In this talk for the students of IIM Udaipur, I have discussed how AI as technology needs to deliver business value in order for AI as a discipline to be seen as relevant to business. I have also spoken briefly about my own research work.
2. DISCLAIMER
• THESE ARE MY PERSONAL VIEWS!
• THE PRESENTATION FOCUSED ON AI
AS A BUSI
• THE PRESENTATION IS BASED ON MY
ONGOING RESEARCH WORK, AND AS
SUCH, LIABLE TO CHANGE
3. TOPICS
• Technology & Society
• Evolution of AI, ML, DL
• Business Value Impact
• Barriers and Challenges
• Future of AI…
4. WHAT IS TECHNOLOGY?
• the use of science in industry, engineering, etc., to invent useful things
or to solve problems
• a machine, piece of equipment, method, etc., that is created by
technology
• the practical application of knowledge especially in a particular area
• a capability given by the practical application of knowledge
• a manner of accomplishing a task especially
using technical processes, methods, or knowledge
• the specialized aspects of a particular field of endeavor
https://www.merriam-webster.com/dictionary/technology
5. HISTORY OF TECHNOLOGY
The development over time of systematic techniques for making and doing
things.
By early 20th century, the term embraced a growing range of means, processes,
and ideas in addition to tools and machines. By mid-century technology was
defined by such phrases as “the means or activity by which man seeks to
change or manipulate his environment.”
Even such broad definitions have been criticized by observers who point out
the increasing difficulty of distinguishing between scientific inquiry and
technological activity.
There is a progressive element in technology, as it is clear from the most
elementary survey that the acquisition of techniques is a cumulative matter, in
which each generation inherits a stock of techniques on which it can build if it
chooses and if social conditions permit.
https://www.britannica.com/technology/history-of-technology
6. DO ONLY HUMANS DO “TECHNOLOGY”?
Essentially, techniques are methods of creating new tools and products of tools, and the
capacity for constructing such artifacts is a determining characteristic of humanlike
species.
Other species make artifacts... but these attributes are the result of patterns of
instinctive behaviour and cannot be varied to suit rapidly changing circumstances.
Human beings, in contrast to other species, do not possess highly developed instinctive
reactions but do have the capacity to think systematically and creatively about
techniques. Humans can thus innovate and consciously modify the environment in a
way no other species has achieved.
By virtue of humanity’s nature as a toolmaker, humans have therefore been
technologists from the beginning, and the history of technology encompasses the whole
evolution of humankind.
In using rational faculties to devise techniques and modify the environment, humankind
has attacked problems other than those of survival and the production of wealth
with which the term technology is usually associated today.
https://www.britannica.com/technology/history-of-technology
7. TIMELINE OF TECHNOLOGY IN THE HISTORY
OF HUMAN EVOLUTION
• ~3.3m: First tools
• ~1m: Fire
• ~20,000: Agriculture, Pottery,
Woven fabrics, Wheel
• ~6000 BCE: Irrigation
• ~4000 BCE: Sailing
• ~1200 BCE: Iron
• ~~850 CE: Gunpowder
• 950: Windmill
• 1044: Compass
• 1250: Mechanical Clock
• 1455: Printing
• 1765 CE: Steam Engine
• 1804: Railways
• 1807: Steamboat
• 1826: Photography
• 1831: Reaper
• 1844: Telegraph
• 1876: Telephone, Internal-
Combustion Engine
• 1879: Electric Light
• 1885: Automobile
• 1901: Radio
• 1903: Airplane
• 1926: Rocketry
• 1927: Television
• 1937: Computer
• 1942: Nuclear Power
• 1947: Transistor
• 1957: Spaceflight
• 1974: Personal Computer
• 1974: Internet
• 2012: CRISPR
• 2017: Artificial Intelligence
https://www.britannica.com/story/history-of-technology-timeline
8. THREE STAGES OF NEW TECHNOLOGY
ADOPTION
• First, the new technology or innovation follows the
line of least resistance, i.e. it is applied in ways that
that do not threaten people – reducing the chance
that the technology will be abruptly rejected.
• Second, the technology is used to improve
previous technologies (this stage can last a long
time), and
• Third, new directions or uses are discovered that
grow out of the technology itself.
Megatrends – John Naisbitt, 1984
9. IS TECHNOLOGY “AUTONOMOUS”?
• The definition of technology as the systematic study of techniques for
making and doing things establishes technology as a social
phenomenon and thus as one that cannot possess
complete autonomy, unaffected by the society in which it exists.
• Of course it must be admitted that once a technological development,
such as the transition from sail to steam power in ships or the
introduction of electricity for domestic lighting, is firmly established, it
is difficult to stop it before the process is complete. The assembly of
resources and the arousal of expectations both create a certain
technological momentum that tends to prevent the process from
being arrested or deflected. Nevertheless, the decisions about
whether to go ahead with a project or to abandon it are undeniably
human, and it is a mistake to represent technology as a monster or
a juggernaut threatening human existence. In itself, technology is
neutral and passive.
https://www.britannica.com/technology/history-of-technology
10. TECHNOLOGY AND SOCIETY
The relationship between technology and society is complex. Any technological stimulus
can trigger a variety of social responses, depending on such unpredictable variables as
differences between human personalities; similarly, no specific social situation can be
relied upon to produce a determinable technological response.
