So a bit different from the other comments. We're a wearable company which of course pairs with an app.
We built iOS first because that's what most of our users have, though the team is mostly on Android.
We built our logic in Rust, so that compiles cross platform. It's really just the UI layer that is per platform.
Though I had initially written our first demo in React-Native, the team wanted to go with native UI.
Lately we've been talking about moving the timeline for our Android app up because we can get AI to help using the iOS app as a foundation.
I haven't seen anything really new and compelling in iOS/Android land. The promise of right once run anywhere continues to be a promise, and the bluetooth libraries across different devices (particularly in Android land) has me concerned.
We define our API into the Rust via protobufs[0] and some codegen.
The wrapper around these is very lean, FFI-based[1], it's abstracted to a library so Android devs never need to know there's rust/codegen involved.
We then define a light Kotlin-based wrapper around the protobuf-generated API[2]. This is a candidate to move to Kotlin multiplatform in the near future (mostly for test speed improvements).
> the bluetooth libraries across different devices (particularly in Android land) has me concerned
At Google IO 2026 there little discussion of core improvements to Android as a platform. It was mostly AI related, nothing about Gabeldorsche etc, which I find more interesting.
> and the bluetooth libraries across different devices (particularly in Android land) has me concerned.
I work on the Android side for a company that develops an external bluetooth device (not a wearable, but similar enough) and the bluetooth library situation has certainly been an issue to be concerned about.
Using the default Android bluetooth SDK "raw" is fraught with potential problems, at the very least you need some sort of added message-queue-like system built on top since it does none of this for you. You also have to contend with a lot of strange device specific issues (of the type "works fine on 99% of Android devices, this one Samsung device has a driver quirk we need to work around", etc).
Things looked pretty good with the Jetpack androidx.bluetooth library as a higher level abstraction, but Google suddenly deprecated it and stopped working on it before it left alpha.
After lots of investigations and iterations I ended up using Nordic Semiconductor's 2.0 Kotlin BLE Library, which is technically also still in alpha release, but is much more stable than that designation suggests.
After having messed with all of the options, I highly recommend Nordic's library as a solid basis for using Bluetooth in modern Android/Kotlin, but in your specific situation I'm not sure if bridging it to Rust would be less or more painful than having an LLM generate a Rust wrapper and message queue system over the raw android.bluetooth library.
why was your team against react native? (for mobile app right that goes along with the wearable right?)
you can ship the actual UI stuff in seconds.
you can write your own plugins for anything native and just call it. we have a app that uses many native SDK calls which we write plugins for and call.
not to mention that once the native work is done you can deploy ui level changes and new features, bug fixes to users over OTA.
i’m just an outsider so i know my lane. but
whatever you do, do not allow bike shedding and ultra opinionated tech nerds to derail company vision timelines.
An upstream metric that can be measured daily is sleep slow-wave activity or delta power.
This is the synchronous firing of neurons which define restorative deep sleep, and one of the primary patterns which we describe as the Neural Function of Sleep.
This Neural Function of Sleep naturally declines with age, but more importantly, through stimulation we can enhance it which research is showing improves immune function, increases HRV, and more.
So while the original post discusses markers they expect to measure every 3 months, our work at https://affectablesleep.com measures the Neural Function of Sleep daily, but not to give you a score, but to actively support how well the brain sleeps, not how long.
Though there are over 50 published peer-reviewed papers in these techniques, I'm curious to see if we begin stimulation in our 30s, prior to the decline in sleep, do we slow the rate of decline as we age, as well as supporting daily function.
That's a strained claim to make. From a quick search assist response: "As of 2026, there are approximately 3.5 billion active Android users and around 1.5 billion active iOS users globally"
How are you serving the larger customer base by being iphone only?
I think they will soon be re-writing this article with the addition of "and rise again".
I work in neurotech/sleeptech and one of the biggest challenges I see in our industry is, if not pure snake-oil, the over-hype of "backed by research".
People have accused us of being snake-oil as well, and I get why they might think that if you haven't read or don't understand the science.
