I started with Whoop a year ago bc, as an AI entrepreneur, I was enthusiastic about Will Ahmed’s idea. In the meantime, I’ve started to partly question how cleanly it has actually been implemented.
Bc for a year now there has been an unresolved super simple, but from a data science perspective seriously relevant problem:
In the journal, there is no way to “deselect” behaviors once they’ve been filled in. There is only YES or NO. But there should also be a “no entry” (= do not consider this behavior in the calculation for that day).
When is this a problem in practice?
- You first click YES for “Took magnesium,” but then you’re no longer sure that you actually took it.
- You use the “autofill” function, but out of the 30 behaviors there are certain things you’re no longer sure about the next day.
- You realize that things like “Homeoffice” shouldn’t be checked on the weekend and want to remove them afterward, but that’s not possible.
- many other situations
What effect does this have?
- You’re forced to decide between YES or NO.
- So it can happen that you state you didn’t take magnesium even though you actually did.
- Since this surely happens to many people who then just say “whatever,” it unnecessarily distorts not only your own statistics but the overall statistics as well.
At that point, your argument of “only curated selection of behaviors” no longer convinces me if, on the other hand, such obvious little things aren’t being fixed.
I think it’s understandable that, bc of “5 min to solve”-bugs like this, I start asking myself: How well are the algorithms in the backend actually implemented, etc.? Maybe you should have some smart people review everything. As is we all know, you can sometimes be blind to your own issues.
Greetings from Germany