Any “theory of invention,” therefore, must remain extremely tentative, and any notion of
a “philosophy” of the history of technology must allow for a wide range of possible
interpretations. A major lesson of the history of technology, indeed, is that it has no
precise predictive value. It is frequently possible to see in retrospect when one
particular artifact or process had reached obsolescence while another promised to be a
highly successful innovation, but at the time such historical hindsight is not available
and the course of events is indeterminable.
In short, the complexity of human society is never capable of resolution into a simple
identification of causes and effects driving historical development in one direction
rather than another, and any attempt to identify technology as an agent of such a
process is unacceptable.
https://www.britannica.com/technology/history-of-technology
11. SOCIAL INVOLVEMENT
• Three key points of social involvement:
• Social need
• Social resources
• Sympathetic social ethos
• In default of any of these factors it is unlikely
that a technological innovation will be widely
adopted or be successful.
https://www.britannica.com/technology/history-of-technology
12. SOCIAL NEED
The sense of social need must be strongly felt, or people will not
be prepared to devote resources to a technological innovation.
The thing needed may be a more efficient cutting tool, a more
powerful lifting device, a labour-saving machine, or a means of
using new fuels or a new source of energy. Or, because military
needs have always provided a stimulus to technological
innovation, it may take the form of a requirement for better
weapons. In modern societies, needs have been generated by
advertising. Whatever the source of social need, it is essential
that enough people be conscious of it to provide a market for
an artifact or commodity that can meet the need.
https://www.britannica.com/technology/history-of-technology
13. SOCIAL RESOURCES
Social resources are similarly an indispensable prerequisite to a
successful innovation. Many inventions have foundered because the
social resources vital for their realization—the capital, materials, and
skilled personnel—were not available. The notebooks of Leonardo da
Vinci are full of ideas for helicopters, submarines, and airplanes, but few
of these reached even the model stage because resources of one sort or
another were lacking. The resource of capital involves the existence of
surplus productivity and an organization capable of directing the
available wealth into channels in which the inventor can use it. The
resource of materials involves the availability of appropriate
metallurgical, ceramic, plastic, or textile substances that can perform
whatever functions a new invention requires of them. The resource of
skilled personnel implies the presence of technicians capable of
constructing new artifacts and devising novel processes. A society, in
short, has to be well primed with suitable resources in order to
sustain technological innovation.
https://www.britannica.com/technology/history-of-technology
14. SYMPATHETIC SOCIAL ETHOS
A sympathetic social ethos implies an environment receptive to
new ideas, one in which the dominant social groups are prepared
to consider innovation seriously. Such receptivity may be limited
to specific fields of innovation—for example, improvements in
weapons or in navigational techniques—or it may take the form of
a more generalized attitude of inquiry, as was the case among the
industrial middle classes in Britain during the 18th century, who
were willing to cultivate new ideas and inventors, the breeders of
such ideas. Whatever the psychological basis of inventive genius,
there can be no doubt that the existence of socially important
groups willing to encourage inventors and to use their ideas
has been a crucial factor in the history of technology.
https://www.britannica.com/technology/history-of-technology
15. HIGH TECH NEEDS HIGH TOUCH!
“High tech / high touch is a formula I use to
describe the way we have responded to
technology. What happens is that whenever
new technology is introduced into society,
there must be a counterbalancing human
response – that is, high touch – or the
technology is rejected. The more high tech,
the more high touch.”
Megatrends – John Naisbitt, 1984
16. HOW DOES TECHNOLOGY IMPACT: 5I MODEL
(AUTHOR’S RESEARCH /WIP)
Integrate
• Adapt self
and/or the
environment
and enable
disparate
systems and
humans to
work together
to achieve a
specific goal
Improve
• Improve
existing
resource
efficiency,
process
productivity
and raise
human values
such as health,
safety, wealth,
happiness, etc.
Insights
• Generate new
capabilities,
knowledge or
insights by
recombining
existing
knowledge or
from existing
data, etc.
Invent
• Create new
possibilities
hitherto
unknown to
humankind
Imagine
• Imagine future
scenarios that
are non-linear
projection into
the world that
doesn't yet
quite exist, in
time, space or
any other
dimension
17. INTEGRATION OF TECHNOLOGY WITH
HUMANS (AUTHOR’S RESEARCH /WIP)
Augment
• Adoption of technology can
help augment human faculties
or capabilities, however in a
passive way
• E.g. calculator helps speed up
basic math, while a car helps
extend the geographic reach, or
a reading glass, etc.
Partner
• In a partnership, humans
delegate part of their decision-
making or execution to a
technology
• E.g. a pacemaker can enhance
human life, or using GPS frees
from the stress of knowing the
route
Replace
• Finally, when the technology
reaches a maturity where it can
be trusted enough to not only
function reliably but also act in
the best interests of humans in
the most ethical manner, it
might replace humans partially
or completely.
• E.g. high-speed stock trading,
weather prediction, or
manufacturing
18. ROLE OF MANAGEMENT INNOVATION IN
ADOPTION OF TECHNOLOGY
There were technological innovations of great significance in many aspects of
industrial production during the 20th century. It is worth observing, in the first place,
that the basic matter of industrial organization became one of self-conscious innovation,
with organizations setting out to increase their productivity by improved techniques.
