How AI Increases the Value of Human Judgment

Artificial Intelligence - September 18 2019

As the insurance industry braces for the rise of machine intelligence, many insurance professionals are plagued with fear and uncertainty about how AI will transform traditional business processes, change how business is conducted today, and impact the workplace. We have all seen the headlines that refer to AI replacing humans and, in some industries like manufacturing, the reality of workforce augmentation is already here. It’s commonplace to see robots working alongside their human counterparts on assembly lines. Top car manufacturers have embraced robotics and leaders in the automotive supply chain are embracing artificial intelligence (AI) to take on tasks such as quality control.

For industries like financial services and insurance, machine intelligence is still making its debut. With the power to help organizations make better business predictions faster and cheaper, machine intelligence promises to provide greater insights. Companies looking to gain competitive advantage are securing early-adopter status while fully embracing the value of machine intelligence and AI.

Today, businesses collect massive amounts of data, but many business decisions are still predicated on the usage of very small amounts of data. What if real-time access to more data could drive better and more accurate strategic decisions. In the case of insurance, imagine if an underwriter or underwriting assistant could assess a risk based on 80% of the relevant data rather than only using 20% of the data provided on submission or application forms? Surely, decisions based on more knowledge would be better for the business.

The Art of Prediction

Recently, we hosted a webinar in conjunction with ACORD featuring Ajay Agrawal, best-selling co-author of Prediction Machines: The Simple Economics of Artificial Intelligence. In Prediction Machines, the authors argue that, as the art of prediction becomes cheaper, businesses will start using it more. Even problems that were not traditionally arithmetic problems will now be thought of as prediction problems that can be solved by machine intelligence. For data-intensive industries like insurance predicated on assessing risk, machine intelligence enables underwriters to better assess risk and make predictions in the face of uncertainty.

If we take a closer look at how machine intelligence is impacting insurance, a new prediction problem is submission triage. Prior to the digital age, we may not have thought of submission triage as a prediction problem, however, with machines, we can triage and route submissions faster and better than humans. With AI, we can extract data from unstructured insurance documents faster and more accurately than a human by parsing and contextually understanding data points and through the use of an API feed data into core systems like rating engines, policy management systems and AI-powered insurance applications, eliminating the need for humans to perform the mundane task of re-keying data. Machine intelligence can also be leveraged to auto-route and prioritize submissions to the right underwriting department for human risk assessment and decision-making.

The Value of Judgment

On the webinar, Agrawal stressed that prediction is only one piece of the puzzle. Let’s not forget the value of judgment. AI creates more opportunities for judgment – increasing the value of human judgement and creative thinking. AI has zero judgment, only humans have judgment. AI is not displacing humans, but rather empowering humans to focus more on the judgment role, as opposed to the combined prediction-judgement routine of decision making. If the art of prediction becomes faster, better, and cheaper, that will give way to more decisions to make. Giving human knowledge workers the opportunity to ditch the mundane, soul-sucking tasks like data entry to focus on more strategic, meaningful and rewarding work.

Business Value of AI

According to Notes from the AI frontier: Applications and value of deep learning published April 2018 by McKinsey & Company, they estimate that per industry AI’s value amounts to between one and nine percent of 2016 revenue. The value as measured by percentage of industry revenue varies significantly among industries, depending on the specific applicable use cases, the availability of abundant and complex data, as well as on regulatory and other constraints. As indicated below, AI creates enormous value in retail, healthcare systems and services, transport and logistics, travel and consumer packaged goods. And there is tremendous opportunity for AI to create value in insurance. Whether in the form of data extraction, underwriting workflows, claims processing, automated application intake and/or chatbots.  

McKinsey Graph

Growing Adoption of AI in Insurance

During the webinar, we polled participants on their adoption of AI. Based on the results, commercial insurance brokers and carriers are realizing the value that machine intelligence can deliver. Although still in the early adoption phase, artificial intelligence projects are on the rise, with organizations taking the plunge to launch pilots and the very early adopters deploying production use cases across specific lines of business hoping to gain competitive advantage and leapfrog their competitors.  

poll projects

We also examined how the early adopters are using machine intelligence. With data extraction being the most popular use case with 42% of organizations deploying AI for real-time data extraction – which is not surprising given the vast amounts of data collected by insurance companies. It’s clear that there is value in leveraging vast amounts of data collected data to make sound business predictions. For example, if we take a closer look at the Submission Triage or Application Intake process, today brokers are completing long questionnaires and providing carriers with enormous amounts of data. Carriers are leveraging the data from 10 to 15 questions to make their decisions. Basically, they are only using 10% of the data they ask for. Now imagine, if by deploying AI, carriers could have real-time access to 200 to 500 data points in seconds to make more-informed risk and pricing decisions based on 30 to 50% more data. It would be far more efficient and better for the business overall.

poll -use cases

AI promises to be the most disruptive class of technologies during the next 10 years due to advances in computational power, volume, velocity and variety of data, as well as advances in deep neural networks (DNNs)," said John-David Lovelock, research vice president at Gartner:

"One of the biggest aggregate sources for AI-enhanced products and services acquired by enterprises between 2017 and 2022 will be niche solutions that address one need very well. Business executives will drive investment in these products, sourced from thousands of narrowly focused, specialist suppliers with specific AI-enhanced applications."

Watch the On-demand Webinar

For more on the value machine intelligence is driving in the insurance industry, listen to our on-demand webinar Emerging Trends in AI, Data Extraction and Underwriting featuring Ajay Agrawal, best-selling co-author of Prediction Machines: The Simple Economics of Artificial Intelligence and Ron Glozman, CEO, Chisel AI. They discuss the value machine intelligence coupled with human judgement can bring to the insurance industry and how AI will open up new markets for insurance.

View the On-Demand Webinar

Browse different topics

Recent Posts