How Can Artificial Intelligence and Artificial Intelligence Help Achieve Precision Nutrition?

Precision nutrition is about better adapting diets and dietary recommendations to different people because, as I have written before, one size definitely does not fit all. Forbes. To determine the best diet for someone, all you need to do is understand that person’s genetics, physiology, microbiome, body type, eating behaviors, stress, social influences, food environment, health conditions, and many other things. These are things that affect nutrition and health. And you keep track of how all of these things might interact with each other and change over time. No problem, right?

Not exactly. It can be really complicated to keep track of and sort through all these different things that are happening at different levels, in different ways, for different people, at different times and circumstances. That’s a lot of “differences.” But these days, when you have something very complex that you need to solve, you have a potential friend in artificial intelligence, that is, artificial intelligence.

One of the big challenges is that science has not yet figured out how all of these different factors may interact to influence how a person’s diet may affect their health. Of course, studies to date have produced insights into how each of these factors may act individually and for certain types of people. But combining these insights is a different matter, and many gaps remain.

This is because no single traditional real-world laboratory, clinical, or epidemiological study alone can account for, measure, and track all the different factors and outcomes for all types of people. No matter how hard you try to design the “perfect” study, you will undoubtedly fail to include all kinds of people and measure every relevant factor and outcome.

Also, even if you design a “perfect” study, you still have to wait a very, very long time to get all the results needed. It can take years or even decades for the effects of nutrition to manifest as different health conditions. Anyone who ate like a garbage disposal and considered ketchup a vegetable throughout their 20s will tell you that.

So if you really want to figure out how to do precision nutrition, you have to somehow combine data from many different studies and fill in the gaps. You also want to find ways to extend the results of a particular study to people who did not participate in that study and to conditions that were not covered. This can all be too complicated for any person, even a team of unassisted people.

Enter the AI ​​and sing Randy Newman’s “You’ve Got A Friend In Me.” These types of computer-aided techniques can keep track of many different things, combining different data sets in different ways and figuring out how they fit together. These techniques can also determine how the results of a single nutritional study can be applied to different conditions and situations, thereby increasing the utility and value of that study. And various artificial intelligence techniques can do this much faster than humans. These are just some of the ways AI can help achieve precision nutrition.

To understand how artificial intelligence can do these, you must first know what artificial intelligence is. These days, AI has become such a sexy term that people can use it without even knowing what the term means, like “Hey, can you do that AI?” They might say things like: Artificial intelligence is an umbrella term that basically covers any computer-assisted technique that can replicate something the human brain would normally do beyond simply following step-by-step instructions. This means that an AI approach can evaluate situations or make decisions on its own. There are already many different types of AI approaches, methods, and tools available, and the list continues to grow every year.

One way to classify AI techniques is along the continuum of how these techniques are designed and operated. At one end are purely data-driven AI approaches. These are “top-down” techniques that start with a body of data and try to understand patterns, trends, and relationships from that data. This is a bit like how a statistician might analyze a data set. But the AI ​​algorithm can do this much faster and perform many different analyzes across multiple data sets at the same time.

Let’s look at a theoretical example. A data-driven AI approach can analyze various data sets, slice the data in different ways, and find that people who eat a particular food item tend to live longer. Let’s call this food item “The Best Food Ever”, this is a completely fictional term and does not bear any proper name. The AI ​​algorithm can then associate the Best Food Ever with greater longevity, but it can’t explain why this relationship actually exists. The Best Food Ever can’t really distinguish between whether consumption actually has a beneficial nutritional effect and whether some sort of coincidence has occurred. Maybe those who tend to eat the Best Food Ever may also tend to eat another food item not included in the data set that actually works. Or maybe people with less stress are more likely to have the time and money to eat the Best Meal Ever. The Best Meal Ever may actually be a red herring, meaning something misleading or distracting, rather than something made from fish.

At the other end of the spectrum are mechanistic or explainable AI approaches. These AI methods attempt to recreate what actually happened, bottom-up, by recreating the actual mechanisms behind a process or decision. They are considered explainable because you know why a result was generated.

This is similar to what scientists do when they design experiments to test what might happen in the laboratory. The difference is that the AI ​​algorithm or model is not limited to a physical laboratory and can serve as a “virtual laboratory” representing an individual, a group of people, a population, or an entire geographic area. The model can then run experiments on the “security” of a computer in ways that would be too complex, too costly, too time-consuming, too impractical, or even too dangerous to do in real life. The mechanical AI tool can then use the results of these experiments to determine recommendations, just like a human would run thought experiments in their head before taking action.

For example, a mechanistic AI approach might be to represent different reasons why a person might choose to eat the Best Meal Ever. It can also represent the different nutrients in The Best Food Ever, how they are broken down in the body, how those nutrients then affect different organs, and how this ultimately affects longevity. This AI model can then look at what would happen over time if different people ate the Best Meal Ever and decide who would benefit from eating the Best Meal Ever and how.

These different AI techniques across the spectrum can be interoperable and integrated. A purely data-driven approach could suggest relationships (e.g., take a closer look at The Best Food Ever) that could guide the creation of more mechanistic AI approaches (e.g., let’s find out what the Best Food Ever actually does to the body). Similarly, a mechanistic AI approach can help identify where data-driven approaches are needed. Let’s say you’re trying to represent the mechanisms by which The Best Food Ever affects the microbiome, but you can’t unravel them because there are no traditional studies that clearly reveal relationships, patterns, and trends. Therefore, it may be useful for data-driven AI approaches to sift through this microbiome data.

Of course, you shouldn’t automatically trust anything the AI ​​tells you. Just as a poorly designed clinical trial or observational study can lead to misleading results, a poorly designed AI approach can also lead to misleading results. That’s why you need to know what’s underneath the AI ​​approach and understand its relative strengths and weaknesses. At the same time, no AI approach (like any real-world study) will be perfect. Don’t let the perfect be the enemy of the good, and don’t let the flaws of an AI approach prevent you from using it for risk aversion.

Integrating more artificial intelligence and other computer-aided approaches to make more precise recommendations is not entirely new and has been done in other fields as well. Fields such as meteorology, finance and aeronautical engineering have long used computer-aided techniques to combine and analyze complex data from disparate sources and create more accurate forecasts and forecasts.

So while AI probably won’t challenge some of the already established nutritional understandings, such as the value of eating fruits and vegetables, the nutrition field is ripe for change. There are so many people out there who are super duper blah blah blah claiming that this diet works for everyone. But not everyone is the same and has the same conditions, and that’s exactly the problem. Achieving more precise nutrition is not easy. but you have a potential friend in artificial intelligence. But like every potential friend, you should treat him right and know what he can and cannot do.