I've seen products that claim to be Vagus Nerve stimulators that are worn on the wrist, nowhere near the vagus nerve. Products that claim to mimic the "magnetic frequency of hormones".
We've also got the current "It's got electrolytes" craze which is kinda crazy that we've truly replicated idiocracy.
For those curious, I'm the co-founder of affectablesleep.com
The modern emphasis on electrolytes is directionally correct but ultimately silly. Most electrolyte satchets and drinks are just overpriced salt water with artificial flavoring and sweeteners. The majority of the Western world actually fails to meet their recommended intake of potassium and magnesium. The former is rare in beverages because of its bitter taste and possible drug interactions. Magnesium is also used somewhat sparingly because certain forms create GI distress. This is very unfortunate since both of these minerals are essential to overall health and of great benefit to the cardiovascular and circulatory systems when taken in recommended amounts.
I dump Lite Salt (potassium and sodium chloride in equal amounts), Stevia, and powdered grapefruit into a bottle and shake to make my own electrolyte drink. It's dirt cheap and tastes like flat Fresca.
My quality of life improved by a significant amount when I started supplementing magnesium. Better sleep and a bunch of other things. It's actually scary how much of a difference something so simple (and cheap) made.
Placebo effect may exist for someone who claims to sleep better with it, but there are effects that are definitely not placebo.
In the water that I drink in the morning I dissolve small quantities of powders of magnesium bisglycinate and potassium citrate.
Before starting to do this, after days with more intense physical effort, I frequently had nocturnal leg cramps. Since I began taking this regularly, I never had leg cramps again.
Understandable thing to test for but, in my experience magnesium has been legitimate for me. I've found it improved my ability to think personally and depression. The 'dosing effects' are the thing that convinced me it is a real effect. If I ran out of it and was say, waiting for it coming in the mail I could go for it for some days without noticing a difference. That level of 'some endurance, don't need it daily' seemed to suggest some real pharmacology behidn it.
I also say for me because if you already have good levels of magnesium in your diet, it will have nothing to improve.
gwern once ran a rigorous N=1 self-experiment on magnesium with self-blinding and week-long blocks, and concluded that it was probably helpful to him: https://gwern.net/nootropic/magnesium#conclusion
You are welcome to review his methodology and see if it still seems like the placebo effect.
Based on his ranking of MP (mood/productivity), he concluded that it helped for the first three weeks and then got worse because he was overdosing on the magnesium. So not much there is going to apply to the general case.
I believe that it is undesirable to ingest excessive amounts of chlorides.
Normally, the amount of chloride intake should not exceed the amount of sodium intake, i.e. one should not ingest other chlorides except table salt. The normal amount of chloride in the body is less than that of sodium, a part of the sodium ions being neutralized by bicarbonate anions. Any excess chloride requires additional work for the kidneys, to eliminate it from the body. Excreting chloride is also likely to take with it some of the useful metallic cations that you have supplemented.
Other metallic cations than sodium should be combined with organic anions, i.e. one should ingest their salts made with organic acids chosen from normal nutrients.
Good choices are potassium citrate (one of the cheapest organic acids that is also a normal component of food) and magnesium bisglycinate (because the magnesium salts of cheaper organic acids, like citric acid, are not soluble in water, while magnesium bisglycinate not only is soluble in water, but it also does not form solid precipitates with other components of food, which would prevent its absorption in the gut).
Both potassium citrate and magnesium bisglycinate can be found as cheap pure powders, which are preferable to any other forms, like capsules, which contain useless excipients for which there is no reason to be introduced in your body.
Besides sodium, the only other exception to the rule of using salts of organic acids is calcium, which can be supplemented as calcium phosphate or bone meal. However, calcium phosphate powder is not soluble in water, so unlike sodium and magnesium salts powders it cannot be put in drinking water, but it should be mixed like table salt in some solid food, before eating it.
It's worth noting that glycine can have negative psychological effects on some people, anxiety, insomnia, etc. Supposedly it's because it's a NMDA receptor co agonist.
