The video argues that MyFitnessPal’s expensive acquisition of Cal AI is misguided, as AI-powered calorie estimation from photos can now be easily and cheaply replicated using modern AI tools. The creator suggests that as AI agents become more capable, the value of standalone apps like Cal AI will diminish, making technical features less of a competitive advantage and shifting the focus to distribution and community.
The video discusses MyFitnessPal’s recent acquisition of Cal AI, a startup that uses AI to estimate the calories in food from photos. The creator speculates that the acquisition likely cost MyFitnessPal tens of millions of dollars, possibly even in the low nine figures. He argues that this move demonstrates MyFitnessPal’s lack of understanding about the current direction of software development, particularly as AI agents become increasingly capable and accessible. The video’s central claim is that the core functionality of Cal AI—using AI to analyze food photos and estimate nutritional content—can now be replicated extremely quickly and cheaply by anyone with access to modern AI tools.
To illustrate this point, the creator describes how he was able to build a similar calorie-tracking feature in under 20 minutes using OpenAI’s tools and Gemini Vision. He demonstrates the process, showing how his agent can identify food items from a photo, estimate their nutritional content, and log the data securely. He notes that this kind of functionality, once considered sophisticated, is now trivial to implement thanks to advances in AI. The implication is that the technical moat that once protected apps like Cal AI is rapidly disappearing.
The video also reviews the history of MyFitnessPal, noting its 2015 acquisition by Under Armour for $475 million and subsequent sale to a private equity firm at a loss. The creator criticizes both Under Armour and the private equity firm for lacking the technical expertise to run a software company effectively. He suggests that MyFitnessPal’s acquisition of Cal AI is another example of a legacy company trying to buy innovation rather than build it, and that this strategy is increasingly risky as software becomes easier to replicate.
A key argument in the video is that while Cal AI reportedly had $30 million in annual recurring revenue (ARR), this revenue is not durable. As AI agents become more capable and customizable, users will have less incentive to pay for standalone apps when they can get the same functionality integrated into their personal AI assistants. The creator points out that the cost of using Cal AI ($2.50 per month) is already low, but even that becomes unnecessary when users can build or access similar features through their existing AI subscriptions.
In conclusion, the creator predicts that the software industry is undergoing a major shift, with AI agents poised to disrupt traditional SaaS business models. He believes that acquisitions like MyFitnessPal’s purchase of Cal AI will become increasingly common, as legacy companies try to keep up with nimble startups. However, he warns that software is no longer a strong competitive moat, and that the real differentiator will be distribution and community, not technical features. He ends by congratulating the Cal AI founders for their successful exit, but expresses skepticism about the long-term prospects of such acquisitions in a rapidly changing software landscape.