Methods of work study, first systematically examined in the United States at the end of
the 19th century, were widely applied in U.S. and European industrial organizations in
the first half of the 20th century, evolving rapidly into scientific management and the
modern studies of industrial administration, organization and method, and particular
managerial techniques. The object of these exercises was to make industry more
efficient and thus to increase productivity and profits, and there can be no doubt that
they were remarkably successful, if not quite as successful as some of their advocates
maintained. Without this superior industrial organization, it would not have been
possible to convert the comparatively small workshops of the 19th century into the
giant engineering establishments of the 20th, with their mass-production and
assembly-line techniques. The rationalization of production, so characteristic of
industry in the 20th century, may thus be legitimately regarded as the result of the
application of new techniques that form part of the history of technology since 1900.
https://www.britannica.com/technology/history-of-technology
19. TECHNOLOGY: GOOD SERVANT OR A BAD
MASTER?
• Things are in the saddle and ride mankind – Ralph
Waldo Emerson
• Human beings are the sex organs of the machine
world. – Marshall McLuhan, Understanding Media: The
Extensions of Man
• “Until technology has the ability to reproduce itself on
its own…at that point, we become dispensable.” –
Nicholas Carr, The Shallows
22. SO, WHY NOW?
After decades of start/stops, finally AI seems to be at the cusp of its (third) resurgence.
23. WHAT IS INTELLIGENCE?
Intelligence can be defined as the ability to solve complex problems or make
decisions with outcomes benefiting the actor, and has evolved in lifeforms to
adapt to diverse environments for their survival and reproduction. For animals,
problem-solving and decision-making are functions of their nervous systems,
including the brain, so intelligence is closely related to the nervous system.
Intelligence is hard to define, and can mean different things to different people.
Once we consider the origin and function of intelligence from an evolutionary
perspective, however, a few important principles emerge. For example,
different lifeforms can have very different types of intelligence because
they have different evolutionary roots and have adapted to different
environments. It is misleading and meaningless to try to order different animal
species on a linear intelligence scale, such as when trying to judge which dog
breed is the smartest, or whether cats are smarter than dogs. It is more
important to understand how a particular form of intelligence evolved for each
species and how this is reflected in their anatomy and physiology.
Q&A – What is Intelligence, https://www.hopkinsmedicine.org/news/articles/qa--what-is-intelligence
24. FROM BIOLOGICAL INTELLIGENCE TO AI
“In my view, however, true intelligence requires life, which can be
defined as a process of self-replication. Therefore, I believe that
superintelligence is either impossible or something in a very distant
future. True intelligence should promote—not interfere with—the
replication of the genes responsible for its creation, including necessary
hardware like the brain. Without this constraint, there is no objective
criteria for determining whether a particular solution is intelligent. It
may eventually be possible for humans to create artificial life that can
physically replicate by itself, and only then will we have created truly
artificial intelligence, but this is unlikely to happen anytime soon. Until
then, machines will always only be surrogates of human intelligence,
which unfortunately still leaves open the possibility of abuse by people
controlling the AI.”
Q&A – What is Intelligence, https://www.hopkinsmedicine.org/news/articles/qa--what-is-intelligence
25. WHAT IS HUMAN INTELLIGENCE?
• Human intelligence, mental quality that consists of the abilities to learn from experience,
adapt to new situations, understand and handle abstract concepts, and use knowledge to
manipulate one’s environment.
• More recently… psychologists have generally agreed that adaptation to the environment is
the key to understanding both what intelligence is and what it does. Such adaptation may
occur in a variety of settings. For the most part, adaptation involves making a change in
oneself in order to cope more effectively with the environment, but it can also mean
changing the environment or finding an entirely new one.
• Effective adaptation draws upon a number of cognitive processes, such
as perception, learning, memory, reasoning, and problem solving. The main emphasis in a
definition of intelligence, then, is that it is not a cognitive or mental process per se but
rather a selective combination of these processes that is purposively directed toward
effective adaptation.
• Intelligence, in total, has come to be regarded not as a single ability but as an effective
drawing together of many abilities. This has not always been obvious to investigators of the
subject, however; indeed, much of the history of the field revolves around arguments
regarding the nature and abilities that constitute intelligence.
Human Intelligence, https://www.britannica.com/science/human-intelligence-psychology
26. WHAT IS AI?
Prof John McCarthy, Father of Artificial Intelligence:
• Intelligence is the computational part of the ability to
achieve goals in the world. Varying kinds and degrees of
intelligence occur in people, many animals and some
machines.
• Artificial Intelligence is the science and engineering of
making intelligent machines, especially intelligent
computer programs. It is related to the similar task of using
computers to understand human intelligence, but AI does
not have to confine itself to methods that are biologically
observable.
What is AI?: http://jmc.stanford.edu/artificial-intelligence/what-is-ai/index.html
27. BIRTH OF AI
• The Logic Theorist by Allen Newell, Cliff Shaw, and Herbert Simon’s was a
program designed to mimic the problem solving skills of a human and was
funded by Research and Development (RAND) Corporation. It’s considered by
many to be the first artificial intelligence program and was presented at
the Dartmouth Summer Research Project on Artificial
Intelligence (DSRPAI) hosted by John McCarthy and Marvin Minsky in 1956.
• In this historic conference, McCarthy, imagining a great collaborative effort,
brought together top researchers from various fields for an open ended
discussion on artificial intelligence, the term which he coined at the very
event. Sadly, the conference fell short of McCarthy’s expectations; people
came and went as they pleased, and there was failure to agree on standard
methods for the field. Despite this, everyone whole-heartedly aligned with
the sentiment that AI was achievable. The significance of this event cannot
be undermined as it catalyzed the next twenty years of AI research.
