Yeah that's a great point, the data really makes you pause and say "is this worth the cost?"
The other wearable that has had an extreme impact on my lifestyle has been a Dexcom CGM through Levels. I could write pages about this, but in some ways it has been more impactful than anything else. Knowing how your body reacts to particular foods, and seeing the immediate feedback is super eye opening.
And it extends into sleep as well, looking at what your blood glucose does overnight based on stress, exercise, what you ate, when you ate it, it's quite amazing to see how this subsystem in your body is reacting.
I'll never forget the first time I ate sushi (which mentally I thought of as a "light" food, even a snack) only to see my blood glucose rocket past 200 mg/dLf.
You can also see the effect of alcohol here. Not just with carbohydrate heavy beers, but with the fact alcohol itself suppresses your blood glucose levels, since it causes your pancreas to spike insulin, which disposes of glucose. So it's yet another way alcohol causes a disturbance in your body's attempts at homeostasis.
Alas, a CGM through levels is quite pricey because of US healthcare bullshit about giving CGMs to people without a diabetes diagnosis, but doing a month or two of CGM once per year I think it's a great investment (approx $200 for three sensors which last a total of 30 days)
In terms of wearables, it's quite unobtrusive because it's literally just a ring. There is no display to become distracted with. The inside of the ring will glow when it's taking measurements (red for blood oxygen, green for heart rate) so that might surprise you when the lights are out, but that's about the extent of how "intrusive" it is.
The battery life is very good, I typically get several days to a single charge, which is far better than the Apple Watch. But most importantly it's comfortable to wear during sleep - indeed it takes the majority of its measurements only while you're sleeping. YMMV but there's no way I can wear an Apple Watch to sleep, while the Oura ring works perfectly. I've worn the last two generations of Oura ring and would definitely recommend.
Sharing my own anecdata: my Oura ring shows that on days I drink, RHR is 10bpm higher and HRV about 15ms lower. Exercise has a huge impact on sleep quality too.
My theory is that when you're young, your body has a much wider margin for error. A few drinks or sloppy exercise routines never seemed to impact my work. As I've aged, the margin has shrunk, and now there's a measurable effect that I can no longer ignore. Aging is in itself a handicap, but now after too many drinks, the next day both my cognitive sharpness and motivation to even care about shoveling bits around a network have declined.
Now I need to actively stack the deck in my favor, rather that engaging in behaviors that further handicap myself. So that means exercise, a sleep routine, zero alcohol most days of the week, reducing stress, all that goddamn stuff I never had to worry about ;)
Another way to think of it is this: at 40 or 50, can you afford to be lugging around a 20% RHR penalty all day? That will surely come home to roost, probably in the form of atherosclerosis. Remember the most common clinical presentation for a heart attack sufferer is not chest pain, but being dead on arrival.
I bring this up because it's worth nothing that our bodies and minds are remarkably good at covering up problems until far too late. When you get those early warning signals (high RHR, feeling of cognitive decline, etc) trust your body, be proactive, and employ countermeasures.
I don't get to exercise my dusty algorithms knowledge nearly enough to follow this nearly as well as I'd like. Are we talking about the bin packing problem and approximation ratios? So, is the intuition here that when the ratio of bins to objects goes up, the worst-case performance for the algorithm goes down into the toilet?
(aside: fascinating example of a US military slide deck, like a 90s fever dream made in Harvard Graphics)
Here’s also a Voices from DARPA episode, which suggests they’ve been funding research since at least 2012: https://blubrry.com/voices_from_darpa/30975140/episode-17-gu...
In terms of ensuring designs are not tampered with seems like watermarking is the direction they’re exploring.
Another fascinating concept was the “silicon odometer” to ensure recycled chips aren’t reintroduced into the supply chain.
Asianometry has a fascinating video on that issue: https://youtu.be/7epnv43jGV8