The amount of glycine that accompanies the at most a few hundred milligrams of magnesium per day that you might want to use as supplement is relatively small in comparison with the glycine content of most kinds of food that you eat anyway.
Bisglycinate is the cheapest form of chelate magnesium, but there are somewhat more expensive forms of chelate magnesium, which use other amino-acids instead of glycine. Someone who is very sensitive to glycine could use those. The non-chelate forms of magnesium are prone to form solid precipitates that pass through the gut without being absorbed.
Much more glycine than in a Mg supplement would be provided by any food with gelatin, so if someone really has glycine sensitivity, that should manifest when eating food with gelatin.
Most of that site reads as a "trust us, sound makes you heal!" I don't see a single thing there that tries to explain the correlation between bumping your brain with sound and "restoration". It doesn't explain what it restores, why it works better than nothing.
Honestly this website doesn't seem to even be trying to sell anything. "Help your sleep" how? "doesn't make you sleep longer" but what does it do instead? Do I feel more restored when I wake up even with a shorter sleep time? It's not even trying to sell me on a specific outcome that I John Everyman is facing.
Snake oil things are typically very light on details and this site is also light on details. Maybe it's a victim of form over function? The site's design looks nice but has very low information density.
Fair criticisms.
However, there are FDA limitations on what you can say regarding devices that both measure and affect biomarkers and neuromarkers while the user is unconscious.
This is why we describe the neural function of sleep, but can't specifically get into details regarding increases in slow-wave activity, 15% decrease in early night cortisol, 14.5% increase in HRV, etc etc. We can link to the research, but can't say "this is what we do".
We are relying on user testimonials, which we are gathering through our beta testing and beyond.
At the same time, we do describe the "clearer thinking", "immune function", "stress" etc about half way down the page.
We decided to go with a low device cost and subscription to make it cost effective to purchase when finances allow us to get to a monthly instead of yearly subscription.
When I worked on Microsoft Band, we wanted to have really nice side buttons. Great feel, not only the material, but the perfect spring tension, a great clicking sound, everything. We had someone who obsessed over them month after month, trying different polishes, colors, materials, spring types, everything.
$2 a button, 2 buttons, $4 total on the BOM. Absurd. Yes the buttons were great, but we couldn't justify $4 for them. But that was the best any of our suppliers were able to do for metal buttons. We had spun off the Xbox accessories team, so we really knew buttons, but we really knew plastic buttons of a certain type. Metal buttons were kicking our butts.
Then someone asked: "have we tried a watch supplier?"
So yeah, like 25 cents per button for the exact same buttons. It wasn't that our other suppliers were ripping us off, they just weren't setup to make the type of buttons we needed, in those quantities, at the price we needed.
So what I'm saying is, your base product is already expensive. It costs the same as an entire smart watch capable of running an operating system. You probably have lots of specialized components, lots of custom hardware, but your base price is already high and I hope you don't have any $4 buttons in the product, they are easy to overlook. I also get it is because you are doing small batch runs, and you are amortizing all the tooling and dev work over a small number of units, but you are not low cost. You are an unknown product selling adjacent to a bunch of snake oil people, if you were just $200 by itself I could maybe see it as an impulse buy to try out, but plus the subscription, ouch.
I have a brand new Microsoft Band still in the box! When I ordered one, Microsoft sent me two and refused to refund me and take the 2nd one back :) I still look at that device fondly.
What makes you think an EEG headband is not just a smartwatch in a different form-factor? We're running an nrf53 plus extra storage. We don't have a screen, but screens are fairly inexpensive. We've got a full EEG system.
We have actually done pretty well at limiting the number of parts. The headband is so thin it's buttonless. We didn't have room for a button and created a fully buttonless interface.
There are parts you wouldn't think would be expensive but are. Bone-conduction transducer, heart-rate sensor, etc etc. Yes, we could ditch the HR and save some money but we feel it will be beneficial.
