At the Movies: Artificial Intelligence, Robotics and the Human Touch
June 23, 2017 | By Deborah Grey
Humans have been preoccupied with robots for a very long time. Films such as Metropolis (1927), Blade Runner (1982), Bicentennial Man (1999), Her (2013) Transcendence (2014), every Star Wars movie, The Avengers: Age of Ultron (2015), and countless others reveal our intrigue with artificial intelligence (AI).
With a little ambition, anyone can study a subject, be it piano or plumbing. Far more impressive is when studying turns into mastery. Inevitably, we all have met people whose deep knowledge, wisdom from experience and natural ability causes us to see them as something special and distinct from their peers – true artists of their trade.
To connect the juxtaposition of AI and human expertise, let me point to another interesting example. Have you seen the movie Hidden Figures (2016)? The film is based on the book Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race. It is a compelling proof point for the importance of human expertise and intuition in the evolutionary and increasingly complex environment of human-machine interactions. The movie chronicles the story of three African-American women whose work calculating complex equations for NASA dramatically shaped the future of US space exploration and altered the course of the ensuing Cold War against the Soviet Union.
Characterized as “human computers,” these women represented mastery of mathematics at its highest level, their talent often outshining the most advanced technology of the day. Indeed, human computers were used through the 1970s by NASA. Susan Finley started as a human computer in 1958. In 2017, she continues to work at NASA as a subsystem engineer for the Deep Space Network.
In perhaps the most poignant example of mathematical mastery in the “Hidden Figures” chronicle, one of the women, Katherine Johnson, was tasked with computing the trajectories for Alan Shepherd’s flight in 1961. This work involved determining the exact path of Freedom 7 from liftoff to splashdown. Using her depth of knowledge and natural talent, Johnson double checked and reverse engineered the trajectory calculation of the most advanced computer of its day. When the computer’s numbers were found to be inaccurate, the astronauts had to rely on Johnson’s calculations. From that point on, astronaut John Glenn who did not completely trust the computer, refused to fly without Johnson manually computing his trajectories. In the movie Glenn says, “Get the girl to check the numbers…if she says the numbers are good, I’m ready to go.”
Now let’s connect the dots of AI and human expertise in the world of insurance. Commercial insurance underwriting, for example, can be an art. Allow me to explain. BOP or Artisan Contractor risks might fit into an automated underwriting structure and platforms being developed today. But the first risk I had to underwrite at ACE was 18 leased fire trucks in Iraq during wartime. This type of risk requires a significant human touch and intervention. There are simply too many coefficients and variables that might be missed in a purely automated risk evaluation process. Another account I had to underwrite was a boat manufacturer who, I found through the company’s website, was manufacturing and distributing much more than boats. The types of coverages this business needed were Product Liability, Defense Base Act and other specialty coverages that even a programmed robot might have missed due to the layering of risk factors.
Similarly, peer to peer insurance for homeowners allows insureds to get immediate quotes and bind a policy within minutes. Yet, if a homeowner has expensive paintings and other anomalies compared to a typical set of coverages, would it be in the best interests of the policyholder and the insurer to use an automated quote process? As Arun Balakrishnan, CEO of Xceedance explained regarding the convergence of insurance and technology:
“I feel very confident that in the next ten years insurers are going to be data and technology driven businesses… The role of the seasoned, more experienced management and underwriting staff is particularly critical in the transformation of the insurance workplace. If we free up their time from having to do menial, mundane and repetitive tasks, their wealth of knowledge can be put to use to make machines and software even smarter. Their role in building the capabilities of AI platforms will be paramount. Those platforms may never be able to replicate exactly the management and underwriting capabilities of experienced people, but they can come close enough to deliver huge productivity benefits for the business.”
I wholeheartedly believe in digital transformation. It adds efficiencies, productivity and accuracy to insurance operations and policyholder services. At a recent conference, speakers expounded the virtues of IoT and remote digital messaging. One example involved the failure of a home hot water heater. In this scenario, home-based digital sensors would automatically notify a plumber, schedule an appointment and order the failed part. The homeowner would not have to be present to manually address any of the work and logistics of the sensors, or to coordinate a connected network of technicians, suppliers and service providers. This was a functional example. Several companies, including Roost, provide sensor technology to alert homeowners about potential mechanical or system failures. And IoT connectivity with the warranty and insurance process cannot be far behind.
With the pace of innovation these days, it won’t be long before this type of technology is ubiquitous and commonplace in commercial properties as well. In fact, in the digital insurance ecosystem, enabling proactive, prescriptive approaches to risk management can provide both property owners and insurance companies cost efficiencies, time savings and a stronger business and service relationship.
The merits of digital enablement aside, there is also the parallel issue that many of us still want to be in physical control of our environments. We often don’t like other people making decisions for us, never mind machines, bots and smart networks. It remains to be seen if our hands-on habits and our fragile egos are able to handle the new paradigms of digital enablement and intelligent machines.
I personally think artificial intelligence and robotics are inevitably tied to our future business and personal interactions. Yet, human expertise will be a critical aspect of training AI and robots; and the human touch will remain key to leveraging the decision-making benefits that will emerge as a result of intelligent technology. Like many, I remain skeptical about the notion that a machine can be taught everything. Even with the most advanced programming, can machines really replace all the intuition and knowledge of humans?
Some say yes. According to a study by Celent, “Applying Conversational Commerce to Insurance: Aligning IT to the Machine World,” 1,820 insureds and 134 insurers were surveyed. What they found was that the biggest driver behind insurers’ adoption of AI and robotics for customer service is demand from policyholders.
The survey showed 37% of policyholders were using smart technologies to communicate with their insurance companies and actually preferred doing so compared to human interaction. Another 21% who had not yet used smart technologies were interested in trying it. To date, 25% of insurance companies have invested in machine learning while another 42% have plans to invest in it. Time will tell if this survey reflects an accelerating trend to rely on AI and robotics for transactional and service activities in insurance.
I do not believe machines will ever completely replace humans, no matter how hard we try to make a machine think and act as we do. While digital transformation can offer endless opportunities — from heightened efficiency and cost-savings to accurate results and predictive capabilities — it’s also a fact that experiential wisdom, personal interaction, raw talent, gut instinct, in short, the human factor, cannot be replaced in the strategic work and nuanced relationships of the business of insurance. As a result, rather than blindly accepting or staunchly resisting technological change, we must focus energy on initially learning how to coexist, then finetuning and optimizing human-machine interactions.
In this period of rapid and, at times, uncertain evolution in the insurance industry, we must realize that it is in everyone’s interest to merge the best of technological progress with the power and unmatchable strength of human wisdom and instinct. It’s also imperative that insurance companies optimize their operations through cutting edge automation and process improvements.
Deborah Grey supports business development and industry relations at Xceedance. Her diverse insurance career includes expertise in P&C insurance, including program business, Commercial Lines underwriting, product development, compliance, captives, fronting arrangements and reinsurance. Deborah has spent the past seven years helping insurance companies select and adopt legacy replacement systems, platform upgrades, BPO, managed services and IT services.