Setting aside how many data points you’ll get after testing, you’ll actually see them on Superpower’s website, on the app, and through the company’s AI service. The website is well designed and easy to navigate. Once your results are available, you’ll see general results, like your “superpower score” and your “biological age.” You’ll then be able to dive into the results individually and navigate different tabs for things like seeing all the data together, taking a look at your “protocol,” and purchasing recommended supplements and other products.
The “protocol” is essentially Superpower’s plan for your health, based on the goals you entered into the platform when you started and the test results returned from the lab. I didn’t really find it to be the most useful tool, but I was much more interested in the lab data anyway. Additionally, there is a strong connection between the protocol and purchasing additional products from Superpower, which put me off a bit. If you want concrete steps, you may find the protocol useful, and not all of them revolve around selling your products. For example, my vitamin D levels were lower than normal, so my protocol involved getting outside, which is effectively free.
Much of what Superpower offers depends on AI, for better or worse. I found it useful, but that’s probably because the data derived from Superpower was unhelpful in the first place. I was told I had an elevated “bilirubin-to-albumin ratio,” a measurement based on two lab results considered to be in the “normal” range. The ratio between the two was, however, “high”. This, of course, was slightly worrying. Most people will see something like this and wonder what a high ratio could mean for their health.
So I asked Superpower’s AI tool, and was told, in other words, that the ratio was technically high and that I didn’t need to worry about it. In her explanation, the AI told me that the ratio “is not a clinically significant elevation – it is a mathematical artifact due to the presence of bilirubin in the upper optimal range rather than the lower optimal range.”
One of two things is true here: either the ratio needs to be adjusted so that customers aren’t unnecessarily alarmed when they get their results, or the AI is wrong. Either way, it’s a problem – and it’s Superpower’s responsibility to fix it. Of course, this is ignoring the very fact that not knowing whether AI is reliable is another problem. AI services may simply not yet be reliable enough for health-related questions, especially when they provide conflicting information about whether a test result is “elevated” or not “a clinically significant elevation.”
