Multimodal models are useful for lots of things! They can accomplish a range a tasks from zero-shot image classification to helping perform Retrieval-Augmented Generation on images. Like many generative model, I find the utility comes not necessarily from outperforming a human, but from scaling a task that a human wouldn't want to do (or won't do cheaply).
The set of techniques for retrieval is immature, but it's important to note that just relying on model context or few-shot prompting has many drawbacks. Perhaps the most important is that retrieval as a task should not rely on generative outputs.
To me, Garmin is the only company here that does a good job with the "hardcore fitness" market, but why does that matter in the first place? The better market to compete in is the "fitness amateur" market, which I see as being many times the former in size. Fitbit competes well here, and Google has no offering to speak of. Seems like a successful diversification move for Alphabet's portfolio. I especially like what Google's software can do to improve FitBit's offering by leveraging AI.
I think the point you're missing is everyone levels up. Even amateur fitness users want a better fitness wearable. Which is why Garmin and Apple have the lion's share of the market. Fitbit is still seen as a step tracker level device, even if they do more. Garmin wearables are laser focused to the workout market and most buyers seem to lust over the idea of needing a more capable fitness device. Garmin and Apple both cover "fitness amateur" in spades but do an upsell far better.
With regard to AI, I'm not trying to be flippant but, those users don't care. No amateur fitness users would even understand how that would apply and likely be less to care than selling predictive analysis to those working out who legitimately track vitals and are trying to shave seconds.
Apple Watch battery lasts a day at most. My Fitbit can go without a charge for 6 days. Fitbit offers sleep tracking.. Apple Watch doesn’t natively, and even if it did, you still have to charge it at night. Fitbits are more minimalistic. They’re so nice. I don’t want a Garmin wearable. A lot of people who use Fitbits don’t use them for the fitness aspect. We use them to avoid the sedentary lifestyle by viewing our steps, and occasionally going on a run. I hope Google doesn’t ruin Fitbit. But judging by their past moves, my Charge 3 will be useless in a matter of 2-3 years
I much prefer my Apple Watch over my shitty fitbit to be honest. Charging isn’t an issue. I only charge it in the morning when I’m getting ready and that’s enough to last me the whole day and through the night.
All that as well as the great GPS capability. I have the Fitbit Ionic and start the Walk exercise when bushwalking ( hiking). It produces very accurate track logs that I can use for mapping when I return. It does consume more battery but far less than high rate logging on the phone
This, I want the basics, not another phone on my wrist. The charge 3 lasts me 10 days and charges in about 15 minutes. I don't want another device to charge daily.
This and the casual smartwatch market where FitBit's versa already sells great (even though it's not great at all). Google's offerings here are sparse, littered with confusing options and half-assed design. A cohesive pixel/Fitbit product can help create a proper Anti-Apple Watch product.
> "fitness amateur" market ... Fitbit competes well here
I'm not so sure about that.
For simple activity monitoring by way of step counting, movement time, sleep time, and similar simple metrics, there is a lot of competition. This isn't just from the extremely cheap (but probably not very good) options, there are several devices out there at about half the price of FitBit's cheapest that seem to do the job just as well (caveat: I'm basing this mainly on anecdotal evidence from friends/family and online).
For very little more than their cheapest watch & step-counter you can get a TomTom sports watch which has built-in GPS for accurate run/cycle/other tracking, breadcrumb mapping, and so on. The price difference for adding a wrist-based heard-rate monitor is about the same in both ranges.
Moving away from the casual fitness market towards people like me[†], their only GPS capable device[‡] costs more than Garmin's 235 which is a more capable device, more than twice the price of the aforementioned TomTom units, in fact you can usually get a Fenix 3 for the same price as the Ioinc, and there are a couple of other well regarded competitors with similar feature sets at that sort of price level too.
Their key advantage is name recognition, at the casual end of the market at least, though that doesn't necessarily help. People often call cheap-n-dirty activity trackers "cheap fitbits" rather than an activity tracker, watch, or other name including the products official name, but they still buy them instead of the actual fitbit. They did in the past seem to have that part of the market cornered, but seem to have let it slip considerably in recent years.
(NOTE: I'm in the UK. Relative pricing of manufacturers/models may differ in different markets.)
[†] I'm a recreational runner, far from the top of any particular class though in recent years I've knocked of a couple road marathons and multi-day trail challenges so I consider myself to be good at putting one foot in front of the other without tripping over either!
[‡] I generally discount phone-based GPS tracking by wrist-mounted devices due to the battery drain on the phone, and I never found it terribly reliable though that may have improved since
