Yeah, I think that's right, but where your clients are is irrelevant. That's what confused me about the author's statement.
I believe most (all?) registrars should let you give them a global DNS server, at which point your domain should be resolvable globally. Whether it's reachable isn't at the registration level, it's at the DNS level. The blog author's comment confuses me. Maybe they mean like having a TLD that matches each country with users? Seems not worth it.
You can get a domain for like $10 a year from many places. The national registrar I looked up was like $7 a year (Brazil).
Generally ML models are trained once and then used in their trained state unchanging. They don't continue to learn.
I went down a slight rabbit hole on this. As far as i can tell (looking at manufacturer claims on their website for devices of the era), 2018 is right on the edge of when a lane finding algorithm would have started to incorporate ML techniques. That's when mobileye released the eyeq4, which claimed to run ML models, and Wikipedia says was used by Ford, BMW, Nissan, VW, and Honda, albeit on only some of their models. The previous Eyeq3 from 2014 seems to be more classical models, but still used some ML techniques, seemingly more for things like speed limit reading.
ML algorithms refer to quite a wide variety of approaches. Generally what separates them is when the parameters of a generic model/algorithm are optimized over some specific data, instead of using an algorithm handcrafted to the particular problem.
You can do custom notes for people in voyager
I think you're almost right about the high level principles, you've just forgotten that entropy scales as the log of the combinations not linearly.
So the entropy of a 128 bit key is not 2^128, it's just 128 and therefore the fundamental limit on the cost required to recover it that you're talking about is very small.
See landauer limit (which isn't quite the same thing since it's about irreversible computers which some quantum computers might not be, but gets at the idea of there being an energy cost to have information)
I believe the specifics of your thoughts about entropy of public vs private keys is also incorrect but not necessary to the core of what you're saying.
There's like two people's devices behind my router. The amount of benefit I get to hiding which device is sending traffic is pretty meaningless. And all my user devices are using privacy extensions anyway.
I don't quite understand what you're trying to prevent. Like if I have an iot device that doesn't support privacy extensions, who am I making the tracking harder for by using ipv4? Anything the device talks to already is probably tracking via the device authenticating in the first place. And anybody in the middle doesn't really learn much by knowing that a particular device is sending opaque traffic vs the house is sending opaque traffic.
Big tech tracks us because we voluntarily login to their services. And they do so mainly to sell us ads. If you don't want them tracking you, hiding IP doesn't help much. Like don't login to their services and block requests to their domains to prevent third party site integrations. If you need to use their services anyway, then you're fundamentally screwed.
Presumably this is
Tap for spoiler
Claude/Anthropic
.The words on the timeline seem related to
::: the words it uses while it does CoT processing :::
I like it!
I don't think the burial matters. It's not random atoms that are turned into C14. It's specifically nitrogen. And I think those interactions mainly happen in the upper atmosphere with cosmic rays. So it's the atmosphere providing the shielding.
The Wikipedia page for this explains it pretty well, especially the Physical and Chemical Details section.
We were in upstate NY, and got extremely lucky with a hole in the clouds right around the sun at totality.
The red at the bottom was unexpected and very cool to see. It's a solar prominence
I don't think there are usually any bacteria in your blood