I’ve used Google Health for 2 months. Here’s what I love (and what I hate)

It’s been greater than two months since Google launched the Fitbit Air. Whereas the $99 health tracker grabbed many of the consideration, one of many largest adjustments that arrived alongside it was Google replacing the Fitbit app with Google Health.

On the time, Google stated Well being would turn out to be a unified platform for all of its well being gadgets, together with Fitbit trackers and Pixel Watches. AI was a significant focus with options just like the Well being Coach, and the corporate additionally promised assist for knowledge syncing with a number of well being platforms, including Apple Health.

The preliminary reception wasn’t significantly optimistic. The truth is, our former editor, Rita, wrote on the time that she “hated” the Google Health app. Since then, although, Google has rolled out a number of updates for each the Fitbit Air and Google Well being to enhance the expertise and squash bugs.

However in any case these updates, has Google really fastened the most important issues with Google Well being? I’ve been utilizing Google Well being (virtually) since its rebrand, and right here’s the place I’ve landed.

I’ve by no means been the kind of one who checks their well being/health app very first thing within the morning and spends 10 minutes analyzing graphs, numbers, and restoration scores. Certain, I like sporting screenless trackers just like the Fitbit Air and WHOOP MG whereas I sleep, however manually digging by means of all that knowledge day-after-day simply isn’t one thing for me.

One in every of my largest complaints with conventional well being apps is that they throw numbers at you the second you open them. Google Well being takes a special strategy. As an alternative of anticipating me to interpret each metric myself, it offers me AI-generated summaries that specify what really occurred and whether or not I ought to care. And that, for me, is much extra helpful than making an attempt to investigate the numbers myself.

Google has additionally fastened some points for the reason that launch of the Well being app. For instance, probably the most widespread criticisms early on was that you just couldn’t customise the house display structure. Fortunately, latest updates have addressed that, bringing the ability to change the order of pinned tiles.

Unpopular opinion, however I additionally actually just like the little AI summaries and notifications Google sends all through the day. They’re fast sufficient to look at in a number of seconds and inform me all the things I have to know with out forcing me to actively preserve observe of my well being metrics.

However the place I’ve discovered Google Well being AI genuinely helpful, although, is when asking questions on my very own knowledge. For instance, I lately requested it to check my sleep over the previous month on days after I exercised versus days after I didn’t.

The Well being Coach famous that my sleep high quality was barely higher on days I skipped a exercise (despite the fact that the distinction was very minor). It’s an fascinating stat, and it’s not the form of stat I’d’ve seen alone.

That’s exactly why I like Google’s strategy. I do know loads of different smartwatches and health trackers can do a lot the identical, particularly when you put on them lengthy sufficient, and most now have some type of built-in AI. However since Google presents this at no further value when you’re already subscribed to Google AI Professional (which I’m already paying for), it’s an amazing value-add in my view.

That stated, the AI is much from excellent.

Overreliance on AI is the place Google Well being begins to crumble for me. From what I’ve seen, Google Well being remembers all the things you inform it. Not “all the things” in a free sense — I imply all the things. Moreover, it appears to imagine that something you point out as soon as is now a everlasting a part of your routine.

Some time again, I requested Well being Coach whether or not it was nice to take magnesium and liposomal glutathione collectively earlier than mattress. It was only a easy, one-off query. Now, virtually each night, it asks whether or not I’m sticking to that routine and strikes up conversations as if these dietary supplements have turn out to be a everlasting fixture in my nightly schedule.

What’s fascinating is that this doesn’t appear to be an remoted situation. I confirmed the identical habits with our Managing Editor, Joe Maring, and located a number of Reddit users complaining about the very same factor. The AI seems to recollect virtually all the things you inform it, even when it in all probability shouldn’t.

Because of this, I now suppose twice earlier than asking health-related questions. I don’t know whether or not the AI Well being Coach will deal with an off-the-cuff query as a long-term behavior, which is counterintuitive to the characteristic’s supposed function. Fairly satirically, I’ve began asking these inquiries to Gemini as a substitute, despite the fact that Google Well being is meant to be the extra specialised device.

What makes issues worse is that there’s no actual strategy to audit what Google Well being really is aware of about you. There’s no dashboard displaying what info it remembers. There’s principally no strategy to inform Google Well being to neglect one thing.

I consider well being knowledge is a few of the most private info we’ve. If Google desires AI to construct long-term context round my routines or dietary supplements, I feel it also needs to give me way more management over that info. On the very least, there ought to be a easy possibility to inform the AI to neglect one thing or cease asking about it.

Other than these points, there’s additionally reliability. The Well being Coach, in the long run, is an AI, and it does hallucinate each now and again. I’ve had a number of situations the place a notification confirmed me incorrect well being knowledge, just for it to right itself after I opened the app and it force-refreshed. I’ve additionally seen loads of similar reports on Reddit, so it doesn’t look like an remoted bug.

For me, these are a few of the drawbacks that Google nonetheless wants to repair. The AI summaries appear fairly helpful, but when I’m anticipated to belief an AI with my well being knowledge, it additionally must be clear about what it remembers and constantly correct about what it tells me.

I preserve coming again to the identical feeling with Google Well being: it’s good. The truth is, it’s genuinely good in a variety of locations. After which it does one thing that utterly undercuts all of that goodwill. The easiest way I can describe it is sort of a meal you actually loved, just for the aftertaste to bitter the entire expertise afterward.

There are many issues the app will get proper. However a product that remembers all the things you say, hallucinates right here and there, and offers you no strategy to see or handle what it’s remembering doesn’t really feel like a completed product.

So, two months later, I’d positively say Google Well being has improved, however I’m nonetheless not satisfied Google has discovered precisely what it desires the app to be. It’s caught someplace between a conventional well being dashboard and an AI well being companion, and I feel it’s excessive time Google absolutely dedicated to at least one path.