The History of Artificial Intelligence - https://sitn.hms.harvard.edu/flash/2017/history-artificial-intelligence/
28. KEY MILESTONES
• 1943: Walter Pitts & Warren McCullogh develop a computer model based on Neural Networks of human brain using a combination of algorithms and maths
they called “threshold logic” to mimic the thought process.
• 1950: Alan Turing proposes the imitation game, aka “Turing Test”
• 1952: Hodekin-Huxley paper of brain as neurons forming an electrical network
• 1956: John McCarthy coins the term “Artificial Intelligence” and organizes Dartmouth Summer Research Project, the first conference on AI.
• 1960s: Research labs established at MIT, Stanford, SRI, etc. to mimic human intelligence by problem-solving or playing games like checkers or chess.
• 1960: Henry Kelley develops the basics of continuous Back Propagation (backprop) model
• 1962: Stuart Dreyfus develops chain rule to simplify backprop.
• 1965: Alexey Grigoryevich Ivakhnenko & Valentin Grigorʹevich Lapa develop Deep Learning algorithms using polynomial activation functions and statistical
analysis at each layer.
• 1970s: MYCIN was able to diagnose certain kinds of bacterial infections based on symptoms input.
• 1970s: A “prospector” expert system uncovers a hidden mineral deposit of porphyr molybdenum (a form of copper deposit) at Mount Tolman in the state of
Washington.
• 1973-80s: First “AI Winter”
• 1981: John Searle proposes “Chinese Room”
• 1980s: Development of Expert Systems bring some successes (e.g. DEC’s XCON)
• 1985-90s: Second “AI Winter”
• 1979: Kunihiko Fukushima develops “Neocognitron” the first Convolutional Neural Network (CNN) with multiple pooling and convolutional networks that
allows computer to “learn” to recognize visual patterns using manually-adjustable “weights” of certain connections.
• 1990s: Focus shifts to “Intelligent Agents”
• 1997: IBM Deep Blue beats World Chess Champion Garry Kasparov (Artificial Intelligence)
• 2011: IBM Watson beats human players on US game show Jeopardy (Machine Learning)
• 2012: Deep Learning
• 2014: Ian Goodfellow creates Generative Adversarial Networks (GANs)
• 2016: Google’s AlphaGo beats boardgame Go master Lee Sedol (Deep Learning)
29. AI EVOLUTION TIMELINE
The History of Artificial Intelligence - https://sitn.hms.harvard.edu/flash/2017/history-artificial-intelligence/
30. TODAY, AI IS
ATTRACTING
LOTS OF
ATTENTION!
https://www.economist.com/technology-quarterly/2020/06/11/an-understanding-of-ais-limitations-is-starting-to-sink-in
32. RESEARCH
• Published: The number of AI journal publications grew by 34.5% from
2019 to 2020—a much higher percentage growth than from 2018 to 2019
(19.6%).
• Pre-published: In just the last six years, the number of AI-related
publications on arXiv (pre-peer-review) grew by more than sixfold, from
5,478 in 2015 to 34,736 in 2020.
• Proportion: AI publications represented 3.8% of all peer-reviewed
scientific publications worldwide in 2019, up from 1.3% in 2011.
• Global shifts: In 2020, and for the first time, China surpassed US in the
share of AI journal citations in the world.
Source: Artificial Intelligence Index Report 2021, Stanford University Human-Centered Artificial Intelligence
33. PATENTS
• Number of patent
publications has
gone up 4.5x in last
20 years.
• Almost 3% of all
patents coming
from AI.
Source: Artificial Intelligence Index Report 2021, Stanford University Human-Centered Artificial Intelligence
36. ACCURACY
• Top-1 accuracy tests for how well an
AI system can assign the correct
label to an image, specifically
whether its single most highly
probable prediction (out of all
possible labels) is the same as the t
arget label.
• Top-5 accuracy asks whether the
correct label is in at least the
classifier’s top five predictions.
Figure 2.1.2 shows that the error rate
has improved from around 85% in
2013 to almost 99% in 2020.
37. ENGLISH
LANGUAGE
UNDERSTANDING
Launched in May 2019,
SuperGLUE is a single-metric
benchmark that evaluates the
performance of a model on a
series of language
understanding tasks on
established datasets.
SuperGLUE replaced the prior
GLUE benchmark (introduced
in 2018) with more challenging
and diverse tasks.
39. ACTIVITY
RECOGNITION
• The task of activity recognition is to
identify various activities from video
clips. It has many important everyday
applications, including surveillance by
video cameras and autonomous
navigation of robots.
• Drinking coffee remained the hardest
activity in 2020. Rock-paper-scissors,
though still the 10th hardest activity, saw
the greatest improvement among all
activities, increasing by 129.2%—from
6.6% in 2019 to 15.22% in 2020.
42. CORPORATE
INVESTMENT
• The total global investment in AI,
including private investment, public
offerings, M&A, and minority stakes,
increased by 40% in 2020 relative to
2019 for a total of USD 67.9 billion.
• Given the pandemic, many small
businesses have suffered
disproportionately. As a result, industry
consolidation and increased M&A
activity in 2020 are driving up the total
corporate investment in AI. M&A made
up the majority of the total investment
amount in 2020, increasing by 121.7%
relative to 2019.
