Are we moving away from the traditional insurance concept of risk sharing and will each insured eventually end up paying for their own, precisely quantified risk, leaving many uninsured, or have we always been doing this, albeit in a much less sophisticated way? Has insurance lost its way? Think again.
The CEO of OWLS™ Software, Tavio Roxo, gave a thought-provoking account at the InsurTech 2021 virtual conference, discussing whether the insurance industry has lost its soul with its current objectives and ultimately with its long-term goals. Tavio’s engagement on this issue began with an intriguing question that attempted to gauge the audience’s perspective on where they see the insurance industry in 5 years and if they believe that by then, will robots ultimately be handling insurance and its processes.
This question is pertinent and not only exclusive to the insurance industry but rather to the current era we are living in. Tavio asserts this is on account of how we are currently living in the zettabyte era whereby we have reached the level of creating, replicating, and consuming about 40 zettabytes of data in the year 2020. If you need to visualise the size of this number it looks like this:
40 000 000 000 000 000 000 000
These are the staggering facts which highlight the large scale consumption and creation of data across the world today:
- Every minute 300 hundred hours of brand-new content gets uploaded onto YouTube.
- Whatsapp users send about 100 billion messages a day.
- 60 million Instagram photos being uploaded each day.
- 250 million smartphone apps are downloaded daily
- And over 300 billion emails sent and received each workday.
Tavio further highlights how this immense data now brings about the question of storage, how do we store, secure, and use this large-scale data to better enrich our lives and decision-making processes? Tavio successfully highlighted that the discussion on storing large-scale data cannot be understood outside the adoption of cloud-based technologies. Microsoft estimates that today about 81% of all enterprises have a multi-cloud base digital strategy, that 67% of all enterprise infrastructure resides in the cloud and the average person, that’s you and me, uses 36 cloud-based services on a daily average.
This presents Tavio’s argument which asserts that if the internet was the opening act, the second act is no doubt the adoption of cloud-based technologies to store and manage data, but the third act is cloud based computational power. To make use of this computational power in order to process these zettabytes of data, one needs a strong computer, a super computer (which is already being used by tech companies such as Google to analyze data in real time). Such a computer allows one to quickly decipher trends and predict to a fair level of accuracy what the likely outcome will be based on hundreds, thousands, millions even billions of data points this is exceptionally powerful and necessary for industries that require fair amount of modelling such as space exploration, vehicle manufacture and design, or renewable technologies design and of course insurance. This is pertinent as the industry moves along the path of digitization where it will start collecting and utilizing more and more data points to drive pricing decisions on risk.
Tavio reinforces the assertion that Artificial Intelligence (AI) is the source that brings together the efficiency of the internet, cloud computing and computational magnitude. With all the large data in the cloud, Artificial intelligence is significant because with super computers, it has the necessary processing power required to decipher complicated and large data sets. AI is the umbrella term used to refer to machine learning, neural networks, deep learning, robotics and computer vision. Tavio focuses on machine learning to show how its algorithms solve problems by being able to learn from the past and predict an outcome in the future based on the learnings from data sets, including actions over time. A good simple example in the real world today, are the machine learning algorithms that seek to identify and eliminate certain types of emails because they are ‘spam’ for example. If you, and thousands of others, mark that item as spam, the algorithm, and now the system, understands that email as ‘spam’. In our world of insurance by way of example, when an email from a policy holder hits the underwriter our machine learning algorithm filters it, assigns it to the relevant department and then to the relevant user, and some systems go so far as to predict the exacting nature of the email request and then pre-populate the Policy Administration System with the expected changes straight out of the email (without human intervention).
Artificial Intelligence overall therefore as a concept seeks to take the above behaviour of predicting based on previous outcomes of data sets and expand on it. It artificially creates a new course of behaviour based on an intelligent interpretation of the data set. Tavio further explains that technology providers such as OWLS™ Software are already thinking about data consolidation and interpretation technologies as well as embedded artificial intelligence, machine learning and prediction technologies into their system, all so that the user can click the button and get an answer. So, in our above example, AI should then understand that a new email that has been received is automatically identified as spam before it even hits the user. Or in the insurance space, to not have to do the policy servicing request because as the underwriter you already know what the request is and have already actioned it.
Tavio concluded by affirming that insurers are able to better understand risks through this convergence of technology and data. While further explaining that the natural evolution of this is that pricing risk becomes more and more accurate and individualised but that we need to recognise that the evolution is going to occur and we will never be at the final stage of things. Ultimately, it will reward an individual and entity who behaves in accordance with the insurer’s expectations and will penalise the individual or entity who doesn’t.
Now, the question to ask is this right? Has the industry lost its soul or is it fully intact and just digitally enhanced?