Artificial Intelligence - Implications for Business Strategy 2017-11-20. MIT- Sloan & CSAIL

Tel: +1 224 249 3522 | Email: mitsloan@getsmarter.com | Website: getsmarter.com
Page 2 of 14
executive.mit.edu
Learning outcomes:
LO3: Propose an initiative for a specific business application of AI.
LO4: Produce a roadmap for an organization to gain strategic advantage using AI.
Name: Winston Dodson M6 U2 Assignment
Instructions and guidelines (Read carefully)
Instructions
1. Insert your name and surname in the space provided above, as well as in the file
name. Save the file as: First name Surname M6 U2 Assignment – e.g., Zadie
Smith M6 U2 Assignment. NB: Please ensure that you use the name that appears
in your participant profile on the Online Campus.
2. Write all your answers in this document. There is an instruction that says, “Start
writing here” under each question. Please type your answer there.
3. Submit your assignment in Microsoft Word only. No other file types will be
accepted.
4. Do not delete the plagiarism declaration or the assignment instructions and
guidelines. They must remain in your assignment when you submit.
PLEASE NOTE: Plagiarism cases will be investigated in line with the terms and
conditions for participants.
IMPORTANT NOTICE: Please ensure that you have checked the Online Campus
for the due date for this assignment.
Guidelines
1. There are 4 pages and 1 question in this assignment.
2. Make sure that you have carefully read and fully understood the questions before
answering them. Answer the questions fully but concisely and as directly as
possible. Follow all specific instructions for individual questions (e.g., “list,” “in
point form”).

Page 3 of 14
executive.mit.edu
3. Answer all questions in your own words. Do not copy any text from the
casebook, readings or other sources. The assignment must be your own work
only.
Plagiarism declaration:
1. I know that plagiarism is wrong. Plagiarism is to use another’s work and
pretend that it is one’s own.
2. This assignment is my own work.
3. I have not allowed, and will not allow, anyone to copy my work with the
intention of passing it off as his or her own work.
4. I acknowledge that copying someone else’s assignment (or part of it) is
wrong, and declare that my assignments are my own work.
For each assignment submission included in Modules 1 to 5 of this program, you were
asked to complete activities that should now help you complete a final, polished AI
roadmap. The roadmap will function as a starting point for the successful deployment
of AI into an organization of your choice, with the aim of achieving competitive
advantage.
For this final assignment, use your answers from previous modules as a guide to
populate the roadmap template below. You should be using the organization that
formed the basis of your ongoing project submissions from Modules 1 to 5 to fill in
your roadmap. Please ensure that you consult the assignment guide in Unit 2 of this
module for details on the specific content required for each section. Furthermore, you
can use the feedback you have received in previous modules to refine your answers
here.
Please adhere to the word limit for each section.
Review guidelines:
Your assignment will be reviewed according to each individual section of the
roadmap, and the structure and logic of your writing. View the detailed rubric on the
Online Campus.
Question 1
Complete the four sections of the following roadmap template.

