: GROW
• Artificial intelligence in human resources management: Challenges and a path forward, by Tambe,
Cappelli, and Yakubovich (2019) [Link]
GROW case assignment question (answer all parts): You are the chief human resources leader for a technology
firm in California, and you have been asked by the Board if your firm should implement GROW technologies in
hiring. What is your recommendation? Analyze your decision in terms of:
1) algorithmic bias
2) impact on firm culture,
3) employee data security and privacy.
[In class 6 you will present you recommendation to your classmates. Please provide the considerations that went
into your decision and support your decision with knowledge from the course lectures, readings, and cases.]
5
Class 6
Part 1: Managing potential AI regulation
It is hard to imagine a business not using the internet. Similarly, in a few years, it will be hard to imagine a
business that does not use the cloud in some way. Almost every firm is entangled with a large technology firm
that provides it with some IT services. For startups, these ties may enable access to hard-to-develop resources yet
potentially impact the nature of their innovations. We will pull from prior research on the telecommunications
industry to introduce the concept of “essential facility” to discuss the pros and cons of these relationships.
Next, we discuss how the external environment, particularly government regulators, impacts the usage of AI at
the firm level. AI is relatively new, and, as such, there is limited regulation. Certain governments (mainly the EU
and California) have created policies to address privacy-related issues. However, liability and fairness issues
remain unlegislated. Many of the most active firms lobbying regulators are the same firms with lots to lose (or
gain) from future regulation.
Part 2: Group presentation of GROW case business recommendations
Come prepared with your ~5 slides 9. Not everyone in the group must present, but everyone could present.
Readings:
• Digital data, platforms and the usual [antitrust] suspects: Network effects, switching costs, essential
facility, by Tucker (2019). [Link]
• Making the digital economy “fit for Europe”, by Renda
[Link]
• U.S. Department of Transportation (DOT), creating a data repository
• Read pages 1-11 of the draft proposed EU AI Act. [Link]
Questions:
• Should the government collect and provide data on roads and crosswalks (i.e., data repository) to all
companies? Why? Would it impact competition?
• Could cloud services be an essential facility in AI development? How do the US and EU differ in their
approach to regulating big tech?
