Business Isn't Like the Movies: There's No Leap into
the Future
Films help us envision technological developments. But we make
advances in things like AI and machine learning step by step, not in sudden
jumps.
BY HOWARD
TULLMAN, GENERAL MANAGING PARTNER, G2T3V AND CHICAGO HIGH TECH
INVESTORS@TULLMAN
The best movies are remarkably effective in showing us the near
and distant future, with rampant imagined technological advances being front
and center in the new visions. Blade Runner, Minority Report, The
Matrix, and even the James Bond films, all suggest and promote a flawless
future from a tech standpoint. There may be evil villains, but the gadgets and
gizmos are always great. And, as you'd imagine, we're willing to buy into
these escapist adventures and fulsome fantasies because we have this
long-embedded idea that stories about the future - regardless of how
far-fetched - help us imagine, visualize, and eventually actually create that
future. They don't call Hollywood "the dream factory" for nothing.
Even more importantly, from a mental health perspective, movies
conveniently end - usually by reassuring us of beliefs we
already hold or offering us lies that we'd like to believe in. Real life,
sadly, persists and drags on. Most of the time, when the made-up world wasn't
being destroyed, and before COVID-19 turned theatres into pandemic petri
dishes, good films were relatively harmless and more or less satisfying, if not
thought provoking. Films were refuges in which to avert our eyes and hide our
heads for a couple of hours from some of the ugliest realities of our
day-to-day lives. Tom Hanks's films from Forrest Gump to Apollo
13 to You've Got Mail were classic examples of the
movies ability to transport us to other times, places and worlds.
We've also learned that watching at home - even on the biggest
of screens and in the darkest of caves - isn't anywhere near as effective as a
diversion as the movie theater for a variety of reasons, including the fact
that the noise, clutter, and chores at home rarely take a break. Nonetheless,
in most cases, even watching a bad movie or reruns of decades' old TV shows is
still better - in bursts and binges - than facing the real world at least for a
relatively few minutes. Except when it isn't.
One of the biggest problems with these fantasies that inform our
views of the future is that they set the bar too high and too soon and they
subconsciously adjust our expectations accordingly. We have also come to
believe that the execution horizon is a fraction of what it actually turns out
to be. As a result, in many instances, what we should see as amazing and
radical steps forward in a number of areas -- real-time data analysis, enhanced
and increasingly reliable and accurate predictions, artificial
intelligence/machine learning -- are happening so often and in implementations
so close to what we've seen for years on the big screens, that we're slightly
weary rather than wondrous. We continually take these things largely for
granted and mentally discount just what difficult and remarkable achievements
they represent.
One simple example is the increasing ubiquity of voice as a control mechanism and operating system for our
homes and vehicles. Ever since 1968, when the HAL 9000 in 2001:
A Space Odyssey spoke up and took command of the spaceship, we've
largely assumed that having machines speak and understand our zillions of
languages, accents, and idioms was a trivial matter. The reality is that -
notwithstanding Alexa, Siri and the passage of more than 50 years - we're still
quite far away from any kind of comprehensive and global speech recognition
solutions.
Having said that, customer support and management companies
like Balto, whose call center systems listen to
live conversations with consumers and alert managers when to intervene or join
discussions, are changing and dramatically improving the performance and
effectiveness of account reps and other customer-facing personnel by
monitoring, interpreting and analyzing millions of conversations daily.
Digital people (think chat bots on steroids) are just around the
corner and companies are already creating increasingly convincing robotic
performers. Soul Machines uses your web cam and
microphone (with your consent) to help its humanistic digital performers hold
convincing conversations with you and actually react to your expressions, tone
of voice, and other micro visual tells by reading and instantly interpreting
your responses.
Image recognition is another area where decades of
indoctrination - especially in military movies and spy films - have conditioned
us to believe that the computer is already all-seeing and able to track and
identify virtually any object. Here again, reality is definitely starting to
catch up and companies like Baidu have created systems that are
demonstrably faster and more accurate than humans at image checking and
sorting. AI-enabled devices are now 20%-to-30% faster and more accurate at
reading X-rays than radiologists, and at detecting certain types of skin cancer
earlier and more accurately than the human eye.
But it's usually a profitable exercise to take a step back from
the gee-whiz hype and to focus on the current practical applications that are
also constantly accelerating and creating new operating environments. These
apps, often AI-driven, provide real and immediate competitive advantages in
somewhat more mundane contexts. In addition to the obvious economic benefits to
the businesses that use them, seeing these new technologies and approaches in
operation is a far more valuable and accurate predictor and a powerful
indicator of what further enhancements and extensions we can reasonably expect
in the near term.
The most important operational innovation in most cases doesn't
come from lightning leaps and bounds but from steady, incremental application
and improvement. This is an iterative and cyclical process of successive
approximation, where the end product is always getting a little bit better
rather than postponed perfection -- which never arrives, except in the movies.
You launch, you observe, you measure, you analyze, you modify and then you do
it all again. If it's done right, it's like a flywheel where each revolution or cycle gains speed and momentum
and builds toward a better result.
One example of the total integration of visual technology,
advanced predictive tools, and important elements of AI and machine learning
into a set of practical and game-changing offerings is Precitaste,
which develops Vision AI for the food industry. Starting with external inputs
about expected weather, folding in real time drive-in traffic volumes combined
with walk-in customers (all captured by on-site cameras), adding historical
daily product order volumes, the QSR Brain system is able to track existing
product availability, instruct the back-of-house staff to begin preparation and
increase units of various products where necessary even before they are
ordered. The system also moderates existing product levels on the fly to avoid
excessive inventory and product waste due to fluctuations in demand.
The overall Precitaste system learns and grows
"smarter" every single day based on the unique experience at each
location, so that it can also create going-forward inventory demand and supply
projections -- which typically exceed 90% accuracy when later compared to the
actual results for given time periods. This permits highly specific analysis,
extremely precise personnel planning and scheduling and largely avoids all of
the typical "one size and one strategy fits all" approaches that are
typical of fast-food chains. The reduction in product shortfalls and the
ability of the QSR Brain to anticipate slack times as well as surges reduces
the stress on staff and demonstrably speeds and improves the overall customer
experiences.
On the one hand, building better, faster burgers doesn't seem to
be that big a deal. But the real message is directional. It's hard to imagine
that any manufacturing business, any production operation involving
perishables, or any customer-facing service organization won't be employing
systems like these to better inform, project and operate its business in the
near future.
More and more of the real-time decision-making will be driven by
data and metrics captured instantly and fed into systems like the QSR Brain.
And out will come improved results, better customer and employee experiences,
cost improvements and material reductions in waste. All automatically, somewhat
mystically, and in real time.
As the saying goes, any sufficiently advanced technology is
indistinguishable from magic.