AI Calorie-Tracking App Error Rate Found
Four popular AI-powered food apps may underestimate calories and fat in meals by about one-third, with high-fat ketogenic dishes causing the greatest discrepancies.
This paper presents a study on the accuracy of calorie counting via four widely used AI-powered food apps. The researchers tested these apps against carefully prepared meals to determine their error rates. The results show that all four apps significantly underestimated calories and fat in some meals, while others were more accurate. One notable exception is carbohydrates, which were measured consistently across all apps. High-fat ketogenic dishes appeared to cause the most trouble, with one app underestimating by as much as 70% in these cases. This finding has implications for individuals relying on these apps for meal planning and nutrition guidance.