This one feels different. After months of building, testing, changing things, and occasionally breaking them, OS Health is about to leave our hands and go to its first outside users. October 4th.
OS Health opens its first outside-user beta on October 4th. This is the product we've been building toward: one clear movement decision for today, informed by the context that actually matters. Now we get to see how it holds up in real life.
@kirbywinfield The one that hurts the most is pro athletes. Even though I never would have been able to compete at their level, the visual removes all possibility.
OS Health enters its first outside user beta on October 4! We built it around one question: What should I do for my body today? The goal isn’t more health data. It’s turning the context that matters into one better daily movement decision. Sign up link in the first comment.
October 4. That’s the date we’re opening the OS Health beta to our first users. We’ve spent months building, testing, fixing and refining the product. Now we get to learn what matters most: what happens when real people start using it in their daily lives. Sign up in comments.
Most people don't have a health data problem. They have a decision problem. Every day we're flooded with wearables, sleep scores, recovery metrics, bloodwork, AI, and endless advice. Yet we still ask: "What's the best thing I should do today?" That's why I'm building OS Health.
Initially couldn’t make sense of this Energy per Habit correlation. “Networking on LinkedIn” is such an objectively miserable task. But, apparently, if I do it, that’s proof I was in a terrific mood 😂
Anyone else got one of these reverse causation habits?
The hardest part of building an AI coach isn't generating recommendations. It's deciding when *not* to change them. Consistency builds trust. Adaption should happen only when the user's context genuinely changes. That's the standard we're building toward.
One thing we've learned while building OS Health: The best recommendation engine isn't the one with the most inputs. It's the one that knows which inputs actually matter today. More context isn't always better. Relevant context is.
Every feature we remove forces us to answer a harder question: Does this actually help someone make a better decision today? If the answer is no, it doesn't belong in the MVP. Building AI isn't about adding more. It's about deciding what matters enough to keep.
One lesson from customer interviews: People rarely ask for more health data. They ask questions like:
"Can I push hard today?"
"Should I rest?"
"Am I doing enough?"
Those aren't data problems. They're decision problems.
A good health coach doesn't ask you to ignore real life. Bad sleep, travel, stress, soreness, and time constraints all matter. We're building OS Health to make recommendations that fit today's reality, not an ideal week that never exists.