43. STARTUPS
The following section analyzed
the trend of private investment
in AI startups that have
received investments of over
USD 400,000 in the last 10
years. While the amount of
private investment in AI has
soared dramatically in recent
years, the rate of growth has
slowed.
44. FOCUS
AREAS
• The “Drugs, Cancer, Molecular, Drug
Discovery” area tops the list, with
more than USD 13.8 billion in private
AI investment—4.5 times higher than
2019—followed by “Autonomous
Vehicles, Fleet, Autonomous Driving,
Road” (USD 4.5 billion), and
“Students, Courses, Edtech, English
Language” (USD 4.1 billion).
• In addition to Drugs, Cancer,
Molecular, Drug Discovery,” both
“Games, Fans, Gaming, Football” and
“Students, Courses, Edtech, English
Language” saw a significant increase
in the amount of private AI
investment from 2019 to 2020. The
former is largely driven by several
financing rounds to gaming and
sports startups in the United States
and South Korea, while the latter is
boosted by investments in an online
education platform in China.
45. EDUCATION
Faculty: The number of Ai focused faculty grew by 59.1%, from 105 in AY 2016–17 to
167 in AY 2019–20.
Courses: The number of courses on offer that teach students the skills necessary to
build or deploy a practical AI model has increased by 102.9%, from 102 in AY 2016–17
to 207 in AY 2019–20, across 18 universities
Students: the number of students who enrolled in or attempted to enroll in an
Introduction to Artificial Intelligence course and Introduction to Machine Learning
course has jumped by almost 60% in the past four academic years.
Doctoral: Among all computer science PhD graduates in 2019, those who specialized
in artificial intelligence/machine learning (22.8%), theory and algorithms (8.0%), and
robotics/ vision (7.3%) top the list.
46. SKILLS
PENETRATION
• The AI skill penetration metric shows
the average share of AI skills among
the top 50 skills in each occupation,
using LinkedIn data that includes
skills listed on a member’s profile,
positions held, and the locations of
the positions.
• For cross-country comparison, the
relative penetration rate of AI skills is
measured as the sum of the
penetration of each AI skill across
occupations in a given country,
divided by the average global
penetration of AI skills across the
same occupations. For example, a
relative penetration rate of 2 means
that the average penetration of AI
skills in that country is 2 times the
global average across the same set of
occupations.
47. HIRING
The AI hiring rate is calculated
as the number of LinkedIn
members who include AI skills
on their profile or work in AI-
related occupations and who
added a new employer in the
same month their new job
began, divided by the total
number of LinkedIn members
in the country. This rate is
then indexed to the average
month in 2016; for example, an
index of 1.05 in December
2020 points to a hiring rate
that is 5% higher than the
average month in 2016.
49. ADOPTION
BY REGIONS
The 2020 survey results
suggest no increase in AI
adoption relative to 2019.
Over 50% of respondents
say that their organizations
have adopted AI in at least
one business function
(Figure 3.3.1). In 2019, 58% of
respondents said their
companies adopted AI in
at least one function,
although the 2019 survey
asked about companies’ AI
adoption differently.
50. ADOPTION
BY
INDUSTRIES
In another repeat
from 2019 (and 2018),
the 2020 survey
suggests that the
functions where
companies are most
likely to adopt AI vary
by industry (Figure
3.3.3).
53. BOARD
Mentions of AI in corporate
earnings calls have
increased substantially since
2013, as Figure 3.3.11 shows.
In 2020, the number of
mentions of AI in earning
calls was two times higher
than mentions of big data,
cloud, and machine learning
combined, though that
figure declined by 8.5% from
2019. The mentions of big
data peaked in 2017 and
have since declined by 57%.
54. GOVERNMENT (US,
UK)
Issues around data privacy,
ethical use of AI and fear of job
losses, etc. is fueling the surge
of interest among governments
for evolving stronger
frameworks for policymaking
and governance.
55. AI
TECHNOLOGY
• What is AI?
• What does ML and
DL mean?
• What are various AI
technologies?
• What are the
limitations of AI?
56. GOOD OLD FASHIONED AI (GOFAI)
• Symbolic artificial intelligence is the term for the collection of all methods
in artificial intelligence research that are based on high-level symbolic (human-
readable) representations of problems, logic and search. Symbolic AI was the
dominant paradigm of AI research from the mid-1950s until the late 1980s.[1][2]
• John Haugeland gave the name GOFAI ("Good Old-Fashioned Artificial
Intelligence") to symbolic AI in his 1985 book Artificial Intelligence: The Very Idea,
which explored the philosophical implications of artificial intelligence research.
In robotics the analogous term is GOFR ("Good Old-Fashioned Robotics").[3]
• Researchers in the 1960s and the 1970s were convinced that symbolic approaches
would eventually succeed in creating a machine with artificial general
intelligence and considered this the goal of their field.
• However, the symbolic approach would eventually be abandoned, largely
because of the technical limits of this approach. It was succeeded by highly
mathematical Statistical AI which is largely directed at specific problems with
specific goals, rather than general intelligence. Research into general intelligence
is now studied in the exploratory sub-field of artificial general intelligence.
