By Charles Dugas, Insurance AI Lead, Element AI —
As artificial intelligence reshapes the insurance industry, insurers that embrace AI early on will gain a key competitive advantage—but it may not be the advantage you think. By building the position and experience that accrue to adopting AI early, insurers can redefine intelligent decision-making for themselves and their industry.
AI – Data at your service
Since its earliest days, insurance has been the business of applying best use of the best possible information. The computing revolution has helped insurers pursue this goal, offering continuous yet incremental improvements to information collection and analysis over the years. Historically, insurers have been prudent about the speed with which they adopt new computing technologies, compared to the quicker pace of peers in other industries.
AI is big news now because it’s not an iterative advancement on the insurance software of yesterday. It represents an entirely new approach to designing software. As a result, insurers may wish to rethink their adoption timing.
The unique strength of AI revolves around its ability to tame information, rendering it useful and at your service. Applied to the decision-making that constitutes the core work of insurance, AI systems change the definition of best information itself.
By automatically performing some informational tasks, and providing recommendations and visualizations for others, AI systems are uniquely capable to provide contextually specific, explainable support for every insurance function: from actuaries, to underwriters, to claims adjusters, to leaders.
An AI system can trawl oceans of digital-age data, catching information that has remained elusive until now, and translating it into actionable insights.
This is news because when everyone can work smarter through augmentation with AI, the definition of the best use of the best possible information needs rewriting.
Everything fit to blueprint
For insurers—even those with an abundance of caution—the competitive imperative to adopt AI is clear. AI does not make the world itself less complex, but it does make complex decision-making easier by triggering a fundamental shift in how data-rich organizations sense and draw conclusions about the world.
As AI techniques advance in the coming years, adoption will become a necessity. And the advantage of early adoption will extend beyond the usual benefit of becoming incrementally more efficient or scalable.
Insurers who adopt AI first can leverage an edge on competitors who are not able to grasp the same information as deeply, widely, or cost-effectively. By moving early to discover their unique vision for how to best augment their people with AI, they will also gather more experience redefining the best use of information over time. Both advantages grow over time and scale with the power of digital platforms.
These first-mover advantages are important because they enable insurers operate with greater precision, accuracy, and reach than ever before—an increasingly complex endeavor in an age of growing risk and increasing demand for personalized services.
Smooth sailing for AI adoption
Though AI presents significant new advantages for insurers, it remains true that any change impacting an insurer’s decision-making must follow a few key guidelines.
The good news for insurers is that the way AI systems become useful is by learning from examples—previous instances in which the good judgment of experienced employees has been tracked in data. By learning directly from insurance decisions of the past, and taking new learning from underwriters and adjusters as they use the system, AI systems help reinforce and scale good judgment.
We can distinguish three rules for improving decision-making with information, and how AI should be deployed to support each rule.
1. The goal of better information is to unlock better decision-making.
AI must be directed at improving decision-making: not just automating decisions or creating new analytical insights.
2. Best use of information is a moving target.
AI systems already handle judgment and uncertainty in a similar way to human experts: by using a probabilistic model for making predictions. AI systems must also learn from underwriters or adjusters how to improve and adapt to new conditions over time.
3. Trust underpins everything.
AI systems must earn and maintain the trust of insurers and their experienced employees by providing explainable decisions and a meaningful interface through which to to evaluate and correct learnings.
With the right AI technology, as well as controls to use that technology responsibly, insurers can ensure smooth sailing no matter which part of their AI journey they’re on.
By understanding the benefits that accrue to those who begin this journey early, insurers can also make better decisions about when and how to begin adopting AI. Those who adopt AI first will rewrite the book on the best use of best possible information, seeing improvements in their competitiveness, bottom line, and perhaps even shift the insurance industry as a whole into a new era of intelligent decision-making.
About the Author
Charles Dugas joined Element AI in September 2017 as Director of Industry Solutions where he leads the insurance vertical, helping carriers and other insurance market participants develop and implement a strategic vision and roadmap for the rollout of AI capabilities across their organizations. Charles holds a B.Sc. in Actuarial Science, a M.Sc. in Electrical Engineering, and a Ph.D. in Machine Learning with Yoshua Bengio. He has 20 years experience developing and implementing analytical solutions in financial services, either as an entrepreneur, a professor, or manager of large corporations’ analytical teams.
About Element AI
Element AI was co-founded in 2016 by seasoned entrepreneur JF Gagné and leading AI researcher Yoshua Bengio, PhD. Located in the deep learning hub of Montréal, Canada, Element AI also has offices in Toronto, London UK, Singapore and Seoul. The company is committed to promoting collaborative research models to develop AI products at scale to help people work smarter. To learn more about Element AI products and services, visit www.elementai.com or contact [email protected].
Source: Element AI