Recent improvements in computing power allow users to rapidly process very large volumes of data. One area of computer science that has benefited from this is Artificial intelligence (AI); this includes an area of AI known as Machine Learning (ML). These technologies are starting to have a greater impact on work. This article will describe what AI and machine learning are, their possible application to insurance, as well as the limitations of AI.
What is artificial intelligence? A recent paper from McKinsey & Company defines AI as “intelligence exhibited by machines, whereby machines mimic cognitive functions… cognitive functions include all aspects of learning, perceiving, problem-solving, and reasoning.”[i] AI involves using machines with a form of cognitive skill to solve problems. A widely used artificial intelligence method is machine learning (ML). Machine learning is a “Major approach to realize AI by learning from, and making data-driven predictions based on data and learned experiences.”[i] Machine learning programs train on data-sets to create analytical outputs for solving problems.
How can AI and machine learning be applied to insurance? A recent paper from the Financial Stability Board noted AI can help underwriters by “sorting through vast data- sets that insurance companies have collected to identify cases that pose a higher risk, potentially reducing claims and improving profitability.”[ii] The paper also notes “AI and machine learning applications can substantially augment some insurance sector functions, such as underwriting and claims processing…Machine learning techniques can be used to determine repair costs and automatically categorize the severity of vehicle accidents.”[ii] AI and ML processes can be used to assist insurance professionals in solving problems by analyzing large volumes of data.
There are limitations to AI. The Financial Stability Board notes that “AI and machine learning tools might also miss new types of risks and events because they could potentially ‘overtrain’ on past events.”[iii] A recent paper from Harvard Business Review notes “the applicability of AI-based systems is still quite narrow…machine learning systems are trained to do specific tasks, and typically their knowledge does not generalize.”[iv] In short, the applications of AI are still limited, and AI can falter in dealing with situations where insufficient amounts of new data are available. AI systems have difficulty reaching conclusions which are where skilled insurance professionals come in; they can formulate questions, reach conclusions, and draw on a broad base of knowledge. As such, AI should be viewed as a tool insurance professionals can use to augment their work. Over time, insurance professionals may need to be retrained to work with AI systems that become more advanced.
In summary, artificial intelligence is a branch of computer science that mimics human cognitive behavior such as learning and problem solving. Machine learning involves systems analyzing data to make predictions. AI can be used to help underwriters by analyzing vast data-sets to help identify risks and determine the chance of loss. AI can help claims professionals determine the severity of a loss. Even so, there are limits to what AI can do. AI solutions are limited to solving very specific tasks, and AI systems lack general intelligence. In closing, AI and machine learning are data-driven tools that can improve the productivity of insurance professionals.
[i] Balasubramanian, Ramnath Libarikian, Ari McElhaney, Doug “Insurance 2030-The impact of AI on the future of insurance” McKinsey & Company April 2018 https://www.mckinsey.com/industries/financial-services/our-insights/insurance-2030-the-impact-of-ai-on-the-future-of-insurance
[ii] Financial Stability Board “Artificial Intelligence and machine learning in financial services Market developments and financial stability implications” Financial Stability Board November 1, 2017 page 14 http://www.fsb.org/wp-content/uploads/P011117.pdf
[iii] Financial Stability Board “Artificial Intelligence and machine learning in financial services Market developments and financial stability implications” Financial Stability Board November 1, 2017 page 25 http://www.fsb.org/wp-content/uploads/P011117.pdf
[iv] Brynjolfsson, Erik, Mcafee, Andrew “The Business of Artificial Intelligence” Harvard Business Review July 18, 2017 page 9 https://hbr.org/cover-story/2017/07/the-business-of-artificial-intelligence
Copyright © 2018 Central Mutual Insurance Company. All rights reserved.
The data for the artificial intelligence which is going to make the work very perfectly which will be going to make it more acceptable for the user to have it where it is mining the part.