Symbolic Artificial Intelligence, https://en.wikipedia.org/wiki/Symbolic_artificial_intelligence
58. MACHINE LEARNING
• Term created by Arthur Samuel in 1959
• Machine learning is a branch of artificial intelligence (AI) and
computer science which focuses on the use of data and
algorithms to imitate the way that humans learn, gradually
improving its accuracy.
• Through the use of statistical methods, algorithms are trained to
make classifications or predictions, uncovering key insights
within data mining projects. These insights subsequently drive
decision making within applications and businesses, ideally
impacting key growth metrics.
Machine Learning, https://www.ibm.com/cloud/learn/machine-learning
59. AI IS THE NEW SOFTWARE!
• Machine learning was a huge leap from programmed instructions and if-then
statements that merely simulated the very human process of thinking and
making decisions.
• With machine learning, the machine no longer needs to be explicitly
programmed to complete a task; it can pour through massive data sets and
create its own understanding. It can learn from the data and create its own
model, one that represents the different rules to explain relationships among
data and use those rules to draw conclusions and make decisions and
predictions.
• A machine learning algorithm is a mathematical function that enables the
machine to identify relationships among inputs and outputs. The
programmer’s role has shifted from one of writing explicit instructions to
creating and choosing the right algorithms.
Rose, Doug. Artificial Intelligence for Business: What You Need to Know about Machine Learning and Neural Networks.
62. AI, ML, NN, DL…
• Machine learning, deep learning, and neural networks are all sub-
fields of artificial intelligence. However, deep learning is actually a sub-
field of machine learning, and neural networks is a sub-field of deep
learning.
• The way in which deep learning and machine learning differ is in how
each algorithm learns. Deep learning automates much of the feature
extraction piece of the process, eliminating some of the manual
human intervention required and enabling the use of larger data sets.
• Classical, or "non-deep", machine learning is more dependent on
human intervention to learn. Human experts determine the set of
features to understand the differences between data inputs, usually
requiring more structured data to learn.
Machine Learning, https://www.ibm.com/cloud/learn/machine-learning
67. LIMITATIONS OF AI
When people talk about AI, machine learning, automation, big data,
cognitive computing, or deep learning, they’re talking about the ability
of machines to learn to fulfill objectives based on data and reasoning.
This is tremendously important, and is already changing business in
practically every industry. In spite of all the bold claims, there remain
several core problems at the heart of Artificial Intelligence where little
progress has been made (including learning by analogy, and natural
language understanding). Machine learning isn’t magic, and the truth is
we have neither the data nor the understanding necessary to build
machines that make routine decisions as well as human beings.
https://hbr.org/2016/11/how-to-make-your-company-machine-learning-
ready
69. WHAT WOULD PEOPLE ALLOW AI TO DO IN
THEIR LIVES?
How far would people allow AI into their lives without
becoming uncomfortable?
It turns out, quite far. Two-thirds or more said they would trust AI
with handling medication reminders, travel directions,
entertainment, targeted news, and manual labor and mechanics.
More than 50% of respondents trust AI to provide elder care,
health advice, financial guidance, and social media content
creation. More than 40% said they would trust AI to cook, teach,
police, drive, and provide legal advice. On the other hand, using AI
in child care ranked at the bottom of the list.
What Do People — Not Techies, Not Companies — Think About Artificial Intelligence?,
https://hbr.org/2016/10/what-do-people-not-techies-not-companies-think-about-artificial-intelligence
70. WHAT DO PEOPLE THINK OF AI?
• More consumers see AI’s impact on society as positive than
negative (45% and 7%, respectively).
• People vary in how they understand AI.
• Only 8% of global respondents think AI is science fiction and will
never materialize.
• Trust in AI depends on experience and expertise.
• Consumers encounter AI on a frequent basis.
• Without a doubt, the probability of job loss due to AI was the
largest concern among respondents.
What Do People — Not Techies, Not Companies — Think About Artificial Intelligence?,
https://hbr.org/2016/10/what-do-people-not-techies-not-companies-think-about-artificial-intelligence
71. SO, IS MACHINE INTELLIGENCE JUST A
MACHINE MIMICKING HUMANS?
• Without a reliable standard for measuring human intelligence,
it’s very difficult to point to a computer and say that it's
behaving intelligently.
• Rose, Doug. Artificial Intelligence for Business: What You Need
to Know about Machine Learning and Neural Networks (p. 5).
Chicago Lakeshore Press. Kindle Edition.
73. (WHY) IS AI IMPORTANT?
• In 2013, one in 50 new startups embraced AI. Today,
one in 12 put it at the heart of their value proposition.
(MMC, 2019)
• Nine in ten AI startups address a business function
or sector (‘vertical’). Just one in ten provides a
‘horizontal’ AI technology.
(MMC, 2019)
75. ECONOMIC IMPACT OF AI
• Artificial intelligence (AI) can transform the
productivity and GDP potential of the global economy.
• Global GDP will be 14% higher in 2030
• $15.7 Tr contribution to global economy by 2030. Of
this,
• $6.6 trillion is likely to come from increased productivity and
• $9.1 trillion is likely to come from consumption-side effects.
• +26% boost in GDP for local economies by 2030
https://www.pwc.com/gx/en/issues/data-and-analytics/publications/artificial-intelligence-study.html
76. HOW WILL THIS ECONOMIC VALUE BE
ADDED?
• Labour productivity improvements will drive initial GDP gains as firms
seek to "augment" the productivity of their labour force with AI
technologies and to automate some tasks and roles.
