26 ISE Magazine | www.iise.org/ISEmagazine
April 2021 | ISE Magazine 27
How the intelligent
enterprise will drive
innovation
Blending AI tools, diverse human thought
can drive future business creativity
By Joseph Byrum
28 ISE Magazine | www.iise.org/ISEmagazine
How the intelligent enterprise will drive innovation
The economic powerhouses of the future are the
ones investing in artificial intelligence (AI) today.
Machine learning and deep learning techniques
are well-suited to creating self-adjusting algo-
rithms that optimize business processes. Busi-
nesses across a number of industries are betting
big that this technology will achieve greater business ef-
ciency.
But this represents just the starting point for what AI
should be able to ultimately accomplish. A good way to
explore the full range of possibilities is to imagine what
an enterprise would look like if it were designed from the
ground up to use AI to optimize not just the core busi-
ness functions, but every aspect of operations from grounds
maintenance to the executive suite. The presumption is
that the whole company – call it the intelligent enterprise –
would become something greater than the sum of its fully
optimized parts.
As well see, the intelligent enterprise of the future has
the potential to become a creative engine ideally suited to
driving innovation in the decades ahead.
AI creates order from chaos
Basic AI algorithms can process mountains of data and
match highly complex patterns. More sophisticated expert
systems can look at these data to formulate possible ex-
planations for why things might be happening. The fuller
implementation of AI is what executives need to manage
a market that is chaotic, complex
and in a perpetual state of flux.
Business leaders have a limited
amount of time to make critical
decisions with only partial or even
contradictory data in hand. Even
for the best leaders, identifying fac-
tors that have the greatest impact
on success often end up involving
quite a bit of guesswork. Big data
was supposed to solve this problem
by providing a scientific basis for
decisions, but in many ways it has made the situation worse
due to data overload. Executives frequently complain of be-
ing data rich and insight poor, meaning they have plenty of
information but it’s not useful without a clear idea of what
to do with it.
Properly implemented, AI offers an alternative to luck
and guesswork, with systems that can make it possible to
build a scalable, insight-driven enterprise. One approach to
making this happen would be to model the intelligent en-
terprise on the classic decision-making framework known
as the OODA Loop, which stands for observe, orient, de-
cide and act. The OODA Loop is a method for managing
complexity and bringing order to chaos so an executive has
greater situational awareness and understanding to make
better informed decisions.
Briey explained, the loop begins by “observing” data
without filtering against preconceived notions of what the
data are supposed to mean. This step looks at facts straight
up, free from the coloring of analysis. The idea is to absorb
as much high quality information as possible to gain an un-
derstanding of the current situation, even though the data
can be incomplete or contradictory.
Making sense of the data is the task left to the orient
stage, which formulates theories to explain the observed
facts. As the facts change, successful orientation means
coming up with new interpretations that serve as potential
explanations for the situation. It’s possible, even desirable,
to come up with multiple explanations, even if they might
seem implausible at the time. The point is to come up with
multiple options.
It’s the task of the next stage to decide which explanation
and which course of action best fit the data. The choice
between interpretations of the facts sets up possible courses
of action to achieve the desired goal. This is why having
more options from which to choose can be beneficial, en-
hancing the ability to pivot quickly when the data take a
turn that might be considered unexpected to those who are
unprepared.
In the final stage, it’s time to act by executing the choice
and evaluating the results. This method then “loops” be-
cause every choice affects the field
of action – that means it’s time to
return to the beginning and ob-
serve the impact of each choice af-
ter it has been made. The process
repeats until the desired results are
fully achieved.
The OODA Loop is a system-
atic approach to thinking through
problems in a time-constrained
environment in a way that forces
constant evaluation and reevalu-
ation of options and results. This keeps decision-makers
from becoming complacent and assuming what has worked
in the past will work again in the future. Instead, this new
way of doing things helps the manager adapt quickly to
confusing, chaotic and constantly changing circumstances
by developing a greater awareness of what is happening,
and why.
