In order to understand if AI can be good for insurance, we need to explore exactly what AI is and what sort of impact it could have, not just within insurance but across a full spectrum of industries such as technology, hospitality and healthcare to name but a few.
The theory behind AI aims to help us understand how we can make humans and machines more alike in a bid to make our daily lives more productive and more efficient. In practice, AI is a tool which can be deployed to help humans' complete complex tasks quickly and could exist as an extension to our very own cognition. With the help of AI, we as humans are able to carry out complex mathematical calculations in a very short space of time which can then go on to inform better decision-making processes, regardless of context.
According to this article by McKinsey & Company, artificial intelligence is:
“A machine’s ability to perform the cognitive functions we associate with human minds, such as perceiving, reasoning, learning, interacting with an environment, problem solving, and even exercising creativity. Voice assistants like Siri and Alexa are founded on AI technology, as are some customer service chatbots that pop up to help you navigate websites.”
There a several ways a computer can harness the power of AI, one of which is a hot topic in the media at the moment; machine learning. Machine learning is a form of artificial intelligence which allows a machine to use data to train its own decision-making processes:
“Machine learning algorithms can detect patterns and learn how to make predictions and recommendations by processing data and experiences, rather than by receiving explicit programming instruction. The algorithms also adapt in response to new data and experiences to improve their efficacy over time. Machine learning has already had impact in a number of industries, including achievements in medical-imaging analysis and high-resolution weather forecasting.”
In essence, with machine learning a computer is able to use reason to learn from huge datasets, much beyond the capacity of even the most intelligent human being. Through this, machines may be able to make pinpoint accurate decisions and take action according to its decisions about a particular dataset.
You may also have heard of deep learning. This is an enhanced type of machine learning which allows for a broader range of data to be introduced into the process. This could include images and audio alongside numerical or written data. Deep learning uses ‘neural networks’ which are set up to mimic the way neurons interact in the human brain. Deep learning allows AI to not only use reason, but also to make sophisticated predictions based on existing data. Deep learning adds extra dimensions to the complexities of regular machine learning through the use of a broad range of data types. A simple way to describe this would be that a machine learning model may be able to recognise the letter ‘A’ as the letter A, whereas a deep learning model may also be able to recognise a hand drawn photograph of the letter ‘A’ as the letter A, something that a simpler machine learning model may not be able to do.