Watches are a thing that exists. EEG sleep headbands aren't. We had to create our own materials for EEG sensing. Everything else on the market have limited life and aren't comfortable enough for sleeping.
Of course, all the plastics and TPU are custom. The fabrics are custom. Our scale is also very small right now, so I fully expect the cost to come down significantly with scale.
Yeah I get the custom parts thing, curved screen, curved batteries, and our optics stack for hr was custom built, now days those are off the shelf and a dime a dozen.
I guess my main point is that your device's price is firmly in the luxury impulse purchase category. It is only low price from the perspective of high earning coastal tech workers.
> Of course, all the plastics and TPU are custom.
Ugh getting the coating right on band took so long. Sunscreen stains everything.
That and a lens cover for the UV sensor that was transparent to UV and not just cheap plastic that'd shatter the second you slightly banged into it.
Does the idea of submitting one's self to using something like this not terrify anyone else? The more true the effectiveness of these products become, the more they have the possibility to do the inverse on accident (or potentially on purpose), no?
I think it should. Our system is closed-loop and we monitor the real-time change in brain-wave activity. The process is very precise, and must be (80ms window for a 50ms pulse).
When we first started, many in the sleep community were against using these techniques. A significant number of the studies look specifically at safety, and often people report to these as "null results" when in fact what was being examined was the potential negative impact.
One example is the study on metabolic function [1], which showed no result in healthy men. It did not harm their metabolic function, though it also didn't improve it (though I'm not sure how you would measure improvement in healthy metabolic function).
For our company, there are many modalities and capabilities we are building for the future, we began with auditory stimulation and this one in particular due to the low-risk and volume of research.
Yeah, my co-founder and I were discussing this yesterday.
It was always in the plans, we're all actually on Android, but we wanted to get the user experience right on one platform and not spread ourselves too thin, but we're thinking AI makes a big difference here, so... no announcement yet, but it was never going to be iPhone forever.
I still can't say for sure we'll have android at launch, but we'd like to.
If you're all on Android, the inability to dogfood iOS will hinder that "get the user experience right", won't it? Not being an app developer, I'm curious why a cross-platform framework wasn't chosen. Aren't there several good ones to choose from these days?
We have have iPhone dev phones, I just mean that for our daily use we don't use iOS. When I'm travelling I use iOS as my 2nd phone as well. So we dogfood it, but it's more like a snack specifically just for our app.
The structure of the app is in rust, so that is shared across platforms. The challenge is in the UI elements, and the cross platform stuff just isn't as good as going native there. You end up building a bunch of custom components anyway.
I'm also a bit wary of BLE on Android. iOS, is a bit more of a known entity.
It isn't just iOS and Android either. Because we support researchers, we also have a web app that they use. It surprising how large our technology footprint is across services, firmware, web, and apps.
1) Though they are comparing parents specifically, without the baseline of what the hunter-gatherer groups sleep was like without children, are they comparing hunter-gatherer group to industrialized people? Or are they comparing parenting?
50% of people rate their sleep as an F, and another 21% a D grade [1]. That feels likely everyone is failing at sleep, not just parents.
2) specifically in mothers, as motherhood has shifted later in life, the early years with young children are now often overlapping with perimenopause, so mothers are hit with the double whammy of sleep disruption. I blogged about this a few months ago [2]
The study is still mostly focused on the antiquated idea that sleep duration is a predictor of sleep quality. The latest research shows sleep regularity is a better predictor of morbidity than sleep duration. I wrote about hot the Neural Function of Sleep dictates this [3]. Studies in shift workers (I can never find the link) shows regularity trumps duration for both subjective sleepiness and cognitive performance.
The article does mention the increase in prolactin during breastfeeding, but the tiredness of parenting doesn't only last through the first year (apparently the average of breastfeeding in Australia is 6 months). The hunter-gatherer societies I'm sure breastfeed for longer periods.
I work in neurotech/sleeptech as the co-founder of affectablesleep.com and we have a keen focus on parents of young children and specifically enhancing the Neural Function of Sleep, not sleep duration which everyone obsesses over [4].
