Beyond LEO requires more fuel and a bigger rocket to launch, meaning more cost. It creates greater latency due to the greater distance. Also, they want these satellites to have a 5 year lifespan because terrestrial ISPs and cellular providers and datacentre operators are continually upgrading their hardware. So they will probably want to de-orbit and replace them anyway, because moving them to a graveyard orbit will result in the graveyard getting very full very quickly.

It also causes issues when satellites malfunction, because they won't naturally de-orbit in a practical amount of time. Failure to reach the intended orbit, resulting in an uncontrollable satellite, is one of the most common modes.

My browser shreds cookies as soon as I leave a site in most cases, as well as all other site date. These days the tracking works based on multiple signals, so even if you delete the cookies, if the IP address and browser signals like user agent and screen resolution match, they will re-associate that identity with you. You need to screw with a lot of metrics to throw them off.

In my country a spam lawsuit against 50 people where only one of them is possibly "guilty" of a civil offence with a relatively small financial loss isn't going to fly. They have largely given up suing people here because such speculative invoicing scams tend not to stand up to judicial scrutiny. At best an IP address identifies a subscriber, who may not be the person who downloaded the file, and who isn't under any legal obligation to help determine who it was, and who can't be held liable as there are no reasonable means for them to prevent such "abuse".

I didn't say no rocket launches. We just need to evaluate the impact, and take steps to prevent damage. We need to know what is in the satellites and what the effect of them burning up is.

I wouldn't say they are doing it wrong, I'd say that there is a fundamental conflict between privacy and anti-bot measures.

For privacy reasons I don't want a unique IP address. I want a shared one, and if it's IPv6 I want it to rotate frequently. That's one of the reasons why I use a VPN. ISPs probably also like it because it means that without extensive logging, for which there is no business justification, they can't identify who downloaded some movie that the MAFIAA et. al. want to sue over.

But of course the anti-bot features would love everyone to have a fixed IP address assigned to their person. Failing that, they seem to prefer to just mass block shared IP addresses and force you to log in.

Stuff burning up does not disappear. It becomes part of the upper atmosphere. Before there were so few things burning up that nobody really bothered considering it, but now we are going to be having tens of thousands of satellites re-entering every year, it needs to be looked into.

Indeed. And a few that are not offenders at all and against quite a few misidentified people since this tech is _unreliable_.

I find IPv6 is worse than CGNAT for these anti-bot measures. YouTube is a good example. With IPv6 enabled all I ever see is "log in to prove you are not a bot". With it disabled and sitting behind CGNAT, I can usually watch some videos.

It already has been making false positive matches. There have been several stories about people randomly accosted as they entered stores like B&M, with the security staff claiming they were criminals and often breaking the law themselves in the process.

Because the database is shared by several different chains, it's something that you can't ignore if it falsely flags you. You need to get them to remove your face from it, and ideally claim some compensation for the misuse of your biometric data. The baseline is £250, but I'd be looking for at least £750 due to the hassle and embarrassment caused.

In some jurisdictions, it is not even classified as an attack.

The problem is a practical one. As soon as you restrict voting, some assholes will use that for their own purposes. If that effect could be prevented, voting could be restricted to those that actually understand what they are voting for or against and democracy would work much, much better.

As a CS PhD? I do not even thing that is possible without a major mental breakdown. I think you are hallucinating.

I was just stating the obvious to complete the argument.

Hmmm. Looks like I installed Adblock Plus on Vivaldi so long ago that I did not remember. My apologies.

The main problem they have is that their monopoly is not actually one. That it is all made up.

The amount of queries needed to steal a model (which is possible though queries) is prohibitive. This is probably just an attempt to keep up the fantasy that China cannot do their own models. That is nonsense. For example, small Switzerland has created its own open model (https://apertus-ai.org/), mostly as an academic project. China for sure can do it. All I see China doing is comparative benchmarks and such and that is perfectly legitimate.

If China can do it, that means the likes of Anthropic are not nearly as special as they present themselves to be. But that false image of superiority is what keeps their cash-flow going, because they are still very far away from becoming profitable (if they ever do). Hence, in order to be able to keep burning the stupid money (from "investors"), they have to push this narrative.