We are now in an era where data is king. Companies that own and manage data far exceed the market cap of traditional services or product-based companies. Many of those companies have weaponized data to create a competitive advantage that allows them to leap far ahead of any competition. A major premise of the data economy is that users don’t know what to do with their data and so companies don’t give access to data they collect on individuals. This causes a lack of trust between providers and their customers that have been the catalyst for the new privacy laws and regulations being enacted around the world that are providing more rights to users.

One way to bridge that trust is to create transparent AI that not only helps customers interpret their data but also explains that interpretation to them. This goes beyond the visualization of the data. The big push towards visualizations or creating fancy graphs is based on the idea that the person viewing that information has the required background to understand it. Not many people are data scientists, and many people are not experts in the field that the data is covering. For example, my father was a journalist, not a fitness guru. When he tries to understand his wearable information, he is often confused by it. After all, the goal of 10,000 steps was not drawn from scientific research but rather from a watch maker in Japan who tried to capitalize on the 1964 Tokyo Olympic games by creating a pedometer called a 10,000 steps meter. He simply wanted to know how 10,000 steps would improve his goal of staying mobile in his 80s. Even in the realm of elite sports, many athletes tell us that they don’t want to see the data, they want us to tell them what to do about it.

If the AI is expected to provide a rationale for its conclusions and help you decide what to do then it needs to understand what works for you not just tell you what would work for the general population. One Olympic skier told us, tell me what works for me not what worked for someone else. A long distance runner in Colorado reminded us there was no one way to run, he said I am 6’4 and my girlfriend is 5’6. I can’t run like her. This is one of the main reasons why people struggle with knowing what to do with data. A one size fits all answer is the easy answer but it’s rarely the right answer. Major League Baseball’s use of analytics show us that a tremendous amount of data can be collected on just one player and using that data individual recommendations can be generated. Wearables and apps are collecting a tremendous amount of data everyday on their users. This data can be used to create AI models built to take advantage of the data from single users. This will allow us to predict what will work best for you.

Personalization is the heart of the Omipar product by Amplio. It’s an AI that integrates biometric and psychometric data into a single model of the individual based on their personal needs. It does not just compare its model to everyone else. Instead, it spends time learning who its user is and how best to help them perform. Not only does the model interpret the collected data so that our users can understand their current state but it also predicts what will happen if they took certain actions. This simulation of future outcomes allows our product to suggest the best course of actions to achieve our users’ goals. It is also important to adapt the model so that it changes as our users change in order to keep challenging them to perform better.

Learn more here.