Page 4 of 14
executive.mit.edu
Executive summary
Summarize your plan for the strategic implementation of AI into your chosen
organization. (Max. 500 words.)
Start writing here:
Mixing two clichés - “Change happens!”. Our challenge in business, as in life, is to not
only to meet these challenges but to take advantage of them while benefitting
ourselves, our associates and our societies. The fast-moving development of Artificial
Intelligence (AI) is a significant change and we may be living through an inflection point
caused by these developments. In order to understand how we meet these many
challenges must first understand AI.
A definition of AI is “intelligence displayed by machines, in contrast with the natural
intelligence (NI) displayed by humans and other animals.”
(https://en.wikipedia.org/wiki/Artificial_intelligence ). An organization that adopts and
uses AI would have AI accomplish the tasks that it can do better while utilizing people
to continue to accomplish critical tasks fit for them. AI is generally broken down into
three subfields:
Machine Learning (ML), Natural Language Processing and Robotics. Machine
Learning is “a field of computer science that gives computers the ability
to learn without being explicitly programmed”(
https://en.wikipedia.org/wiki/Machine_learning )Using algorithms and the power
of modern computational abilities these techniques can use the statics of
common data to make decisions.
Natural Language Processing “is a field of computer science, artificial
intelligence …... concerned with programming computers to fruitfully process
large natural language data.” (https://en.wikipedia.org/wiki/Natural-
language_processing ). This enables people to powerfully access large
sources of data, in almost the same way that those people speak with each
other, to research large sources of data and find answers that they could not
get before.
Robotics are “technologies …… used to develop machines that can substitute
for humans and replicate human actions.”
(https://en.wikipedia.org/wiki/Robotics ). The proper use of robotics can free up
human productivity to achieve tasks more readily achievable only by humans.
Another concept that we will need in our transition will be the O-ring principle. The O-
ring principle is a concept developed for economics used to describe how, in an
economic system, the total value created is related to the value of each
subcomponent. The name “O-ring” come from the 1986 Challenger disaster where the
disaster was found to be caused by the failure of a simple, yet important, simple
component. Thus, as sub-components, humans are O-rings, important part of an
organization but our role is defined by what value we bring.

Page 5 of 14
executive.mit.edu
The O-ring principle is used to define human’s role within an organization where many
of the tasks that were previously performed by humans, and our natural intelligence, in
a new organization where tasks are now or are planned to be, performed by AI. In
effect, what value do we bring to the organization when other tasks are performed by
AI?
The following is a presentation using a case study, and applying the concepts above,
the current state of that organization will be described. Then the organizations future
state after an AI initiative. Finally, we will then propose an Action Plan needed to
implement that AI initiative in the case study. In each case we cover the entire
organization from three perspectives; people process and technology.

Page 6 of 14
executive.mit.edu
Current state
Describe the current state of your chosen organization. (Max. 500 words.)
Start writing here:
The organization is one of the major production sites for a large pharmaceutical
company with yearly sales of ~ $1B. The treatment is one of the pioneering “silver
bullets” of the biologics revolution. Despite numerous attempts by competitors it
remains one of the only effective treatment for a deadly inherited disease. A few
years before I joined the product was found to be contaminated. A series of FDA
inspections found numerous systemic problems from production, personnel training
and processes to facilities. A major crisis ensued. Any disruption in the supply of
the treatments would have dire consequences for patients. Production was allowed
to continue while a plan was developed. Doing this is like changing the tires on a
moving automobile.
The plan that was eventually agreed upon was enforced by a consent decree with
the federal government and it would allow continued production of the treatment
under the close onsite supervision of an external 3rd
party. The entire plan would
require extensive new processes, employee training and production site remediation
and would cost the pharmaceutical corporation hundreds of millions of dollars. This
was all in accordance with a Consent Decree signed between the company and a US
Federal Judge and enforced by threat of legal actions.
The external 3rd
party onsite supervision would generate many times the number of
process and product deviations associated with a site this size. A less formal system
of findings generated by the 3rd
party would also generate a significant number of
issues that had to be addressed. Because of the nature of the situation many of these
deviations and findings were also very complex. All of this activity departments
organic to the site that were 3-4 times the size of industry standard benchmarks.
All of these deviations and findings were captured and addressed in industry
standard, benchmarked IT systems and the process and personnel used were
similarly industry standard benchmarks as required by the FDA.
A summary of the current state is:
• Operations required OVER-Inspection
• Progress required the demonstration of Process Capability
This expensive and time consuming brute force process is working. The treatments
have always been available to patients and the site has shown that it is capable of
compliant production with much less onsite, eternal 3rd
party supervision.

Page 7 of 14
executive.mit.edu
The Pharmaceutical / Biologics / Medical Device is one of the most regulated
markets in the world. The FDA has an extensive set of regulations based on Current
Good Manufacturing Practices. So even compliantly conducting business in such an
environment is expensive. The effort and thus costs are proportional to:
• the longevity of the market (how long have all similar treatments been
available) – new treatments require more monitoring
• the number of competitors – fewer competitors requires more monitoring
• complexity / novelty – both require more monitoring
• complexity of production process – more complex requires more
monitoring.
Compliant operations require expensive compliance processes depending on the
organization’s position in the market. The costs of compliance are table stakes the
organization has to play to be in the game.

