An Intelligent Approach to Artificial Intelligence

Marvin Stone, SVP, Business Integration / Innovation Council Leader, Stewart Title

Marvin Stone, SVP, Business Integration / Innovation Council Leader, Stewart Title


Everywhere you turn, everyone is talking about artificial intelligence. Today’s executives are besieged with news of early-stage startups and AI-focused tech titans looking to disrupt virtually any market. With AI dominating the headlines, many industry leaders have noticed a lingering feeling that AI is the answer—the answer to what, however, remains unclear.

"AI is advancing at an incredible pace, resulting in ever-increasing capability at lower and lower costs, which will only fuel the number of disruptors looking to enter the fray"

Compounding this growing feeling of anxiety is the fact that AI represents an enormous step forward in the evolution of technology. The tech of the past replaced physical tools with simple systems: the word processor swept aside pen and paper, and a single database could replace rows of filing cabinets. We told computers the rules, and they dutifully carried them out. Modern AI applications recognize patterns and solve problems independently, though. They more closely resemble the complexity and creativity of the human brain than any previous technology. It’s a new game, and the old rulebook won’t help us here.

The Rise of AI

Over the last ten years, AI moved from the silver screen to the computer screen. Numerous uses of AI began to emerge in fields as diverse as medicine, social media, and perhaps most significantly, online retail. Amazon has spent the last decade pouring resources into extremely sophisticated AI across many of their divisions. When NASA developers turned to Amazon AWS to process nearly 200,000 Cassini images in a few hours for less than $200, they knew the decision wasn’t rocket science.

The tools that analyze image, video, text, and voice patterns form the foundation for more specialized applications in areas like language translation, text-to-speech, voice search, facial recognition, and even chatbots. Lately we’ve been hearing about even more advanced use cases such as product recommendation, sentiment detection, fraud prevention, industrial robotics, and even advanced medical diagnostics.

VC and the AI Imperative

Amazon, Microsoft, Google, and IBM (think Watson) offer more tools than most AI startups could ever dream of—most at very low cost. Because the ubiquity of these tools eliminates barriers to entry, the number of AI startups focused on the insurance industry continues to swell. Just a few examples are:

• Bright Health, which is focused on AI decisioning in healthcare and raised $240m
• Metromile’s on-demand insurance that has raised over $200m
• Oscar Health, who leveraged AI and big data in new ways and raised nearly $1b

Dozens of AI-focused insuretech startups have raised early rounds of $1m to $20m or more. Lemonade, the poster-child for insurance industry disruption, has raised over $130m in just two years.

With no capital expenditures and no limit to the scale that can be achieved, industry analysts expect the number of cloud-based insuretech AI startups will continue to grow.

Insurance Industry Use Cases

When it comes time to garner budget resources for AI R&D, it is good for executives to understand how AI is already being used. Here are just a few of the more interesting, and high return, use cases:

1. Automation - Most technology professionals in large insurance-related enterprise environments have a multitude of homegrown or third-party applications that are difficult, if not impossible, to integrate. AI-powered Robotic Process Automation (RPA) has flourished among financial services firms in recent years.

2. Document Classification - Many processes still rely upon document identification and classification. Determining what type of document is being viewed, whether it was completed correctly, and then “bucketing” those documents by difficulty, or by exception, can be a high-return activity—even if it’s not worthy of a press release.

3. Risk Elimination - Disruption may not come from inside the enterprise, but from external uses of AI and big data by outsiders. An often-quoted report by KPMG envisions a world with “radically safer” vehicles—including autonomous vehicles—that could someday bring a reduction to the auto insurance industry of 50 percent or more.

4. Customer Experience - Expert chatbots can carry out full conversations with customers (B2B and B2C) that take place via text chat or even via speech on phones and smart speakers. Customers of all kinds want choice, and bots bring choice around the clock.

5. Improved Claims Handling - Image recognition is an AI toolset that has improved greatly in recent years. It is possible that AI routines could detect “before and after” differences to determine amount of loss in a wide range of use cases.

Your AI Action Plan

Few insurance industry IT organizations have resources on the bench. Fewer still have resources dedicated to AI. That being the case, it’s time to consider an easy-to-implement action plan:

1. Watch - Have your team start by watching YouTube videos showing the tech-giants’ AI capabilities. These short videos are fascinating and will pique interest, making them perfect for informal lunch-n-learn settings.

2. Educate - As previously stated, artificial intelligence can be tough to grasp for executives who have only been trained in the ways of rules-based computing. Begin with very high-level briefs or presentations moving from general to industry-specific AI topics.

3. Learn - MOOC’s such as edX can open your team to deeper AI learning opportunities from University of California San Diego, MIT, Berkeley, CalTech, Columbia University, and others. Amazon and Microsoft also have excellent self-paced learning programs.

4. Experiment - One aged IT maxim that still holds true today is “what is easy and fun, always gets done.” Developers and infrastructure leaders like working on new technologies—so much so that this kind of cutting edge work is usually considered reward in and of itself. Check out Amazon’s SageMaker as a starting point.

5. Monitor - Use Twitter and Google Alerts to keep an eye out for potential disruption from existing competitors and startups alike. Be sure to follow #AI and #Artificial Intelligence on Twitter.

6. Envision - Work with risk executives, the leadership team, and P&L owners to target specific use cases (e.g. Amazon retail) that might be improved with AI to develop a roadmap. As the saying goes, if you don’t know where you’re going, any road will take you there. Even a rough road map is better than nothing at all.

Insurer, Disrupt Thyself

The insurance technology field is more demanding than ever given increased regulation, growing cybersecurity threats, and more. Existing infrastructure and application sprawl continue to consume over 80 percent of IT budgets. Insurance executives must purpose to focus efforts on the technology voted most likely to disrupt—artificial intelligence.

AI is advancing at an incredible pace, resulting in ever-increasing capability at lower and lower costs, which will only fuel the number of disruptors looking to enter the fray. Somewhere today, another VC-funded startup with a clear vision, a fresh pile of money, and laser-focused ambition is gaining on you like an inexorable self-driving Tesla. Remember, “objects in mirror are closer than they appear.”

Weekly Brief

Top 10 Artificial Intelligence Solution Companies - 2020

Read Also

How Insurance Solutions Can Help Protect Tech Companies as They Grow

How Insurance Solutions Can Help Protect Tech Companies as They Grow

Andrew Zarkowsky, Head of Technology Industry Practice, The Hartford Nick Kreczko, AVP, Hartford Stag Ventures and Dan Neubelt, Senior Analyst, Hartford Stag Ventures
The Future of Insurance is Embedded

The Future of Insurance is Embedded

Alexandre Moreillon, Principal, CommerzVentures
Risk and Recommended Controls for Models and Artificial Intelligence Systems

Risk and Recommended Controls for Models and Artificial Intelligence Systems

Eric Bonnell, SVP, Manager - IT Risk, Atlantic Union Bank
The Recipe of Success for Insurtech Startups

The Recipe of Success for Insurtech Startups

Michael Baptista, Founding Partner, Assembly Capital Partners