AI systems in the intelligent enterprise would automate
this process by analyzing data and using expert systems
to formulate potential courses of action. Human opera-
tors would take this information and make decisions about
which way to act. The AI systems would continually track
T
Executives frequently complain
of being data rich and insight
poor, meaning they have plenty
of information but its not useful
without a clear idea of what
to do with it.
April 2021 | ISE Magazine 29
conditions and results and provide updates in real time.
This would provide needed context and factual backing
for the decision-making process, resulting in higher quality
choices. While this would seem to make AI the star of the
intelligent enterprise, it is actually just half of the puzzle.
Humans have an equally important role.
Humanitys role in the intelligent enterprise
Many imagine a future in which artificial intelligence fi-
nally replaces humans by doing all of our work so we can
lie back on the beach and enjoy the fruits of machine labor.
While this may seem supercially appealing, such a future
is as unrealistic as it is undesirable. Besides the inevitable
overcrowding on the beaches this would cause, it would
not be an efficient use of resources.
AI, and machines in general, are good at some things and
terrible at others. Algorithms have unmatched vigilance
that allows them to monitor the smallest details around the
clock without complaint. Machines never forget and are
unmatched when it comes to numeric computation.
On the other hand, AI is woefully inadequate when it
comes to recognizing nuance or subtle contextual changes.
Machines are very literal devices with no ability to detect
irony or appreciate a simple joke. Joke recognition itself
How the Turing Test still has AI stumped
Can machines think? This was a valid question
generations ago before today’s level of artificial
intelligence and machine learning became a
reality.
And yet a simple test created by famed
mathematician, scientist and World War II code
breaker Alan Turing has proven to be elusive even
for today’s modern AI programs and shows that
human brains are still better able to handle some
types of decision-making.
In 1950, Turing published a paper, “Computing
Machinery and Intelligence,” describing a way to
determine if a machine had intelligence. It involves
a game with an evaluator and three respondents,
two humans and a computer. The evaluator asks
open-ended questions to each and determines by their answers which is the machine. If the computer can fool the judge, it is deemed
to have intelligence.
The simplicity of The Turing Test was also its genius, and it turns out, its reliability. Since the theory was proposed, no AI system
has been able to effectively beat it. While humans are harder to fool with trick questions, even the most advanced machine lacks the
ability to spot such nuances.
Experts explained why in a recent story in Forbes.
“When you ask a GPT-3 system how many eyes the sun has, it will respond that there is one and when asked ‘Who was the president
of the U.S. in 1600?,’ the answer will be Queen Elizabeth I,” said Noah Giansiracusa, an assistant professor of mathematics and data
science at Bentley University. “The basic problem seems to be that GPT-3 always tries in earnest to answer the question rather than
refusing and pointing out the absurdity and unanswerability of a question.”
As AI technology advances, the belief is that it someday may develop the ability to decipher such responses. New tests have been
developed, according to Druhin Bala, CEO and co-founder of getchefnow.com, that can better gauge AI’s perceptive abilities.
“If Alan Turing was alive, he might be shocked that given 175 billion neurons from GPT-3 we are still unable to pass his test, but we
will soon,” said Ben Taylor, chief AI Evangelist at DataRobot.
“The Turing Test is brilliant in its simplicity and elegance, which is why it’s held up so well for 70 years,” said Zach Mayer, vice
president of data science at DataRobot. “It’s an important milestone for machine intelligence, and GPT-3 is very close to passing it. And
yet, as we pass this milestone, I think it’s also clear that GPT-3 is nowhere near human-level intelligence. I think discovering another
‘Turing Test’ for AI will illuminate the next step on our journey toward understanding human intelligence.”
30 ISE Magazine | www.iise.org/ISEmagazine
How the intelligent enterprise will drive innovation
isnt too relevant at the enterprise level (though it could ac-
tually be important for automation in the human resources
department), but recognizing nuance is critical to under-
standing complex adaptive systems in general. For example,
understanding the subtle shades of human behavior is es-
sential for identifying the potential underlying causes of
market shifts that are rooted in the collective decisions of
its human participants.