> "That feels likely everyone is failing at sleep, not just parents."
This doesn't really mean anything for comparing parents with non-parents, since it's self-reported so "failing" could mean "missing several hours of needed sleep each night" to one person and "failing to hit higher-than-needed sleep target twice a week" to another.
I used to work in WebRTC back in it's earlier days and our team developed the open-source rtc.io. (https://github.com/rtc-io)
I never would have imagined that OpenAI is sending the full audio of a request to their servers. I had always assumed the audio was transcribed locally and then sent to the server.
The only reason I can think they'd want the full audio is for later model training, which, ok, fair-enough, but this can still likely be done without the limitations of WebRTC.
I can't think of a single time I've looked at a product tour and thought "well, I'm really glad they told me that, I never would have figured that out.
What the product tour I think often misses is that people don't want to learn your entire tool at one time.
They came to do one thing, that one thing needs to be brain dead simple.
Over time, you can show people what else they can do. But a product tour isn't the way to do that.
I think progressive UIs where you expose more and more to the user over time is the way to go.
If you're thinking "but I have so many features and capabilities this person needs" you probably haven't identified what the one thing people are paying you for is.
Computers were nowhere for ever, then everyone had them.
The internet was tiny, then everywhere.
Smartphones were a teensy market, then everyone had them.
GLP1s were for a small group of diabetics, now a significant portion of the population take them.
This is how things playout time and time again.
Does it mean the commentors 10 years is correct? No. But it also doesn't need to be incredibly optimistic. All it takes is getting the robots right, and there are multiple companies who seem very close.
Mostly just the cost, yeah. It will be like buying a car. The economics will have to make sense for regular people, while it starts popping up in tons of places and become a status symbol.
Digital computers existed for ~10-20 years before hitting the consumer market. It took almost a half-century for the microprocessor to become a ubiquitous appliance.
Interesting comment they have towards the end about "targeted memory reactivation can disrupt sleep".
It is important to note the study they are referring to is "targeted memory reactivation with sleep disruption", there are methods of doing targeted memory reactivation without sleep disruption.
I work in neurotech/sleeptech as the founder of affectablesleep.com, and though we are mostly focused on slow-wave (deep) sleep, we have been looking into memory reactivation, lucid dreaming and other stimulations for additions.
Sorry for the off-topic, but I was curious about Affectable so I opened the website. I saw it's very thin and light and comfortable, but I struggled to find out what "it" is and what it does for me. It's kinda buried.
I was interested enough to click through the different links in the footer. And just as I reached the purchase page, I see that it requires "an iPhone running iOS". Unsure why it requires an iPhone; and no info on a timeline for iPhone-less customers. But that immediately rules me out as a customer.
I feel like the landing page would be a lot better if it started out focusing on what it is & how it can help me.
Apologies again for the unsolicited advice. Just wanted to share my impressions in case it's helpful.
I used to do this regularly when I first started coding, I called them "Codemares". They were like nightmares with the shouting of commands I didn't quite understand would invade my dreams.
It seems to me that this is the purpose of nightmares. I especially noticed this after having kids. They are not by default scared of snakes and such but if they see a nature documentary of a snake biting something or even a cartoon bad guy, it's enough to trigger bad dreams which reinforce the fear and it's far stronger the next day.
IMO this is under-appreciated in current AI models. RL is not very effective in avoiding crocodiles for example, by the time like 5 of your tribe-mates are eaten it's far too late. You need some mechanism that ensures the danger is learned after just a single incident.
We built iOS first because that's what most of our users have, though the team is mostly on Android.
We built our logic in Rust, so that compiles cross platform. It's really just the UI layer that is per platform.
Though I had initially written our first demo in React-Native, the team wanted to go with native UI.
Lately we've been talking about moving the timeline for our Android app up because we can get AI to help using the iOS app as a foundation.
I haven't seen anything really new and compelling in iOS/Android land. The promise of right once run anywhere continues to be a promise, and the bluetooth libraries across different devices (particularly in Android land) has me concerned.
reply