Instructor: David Schuff, Section 003

AI Aiding Diets

food photo

Artificial Intelligence is becoming something we can almost no longer avoid or live without. As humans, something else we cannot live without is food, because it fuels our bodies. Today, there is a concern with the negative effects “bad” food can have on us, most importantly, obesity. More than one in three adults today are what is consider obese, according to their BMI. With this being an ongoing epidemic in the United States, several startup companies hope to help diets by using AI. Edamam is a startup focused on helping you get easier and quicker nutritional information on the ingredients in a recipe. Their database of over 50,000 products allows them to use AI to do all the necessary nutritional information in the background. By creating an easy to use software where users can easily track nutritional information, more people will be more conscious of their eating habits. Passio follows a similar goal but uses AI to help users track food based on image recognition. Thanks to recent advancement in AI and image recognition, the error rate is around 3%, where as a human brain error rate is 5%. The last company to take on AI in the food industry is Habit. They use not only AI but also genetics, from a sample kit sent out, to help users find personalize diets to their needs. Habit goes one step further to offer prepared meals based on the genetic data found from the sample kit. With so many unique opportunities for AI to be implemented in the food industry, do you think AI can combat obesity?


Artificial Intelligence in the Food Industry: How to empower consumer decisionmaking


5 Responses to AI Aiding Diets

  • Hi Dante — this was an interesting post, nutrition doesn’t initially come to mind when thinking about Artificial Intelligence. I think one of the most important aspects of the apps you described is the educational potential. Making nutritional information more easily accessible to people will help them make healthier, more informed options about the food they’re consuming. These apps could even go a step further by explaining the nutritional value of the various components of the meal (ex: this food is rich in Vitamin A which is good for…) instead of just listing the calorie and fat content. Having a simple and convenient app (like the one that lets users track food based on image recognition) will integrate well with people in their daily lives and will encourage them to use the apps more frequently. I think Artificial intelligence in the health and nutrition space has the potential to make a significant impact on obesity, diabetes, and other diseases, especially if the recommendations are customized to that person’s unique medical history and desired goals/diet plan. Thanks for sharing!

  • The article you have provided is very interesting, especially because I’ve never really noticed the useful connection between Artificial Intelligence and nutrition. Based on the information I have obtained from the article you have provided and some additional research, in reference to your question, I think AI does have the potential to semi-combat obesity, but not necessarily to reverse the negative effects of it, nor the impact it has already made on society. The beauty of the applications you have described rests in their ability to empower consumer decision making. While these apps are designed to encourage users to eat healthier by informing them, consumers still must take the initiative to regularly use the apps and follow the nutritional suggestions they give to truly have a healthier diet. Because of this, I see potential for such forms of AI to be successful in combating the effects of childhood obesity in particular, because if children learn at a young age to regularly track their nutrition, it can have many positive effects on their overall health. The problem is, while it does make access to this nutritional information easier and more convenient, that does not necessarily mean that people will regularly use these apps. In order for this type of technology to truly combat obesity in the fullest, it must break down that barrier to ensure regular use by consumers.

  • Hi Dante – I really like the topic of your article, because I never would have thought that nutrition could be a use case for artificial intelligence. I think this use case has incredible potential in curbing obesity, but also for those who are interested in pursuing a more sustainable diet such as vegetarian or vegan but worried about not getting a balance of nutrients (i.e too many carbs, not enough protein, etc).

    I do take exception with the author applying the error rate of machine learning API’s dropping to 3% for the image recognition application. I believe that number is derived from controlled lab tests where researchers train a machine learning program on pictures of dogs and after a while it is able to distinguish between a dog and a cat. I find it hard to believe that an app would have that narrow of an error rate with identifying different ingredients in an uncontrolled environment with foods that can be prepared in a multitude of different ways. As of Summer 2017, an MIT learning algorithm that identifies ingredients had an error rate of 35%. I think this is incredibly promising, but we are far off from hitting that 3% target.


  • I think this is a really interesting area where I did not expect to see AI. Even though AI can help users gain more information about the nutritional value of the food they are eating, I do not think it can combat obesity. It can help educate people on which foods are healthy and which food they should avoid; however, in the end AI cannot physically stop someone from eating foods that are not healthy. The majority of people know that processed and fried foods are not good for them, but that does not stop them from eating them. I do think AI can be very helpful in helping someone decide which foods to include in a meal or recipe based off the nutritional information. Although AI can help educate, I do not think it can help combat obesity.

  • This is a really interesting application of AI. While I think this is a great idea and a product many people would use, I do not think it will have a huge impact on overall health in our country (or world). I think that, as with most health innovations, these products are used by people who are already very health conscious, and are just looking for the next thing to help them be healthier. The other main group of users will likely be people who jump on board for a bit, but then, for the same reasons other dieting and wellness routines failed, will quit. Overall, I think this is a really cool product, but one that is unlikely to have a major impact on widespread health.

Leave a Reply

Your email address will not be published. Required fields are marked *

Where and when do we meet?
Alter Hall 603
12:30-1:50 Tuesdays and Thursdays
Office Hours
Tuesdays and Thursdays, 2:00-3:00 PM
Subscribe to Course Site via Email

Enter your email address to subscribe to this course site and receive notifications of new posts by email.

Join 13 other subscribers