• Our research also shows that 45% of total economic gains by 2030 will
come from product enhancements, stimulating consumer demand.
This is because AI will drive greater product variety, with increased
personalisation, attractiveness and affordability over time.
• The greatest economic gains from AI will be in China (26% boost to
GDP in 2030) and North America (14.5% boost), equivalent to a total of
$10.7 trillion and accounting for almost 70% of the global economic
impact.
https://www.pwc.com/gx/en/issues/data-and-analytics/publications/artificial-intelligence-study.html
77. BENEFITS
• the largest shares of respondents report revenue increases for inventory and
parts optimization, pricing and promotion, customer-service analytics, and
sales and demand forecasting. More than two-thirds of respondents who
report adopting each of those use cases say its adoption increased revenue.
The use cases that most commonly led to cost decreases are optimization of
talent management, contact-center automation, and warehouse automation.
Over half of respondents who report adopting each of those say the use of AI in
those areas reduced costs.
• The survey findings show that some companies using AI are seeing that value
accrue to the enterprise level. Twenty-two percent of respondents say that
more than 5 percent of their organizations’ enterprise-wide EBIT in 2019 was
attributable to their use of AI, with 48 percent reporting less than 5 percent.
• Additionally, in half of business functions, a larger share of respondents report
revenue increases from AI use than in the previous survey, while revenue in
most other functions remained stable. At the same time, cost decreases have
become less common in most functions
78. WHERE TO USE AI?
• If a typical person can do a mental task with less than
one second of thought, we can probably automate it
using AI either now or in the near future. – Andrew Ng,
https://hbr.org/2016/11/what-artificial-intelligence-can-
and-cant-do-right-now
92. PRACTICES FOR MANAGERS IN AN AI WORLD
• Leave administration to AI
• Focus on judgment work
• Treat intelligent machines as “colleagues”
• Work like a designer
• Develop social skills and networks
How Artificial Intelligence will redefine Management, https://hbr.org/2016/11/how-artificial-intelligence-will-redefine-management
93. ADOPTION CHALLENGES!
• While AI as a technology is fast
maturing, there are major barriers and
considerable challenges with adoption
of AI in businesses!
• Historically, AI has been overhyped
• Is there a risk to the future prospects of
AI in Business?
94. AI HYPE
• In 1970 Marvin Minsky told Life Magazine, “from three
to eight years we will have a machine with the general
intelligence of an average human being.”
95. FIRST “AI WINTER”: 1974-80
• In the early days, AI pioneers like Marvin Minsky made bold claims
that, “Within our lifetime machines may surpass us in general
intelligence.” However, limited and expensive computing power
and storage as well as a paucity of data meant that early
solutions could only solve rudimentary problems.
• Basically, AI promises by interesting “toys” failed to deliver and the
research fundings were cut!
• …but AI researchers continued underground research!
• Ended with the introduction of rule-based “Expert Systems”
96. SECOND “AI WINTER”: 1985-90S
• Expert systems were seen as slow, clumsy, difficult to update and
not able to “learn”
• Broken overpromises lead to disillusionment among investors.
• DARPA concluded AI would not be the next wave.
• AI achieved the dubious status of “pseudoscience”!
97. WHY AI PILOTS SELDOM MAKE IT TO
PRODUCTION?
• A 2017 Deloitte survey found that the number one obstacle to
successful AI deployments was that it was “difficult to integrate
cognitive projects with existing processes and systems.”
98. WORKER DISPLACEMENT
• The first challenge posed by AI will almost certainly be how to
support the people whose skills will be obsolete through
automation. - Rose, Doug. Artificial Intelligence for Business:
What You Need to Know about Machine Learning and Neural
Networks
99. ENTERPRISE ADOPTION ISSUES
• WSJ, 2020: AI Project Failure Rates near 50%
• Gartner, 2018: 85% of AI projects won’t deliver for CIO
• Issues of skills, talent, existing business models, scaling, up, trust
and fairness, etc.
102. REVENUES AND PROFITABILITY
• Only four Western and two Chinese AI companies report income, and all have
big losses.
• CrowdStrike and c3.ai both did IPOs and had losses equal to 30% and 40% of
revenues respectively in 2019, and 13% and 40% respectively in 2020.
• Nest’s losses were 85% of revenues in 2017[1] and DeepMind’s losses were four
and 1.7 times its revenues in 2018[2] and 2019 respectively[3] causing Google to
write off $1.3 Billion in debts.
• For China, Megvii’s cumulative losses had reached $1.4 billion by the first half of
2019, or 2.5 times cumulative revenues[4]. CloudMinds reported a net loss of
US$97.48 million in the six months of 2019 or 2/3 of its reported revenues[5].
• Furthermore, the big funding received by many Chinese (e.g., CloudWalk, Yitu,
SenseTime) and American startups (described below) suggest most if not all AI
startups are unprofitable.
Source: https://medium.com/swlh/few-ai-startups-release-revenue-numbers-much-less-income-numbers-will-any-of-them-ever-make-a-b99b48e54e74
103. THE THIRD AI
WINTER…???
This seems like the
perfect setting for a
very high-potential
but an overhyped
market with the
technological
successes much
ahead of the real
business results.
Could we be headed
to the third AI winter,
and if yes, how could
we avoid it?
104. FUTURE OF AI…
• AI is often seen as something
in the future!