Where machines are weak, humans often excel. Our
brains are hard-wired to instantly recognize familiar faces
in a crowd and detect emotions from slight shifts in tone
or facial movements. Our ability to appreciate nuance con-
tributes to our ability for creative expression that machines
simply lack.
Likewise, where machines excel, humans are weak.
We’re terrible at multiplying 10-digit numbers and re-
membering details with precision. Humans also become
bored easily with repetitive tasks. So if our goal is to drive
across Kansas on a flat, straight road
on a sunny day, AI is going to do a
much better job than a human. But
if the task is driving on a snowy day
in Boston, youre much better off if a
human takes the wheel.
The intelligent enterprise is de-
signed to take advantage of the com-
plementary nature and the symbiosis
of man and machine. From biology,
we know different species of living things have founds ways
of living in harmony with one another. Commensalism is
the term used to describe what happens when one side gets
most or all the benefit in a symbiotic relationship but the
other side doesnt mind being used. Orchid flowers, for ex-
ample, often attach to trees to access the sunlight they need
to survive. They arent parasites since the tree is not harmed
by the flower’s presence but the tree does not derive any
particular benefit.
A better arrangement is known as mutualism, which de-
scribes species that form a partnership allowing both sides
to benefit. In the wild, for example, zebra herds are often
spotted teaming up with ostriches. This peculiar collabora-
tion between bird and mammal starts to make sense if one
evaluates the relative strengths and weaknesses of each. The
ostrich has mediocre senses of hearing and smell, but supe-
rior eyesight. Zebras are the opposite, with terrible vision
and great senses of hearing and smell. By grouping togeth-
er, they draw upon a superior sensory defense that greatly
enhances their joint chances of survival against predators.
Optimizing humans to deal with AI
For man and machine to work in similar harmony, human
employees will have to make the most of their strengths.
That means workers in the intelligent enterprise should
come from a different background, upbringing and edu-
cation to maximize diversity of thought. Each new hire
would be expected to bring new perspectives and insights
to the table to avoid one of the most common pitfalls of
large organizations: groupthink. As we saw, having mul-
tiple perspectives is a key aspect of success when using the
OODA Loop, and drawing upon diversity through hiring
decisions can deliver those unique perspectives.
But it takes more than cognitive diversity to build an ef-
fective team. The workers of the intelligent enterprise would
need to draw from a common foundation of training that’s
rigorous enough to allow the workers to extract maximum
value from AI assistance. This would mean taking individu-
als who could be from opposite extremes on the educational
spectrum – say a liberal arts graduate who studied history in-
stead of math and an electrical engineer – and sending them
through a common set of programs like Certified Analytics
Professional and Project Manage-
ment Professional.
While this is obviously a tougher
path for the historian to take, it en-
sures the unique perspective that this
calling has to offer can be effectively
communicated to other team mem-
bers. It creates a common language
that every other employee can un-
derstand. You keep the benefits of
cognitive diversity without losing the scientific rigor.
Organized in this fashion, humans can work well with
one another and with machines, as the workforce is primed
to bring new perspectives to every situation. The human op-
erators can then make the most of their creativity when us-
ing these tools at every level of the organization. All of these
factors combine to make the intelligent enterprise a unique
entity that’s more than the sum of its parts. By honing the
adaptability and understanding of the relationship between
man and machine, the intelligent enterprise of the future
will drive innovation beyond anything we’ve seen before.
Joseph Byrum is chief data scientist at Principal. He was previously
senior R&D and strategic marketing executive in life sciences-global
product development, innovation and delivery at Syngenta. In that
role, he was chief architect of initiatives that won Syngenta the
2016 ANA Genius Award in Marketing Analytics and 2015
Franz Edelman prize for contributions in operations research and
the management sciences. His bachelors degree in crop and soil
science and his masters degree in genetics are from Michigan State
University. He earned a masters degree in business administra-
tion from the University of Michigan Stephen M. Ross School of
Business. His Ph.D. in quantitative genetics is from Iowa State
University. He is an IISE member.
The intelligent enterprise is
designed to take advantage
of the complementary nature
and the symbiosis of man
and machine.