Page 8 of 14
executive.mit.edu
Proposed initiative
Outline your proposed initiative for using AI in your chosen organization to achieve
competitive advantage. (Max. 800 words.)
Start writing here:
Biologics production must focus on compliance. The result is that this imperative
reduces those organizations ability to compete on Porter’s other generic strategies,
cost leadership or differentiation. So, complying with regulations in the industry are
like “table stakes” for all participants so this Focus must be accomplished as
efficiently as possible. Competitors surely won’t stop trying to do so, so adopting
AI will not only become a virtue, temporarily but like an arms race, become a
requirement to survive the threats from others in the industry. An intelligent
organization built on machine learning would turn a “vice into a virtue” and move
from focusing on compliance to Porter’s cost leadership or differentiation.
This intelligent organization will operate with a set of interlocking ML
methodologies, coordinated by a neural net and linked together by a Blockchain –
the sum will be greater than the parts. A Blockchain enables remote auditing and
the accumulation of knowledge.
The Blockchain used to document these activities will be accessible by the
appropriate regulatory agencies, enabling remote auditing by the regulatory
agencies’ machine learning algorithms eliminating most on-site inspections.
An organization transformed using well established program, project and change
management techniques and integrating AI will be an intelligent organization
capable of practicing collective intelligence. The roles of the people will require
Social Intelligence along with the new skills needed to leverage their new AI tools.
These AI tools will create an experience database that will enable the entire
organization to build tools to leverage past experience. This organization will be
more productive, profitable and prepared to respond to market challenges by
shifting among, as appropriate, Porters three generic corporate strategies.
The paragraphs below describes the projects in the Strategy Deployment Matrix.
Biologicals production - processes requiring human actions are frequently required
by FDA to be Validated (operator) or Witnessed (another operator). Current state,
signatures on documents or via electronic signatures in IT systems.
Quality Assurance – a highly regulated environment i.e. FDA, requires frequent
review of controlled documentation. Natural language interfaces, both text and

Page 9 of 14
executive.mit.edu
speech, to query these documents would both speed up as well as ensure
comprehensive access to that information.
Production Scheduling – currently accomplished utilizing Enterprise Resources
Planning (ERP) technologies. Neural Nets (Google’s Alpha Go, IBM’s Watson),
would be used instead. Using current standards, these Neural Nets would first try
every scheduling combination possible, then combine these with past experiences
with scheduling decisions. Finally, optimized with real-time reinforced learning.
Training / Human Resources – Use pattern recognition to detect trends in deviations
and delays in production. Combine this analysis with Validation/Witness data from
production and documentation access data from Quality Assurance to determine
root causes of deviations and production delays due to gaps in the training of
current staff.
Input Supply Logistics – biologics production is critically dependent on supply.
Currently supplies are stored locally, many in expensive cold-chain warehouses.
Self-driving / robot transportation an economical form of just-in-time delivery of
time critical components. This will eliminate most of the costs and risks to onsite
storage of critical supplies.
Biologics Production– occurs in aseptic conditions (clean rooms) that require
elaborate gowning and cleaning procedures for operators entering the production
areas. All supplies also require humans to transport them into the production areas.
Supply Robots like Kiva’s and Sadiki’s could perform many of the supply tasks
now accomplished by humans.
Quality Control – like production operators, QC technicians need frequent access to
the aseptic production floor to take samples and transport them from the floor to the
QC lab. The transport robots described above could take the samples taken by the
production operators directly to the QC technicians in their labs.
Facilities and engineering – like production operators and QC technicians facilities
and engineering need frequent access to the production floor. And due to space
limitations often can’t take all the parts needed. So, also as described above, a
transport robot could bring them the parts required on a just-in-time basis.
Input Supply Logistics – biologics production is critically dependent on supply.
Currently supplies are stored locally, many in expensive cold-chain warehouses.
Self-driving / robot transportation is an economical form of just-in-time delivery of
time critical components. This will eliminate most of the costs and risks to onsite
storage of critical supplies.
Biologics Production– occurs in aseptic conditions (clean rooms) that require
elaborate gowning and cleaning procedures for operators entering the production