• What does the future of AI
look like?
• Would it serve the society or
turn the society into a
dystopian nightmare?
105. POTENTIAL OF AI
• AI is the new electricity – Andrew Ng
• AI is more profound than fire or electricity
– Sundar Pichai, Google
• People should stop training radiologists –
Geoff Hinton, 2016,
https://youtu.be/2HMPRXstSvQ
106. WEAK (NARROW) AI VS STRONG (GENERAL)
AI
• Machine learning still qualifies as weak AI, because the
computer doesn't understand what's being said; it only matches
symbols and identifies patterns. - Rose, Doug. Artificial
Intelligence for Business: What You Need to Know about
Machine Learning and Neural Networks (p. 19).
110. RESEARCH QUESTIONS
• RQ1: What is Artificial Intelligence in the context
of business?
• RQ2: What are AI-enabled Business Models
(AIBM)?
• RQ3: What are the antecedents, barriers and
consequents of AIBMs?
• RQ4: What is the change management
framework in the context of AIBMs?
111. WHAT IS AN AI COMPANY?
• Per MMC Ventures, an investment firm in London, 40% of the
2,830 European AI companies don’t use any machine learning.
“AI-focused” startups receive 15% more funding! (MIT Technology
Review)
• Prof Ajay Agrawal et al have proposed an “AI Canvas” involving
Prediction, Judgment, Action, Outcome, Input, Training and
Feedback. However, it is not clear which of these really
constitute an AI company? (HBR)
• Stages of Human In The Loop (HILT) machine learning systems
include Fully Manual, Augmented Manual, Semi-automated, and
Fully Automated. What does it mean to be an AI company at
different stages?
112. RQ1: WHAT IS ARTIFICIAL INTELLIGENCE IN
THE CONTEXT OF BUSINESS?
There is a need to understand it differently from the underlying AI
technology in terms of what does it mean for a company to be an
“AI company”? In a 2019 survey of 2,800 European “AI companies”,
almost 40% of them did not use any form of machine learning
(MMC Ventures, 2019). Similar data from other geographies can’t be
ruled out. A clarity is required to establish the baseline for building
upon the research.
113. RQ2: WHAT ARE AI-BASED BUSINESS
MODELS?
Business models are described using tools such as
Business Model Canvas that recognize the power of
capital resources, key relationships and value-adding
processes, etc. However, adoption of AI inside an
organization could dramatically change the way an
organization operates on those aspects. For example,
how does a smart car become part of a smart fleet, or
autonomous agents auto-negotiate smart contracts and
conduct business unassisted by humans, or how humans
and algorithms conduct a medical procedure assisted by
each other? We propose to establish a framework to
describe various types of AI-based business models.
114. CURRENT STATE OF RESEARCH
• Although AI is linked to changes in business models, few studies have
investigated whether and how AI leaves its mark.. The Artificial
Intelligence paradigm has the potential to bring about many positive
financial and societal effects. Nonetheless, this study shows that
incumbents are still struggling to perform AI-enabled business-
model innovation. Using AI with transformed business models may
lead to failure if transformations are not linked to established actors
and other businesses in the wider set of ecosystem relationships. If
companies fail to build an ecosystem around their AI-based business
model that is fit for purpose, they may not succeed in the long run.
• Burström, T., Parida, V., Lahti, T., & Wincent, J. (2021). AI-enabled
business-model innovation and transformation in industrial ecosystems:
A framework, model and outline for further research. Journal of
Business Research, 127, 85-95.
115. RQ3: WHAT ARE THE ANTECEDENTS AND
CONSEQUENCES OF AIBMS?
• In the initial literature survey, we have come across over twenty-
five factors, including accuracy, speed, control, algorithm aversion,
human in the loop, data privacy, fairness and ethics, fear of job
losses and human rights abuse, etc.
• There is a lack of comprehensive and systematic research in this
field, and these factors are mostly referred to anecdotally and only
understood at a conceptual level. Further, there is a lack of
academic rigor in understanding these causal relationships.
• We propose to establish the causal framework that helps
establish clarity, and also paves the way for more successful AI
adoption in business models.
116. WHAT ARE THE CHANGE MANAGEMENT
FRAMEWORKS FOR AIBMS?
• Our preliminary research has thrown 25+ potential
variables that affect successful adoption.
• However, they don’t all affect consistently, and widely
vary across industries, etc.
• Given that existing businesses must transition to being
an AI company, a robust change management
framework is required to demonstrate what success
looks like, and how firms could establish their own
change journey
• We plan to study TAM, UTAUT (Venkatesh, 2003), NASA
Technology Readiness Levels, Innovation Ecosystems,
etc.
119. CAUSAL FRAMEWORK (PROPOSED/WIP)
AI-enabled Business Model (AIBM)
Industry Characteristics
Maturity of AI Technologies
Human-In-The-Loop (HILT) /
Decision-making processes
Predictive Capabilities
AI Expertize
Data Quality and Availability
Firm Characteristics
Process Efficiencies
Regulatory Framework
New Insights
Industry
Design
Social Impact
Economic Impact
Technology,
Talent
and
Firm
Industry
Society
Organization Design
Societal Design
Societal Values and Norms
Economic
Conditions
Business Performance
Employee Engagement
Planet Impact
Industry Business Models
Storage and Compute
Business Alignment
Antecedents Moderators and Mediators Consequences