Your Smartphone and Machine Intelligence will Decide what You Eat – an Interview with Victor Chapela
Victor Chapela (CEO of Suggestic) shares some insight into human decision making and how machine intelligence will augment our decision making via our smartphones, enabling us to make better choices so as to reduce our risk of chronic diseases.
I recently held a video based call interview with Victor Chapela, CEO of an AI enabled smartphone health app company Suggestic. We’d agreed to chat off-the-record to start with and then to switch to recorded at the start of the intended interview. However as part of checking recording I captured some of the off-the-record material. Victor kindly gave permission that I could also publish it.
Below is part 1 of the transcript.
Victor: …and by capping expenses, the only way insurance companies can make more money is by paying more claims, because they can only keep 15% to 20% overhead on top of whatever they pay. That’s by law, so what that means — and it’s shocking, is that the insurance companies, to better serve their fiduciary duties with their investors, need to pay out more claims. And pay more claims is not necessarily aligned with making people healthier. Pharma is a similar story. They can only make money on things they can patent, on things they can protect. And a diet or a healthy lifestyle or keeping you healthier in the long term are not one of those things. So incentives are broken at the highest levels.
On the other side, healthcare providers have become more efficient by clustering up people in hospitals. But once you start investing money in hospitals your main driver is to keep the beds full. It’s like a hotel. The more rooms you have full, the more surgery rooms, the more beds that are taken, the more money you make. Again, your fiduciary duties to your shareholders – by law – require them to maximize profit — they even have in many hospitals a quota for their doctors to admit patients into the hospital.
So many high level incentives in the healthcare industry are not aligned with long term health of individuals. Businesses strive to maximize their profit.
One of the few stakeholders that cares long-term for your health is your primary care physician; many healthcare practitioners have a long-term benefit of you getting healthier because you’ll stick with them longer.
And if you go the other way, at a country level it’s also very interesting. Diabetes and chronic diseases will bankrupt almost every country in the near future if the trends continue, especially because of diabetes. But it doesn’t stop with diabetes. Diabetes is part of the cardiometabolic syndrome. Countries and governments around the World have this problem and are searching for solutions. Governments have aligned long-term incentives to help their residents achieve long-term health. Unfortunately, changing regulations and industry-wide incentive structures is very difficult and takes a long time.
Finally, and most importantly, the individual is the one who will be the most benefited. We need to empower persons to increase their own long-term health. It has to be easy and it has to be affordable for us to have a real impact.
I personally believe the best way to help out the world is by helping people be healthier. In that very key aspect, eating correctly, which is a pillar of having a controlled glucose level, having controlled blood sugar levels, and having controlled fat levels in your body can lead to the prevention of diabetes, heart disease and even cancer. It’s not about surgery. It’s not about insurance. It’s about helping everyone lead and enjoy a healthier life.
The main opportunity we found while building Suggestic is to help people to increase their willpower. If you think about it, your willpower – your self-control – is really made up of two parts: your level of motivation, and how easy it is for you to decide and act. First is motivation, which a lot of products out there are trying to help you with by trying to motivate you and teach you, but it’s not enough to keep you eating healthy. The second component is the easiness of the task. What happens is that we are inherently lazy, we are evolutionarily programmed to save energy, so we tend not to do things that require too much effort if we can possibly avoid it. The whole creation of habits is related to this. We create habits to reduce effort. And therefore changing habits requires a lot of effort and runs contrary to our nature. There’s a very good book called Thinking, Fast and Slow, by Dr. Daniel Kahneman, have you read it?
Lee: I have that book
Victor: That’s an incredible book. You know, it’s about this two ways of thinking system 1 and system 2. System 1 is pretty automatic and fast and system 2 is slow and requires a lot of effort. We manage that effort very carefully. One of the things that takes a lot of mental effort is reading and choosing from a restaurant menu. That’s system 2. You have to read it. You have to figure it out. You have to choose what you want to eat.
On top of that, If you have to filter what you can eat, based on a list of dietary rules that somebody gave to you, your system 2 is overwhelmed. Additionally, system 2 willpower is depletable, so if you tried very hard to figure out what to eat, by the time you get to dessert it will be as easy as pie -pun intended- to make a bad choice. Actually, it will be the only option. You will have depleted your willpower by then and will go directly to the not so healthy dessert.
This keeps happening again and again throughout the day. You made an effort to eat correctly and then you start craving a snack. And because you made that effort to eat correctly, now you don’t have the willpower to avoid the snack. This actually gets worse because most of us also need to be using some of our available daily willpower for work, for study or for personal relations.
What we’re doing at Suggestic is helping out by constraining choice, we present you with a reduced set of options that are specifically chosen for you. We’re supporting you in two main ways. We are suggesting what to eat at a restaurant and what to buy at the grocery store. Those two moments represent over 80% of your meal decisions.
One day you will be at a dinner party, where the food is what it is, but if you have been eating correctly, without expending all of your willpower, it’s more likely that you will make better choices at that place, like not eating the canapes or whatever they’re serving.
Now, if you extrapolate these ideas, what we are really doing is figuring out how to get people to comply with their own preferences and requirements. Our system is built in such a way that you can have a large number of data points taken into account, from your genetic code to your preferred diet, and all those things are used to filter out the real world in real time for you.
The first step is solving the willpower problem. Helping people adhere to a better and more personalized diet.
Lee: Can you backtrack there? What did you just say about willpower?
Victor: We are helping people increase their willpower by making it easy for them to make healthy choices.
Lee: You’re taking the load off their system 2 willpower store. You’re stopping them from depleting their willpower.
Victor: Yes. That helps them adhere to their diet or combination of diets that is best for them, their personalized diet. In terms of personalization, what we’re doing is we have built a system that is able to learn over time from you and from people like you.
Lee: It sounds like a recommendation engine, except for nutrition, right? People like you also like this movie. People like you have better blood glucose levels when they ate X.
Victor: Yes. Let me think about that to see if there’s any difference. Generally speaking, yes, it’s a recommendation engine, it learns in a similar way. We use machine learning for that. It “understands” what are those things that you should be keeping track of and then it actually keeps track of those things for you. That’s where we are right now. Because we are a startup, you normally have to cut things into the smallest viable product, the minimum viable product. And you could summarize it as a nutritional recommendation coach.
Lee: You have to make a component or atomize what you’re doing in order to extract maximum value and maximum leverage.
Victor: Exactly. So we are now focused on what I’ve been saying. I believe that within the next five years we will be allowing our technology to make most of our food and health choices for us. We’ll have to start using it and see how it works, in the same way we experienced getting out of traffic consistently with Waze. And every time we second guessed Waze, we actually got into a worse spot. We eventually learned to trust that technology and the same will happen for health.