Page 10 of 14
executive.mit.edu
areas. All supplies also require humans to transport them into the production areas.
Supply Robots like Kiva’s and Sadiki’s could perform many of the supply tasks
now accomplished by humans.
Output logistics – one facility’s output logistics is another’s inputs. Finished
product (medical treatments) many times must be transported for filling and
packaging etc. Special transports (cold-chain) are required so valuable finished
product is expensively stored onsite. Once again, self-driving robot transportation
could provide frequent outbound transport eliminating local cold-chain storage

Page 11 of 14
executive.mit.edu
Plan of action and criteria for success
Detail your plan of action for using AI in your chosen organization. (Max. 800 words.)
Start writing here:
Change Management is commonly divided into 3 areas; people, process and
technology. People are grouped into leadership and process participants. Process
involves the change process itself and then the various operational processes, they
are; production, quality assurance, production scheduling, training / human
resources, quality control and medical affairs. Technology is machine learning with
its many subsets as applied specifically to each of the operational processes listed
above.
The Strategy Deployment Matrix that the IC will use is included below. It includes:
• Generic Corp Strategies
• Business Level Metrics & Goals – with time
• Operational Level Metrics and Goal
• Departments / Functions / Processes
• Projects / Initiatives
• Roles and Responsibilities
• Artificial Intelligence Categories
• Implementation Schedule
• The interaction between each section above, when appropriate
The 2 groups of people are leadership implementation committee (IC) and then
process leaders (PL). The IC will implement the change and so will approve the
plan, manage the resources and mitigate risks. The PL’s will work with the IC and
will lead and execute the changes. They will also train and recruit process
participants (PP). Participants will learn then execute the post change process.
The change process ends at successful implementation. PL’s lead changes to each
of their respective processes and will design their individual process while working
with other PL’s and the IC to ensure that all the processes support each other and
their requirements. Also, they develop the specifications needed for the technology

Page 12 of 14
executive.mit.edu
implementation. Each process may require a unique ML methodology to support
the overall goal.
These initiatives could be considered an augmentation of current business and IT
strategies by using ML to enable compliant operations. It will require additional
and new IT resources. Training for process participants new recruits would fill
training and experience gaps.
The CIO leads the technology implementation utilizing the IC, must understand the
various ML mythologies and must ensure that the skills needed by PL’s /
participants to work in the post implementation environment are in place. Through
the IC the CIO must also provide any surge resources needed during the
implementation as well as training PL’s to lead and manage in the new intelligent
organization.
Each process will be responsible for writing the specifications for their process to
include use cases, descriptions of data sources and to document how the various
Many of the technical requirements would be dictated by FDA requirements to
demonstrate process capability. An example would be self-driving robot transports
to be able to maintain cold-chain conditions. Another would be that robots
transporting product onto the production floor to be able to be automatically
cleaned to a clean room standard. The CIO, production and supply chain would
lead the technical changes required for just-in-time supply. The CIO the wireless
communication standards and the data structure required to support the various IT
production systems. The leaders of production and supply chain would develop the
timing standards governing what just-in-time means.
(The Strategy Deployment Matrix is on the last page below. Exporting all the text to MS
Word yields a count of 318).

Page 13 of 14
executive.mit.edu

Artificial Intelligence - Implications for Business Strategy 2017-11-20. MIT- Sloan & CSAIL

Artificial Intelligence - Implications for Business Strategy 2017-11-20. MIT- Sloan & CSAIL

More Related Content

Similar to Artificial Intelligence - Implications for Business Strategy 2017-11-20. MIT- Sloan & CSAIL

Recently uploaded

Artificial Intelligence - Implications for Business Strategy 2017-11-20. MIT- Sloan & CSAIL