I mean, even if the NPU space can’t be replaced by more useful components easily or cheaply, just removing it is sure to save a small amount of power which equates to a possibly not so small amount of heat that needs to be dissipated, which takes not insignificant amounts of and/or requires slowing the system down. Additionally, the pathways likely could be placed to create less interference with each other and direct heat transfer which is likely to mean more stability overall.

Of course without a comparable processor without the NPU to compare to, these are really difficult things to quantify, but are true of nearly all compact chips on power sensitive platforms.

Yes, but I’m saying they’re making these laws and saying they need it. Many people agree that they need it and because they think they are still secure because they’re using an “encrypted connection”, assuming they don’t think they need to be secure from their government, they are supporting it. If they see that by letting the government steal their data they are also letting that scammer that keeps scamming their grandmother for her credit card to get that credit card number without even needing to scam her anymore, they may think twice about supporting the policy.

Problem is many of us are stuck with very low upstream bandwidth due to cable company ISP monopolies and/or data caps or just were running things on a small raspberry pi or something and the malicious requests will create extra expense or flat put denial of service for real traffic.

Oh totally, but that’s the intended purpose. The thing is they’re saying they can do all that and still allow people to have a secure connection to their bank or whatever, but that’s impossible. Eventually, backdoors always lead to making the security worthless whether it’s bad design like putting hinge screws outside of the door so thrives can just use a screwdriver to remove the door, or a backdoor for locksmiths or government, it’s a weak link it doesn’t matter how thick the door is if a screwdriver removes it or how hard the encryption is to break if it can be bypassed by getting the code used by locksmiths or government, bad actors will get ahold of it and use it.

Problem is scraper bots are way more aggressive and harder to block. If they were ignoring Reddit because they were taking content from IA but IA is willing to obey robots.txt whereas scraper bots are not, they just shifted the load of serving the bots or playing whack-a-mole with their block evading mechanisms. They aren’t going to stop the bots. It may result in being able to negotiate a license with the bigger guys, but that’s likely not going to make up for the money they spend on dealing with the bots in the long run. Of course companies like this don’t really think long term, it just looks good to investors this quarter.

To a point yes, for the crawler bots, but Anubis uses a lot more resources to keep the bots busy than a simple firewall ignoring the request. And if there’s no response vs a negative response, the requests are likely to fall off more quickly. And the even more significant load might be from malicious login attempts which use even more resources and Anubis likely won’t be as effective on those more targeted attacks depending on the types of services we’re talking about. Either way, firewall blocks are way, way less resource intensive than any of that, so as soon as you open up that firewall and start responding to those malicious or abusive requests they will become progressively more resource intensive to mitigate.

And this is the same reason why encryption backdoors would basically make encryption worthless. Doesn’t matter how strong the metal/encryption is if a backdoor exists to be the weakest link.

I just use the Play Store in sandboxed mode on GrapheneOS. The other stores out there haven’t looked all that trustworthy, especially ones that modify the apks. Some are even on common spam blocklists which may be for things other than abuse by the store portion of the site, but it reduces trust.

For anything open source I use Obtanium, the few proprietary apps I need I use the vanilla play store.

Just use a VPN outside the country or lease a low end VPS outside the country with decent bandwidth and host your own VPN endpoint. They can’t force ISPs to block all of the VPN protocols in general or they’ll lose a ton of businesses that rely on it for basic security of remote/traveling employees.

It’s very time consuming to detect and correct the small mistakes that LLMs make. Beyond one or two lines of code, it becomes much more time consuming to correct the multitude of subtle mistakes vs coding it myself. I use code completion that comes with my IDE, but that is programmatic completion, not LLM, and is much, much more accurate and in smaller chunks that are easy to verify at a glance. I’ve never known any experienced developers who have had a different experience. LLMs can be good for getting a general idea of how to code something in a new language or framework I’ve never touched before and more to help find actual examples rather than use the code directly in the IDE, but if I were to use LLM code directly that would be in a test project, never, ever in production code. I would never write production code in a language I’ve never used before with or without an LLM’s “help”.

I think it’s safe to say that all of the LLMs have been training their systems on any site they can get their hands on for some time. That’s why apps like Anubis exist trying to keep their crawlers from killing their bandwidth since LLM companies have decided to ignore robots.txt, copyrights, licenses, and other standard practices.

You can mitigate some risks with software like fail2ban to slow down some of the hacking attempts, but you will still be susceptible to, sometimes unintentional, denial of service attacks from ever persistent “AI” crawler bots as well as the constant barrage of automated hacking attempts. If you’re bandwidth is not able to handle it or you have bandwidth caps, you’re likely going to have issues.

Curious if the account is a legitimate Rus Education account that got hacked. I’ve been seeing that a lot lately where random legitimate businesses are contacting people about totally unrelated subjects. Seems scammers are probably using hacked business accounts to allow them to send messages to anyone without restrictions.

Are you wanting something that you don’t have to download from GitHub yourself (so a project that hosts a docker container somewhere and just code is in GitHub is OK), or are you looking to boycott any project that is not boycotting GitHub and so any part of that project should not use GitHub for any code at all in which case possibly even dependencies should not be on GitHub even if they publish their distributions elsewhere? Or somewhere in between?

Local storage backups (local to the servers wherever that is, so “relatively local”) should be the initial backup, then those backups should be what’s synced to the off-site/third-party provider, generally. But it really depends on the types of tech and how those backups are generated.

There are definitely simple text based filters and alert systems, but I haven’t seen more advanced systems that use “AI” or other context aware technology or automated image processing.

If this was a self-hosted forum, yes, that’s an option. But for professional purposes, a dedicated off-site backup provider is better than having storage at an office site.

I’m not sure how “suspicious behavior” could be relevant to a seller issuing you a refund. Nor do I understand why a government ID would help with that kind of situation. If they are saying it’s because the credit card might be stolen, that doesn’t really make sense for a refund. If they’re saying the account might be hacked, then again, I can see limiting purchases, but not refunds. Are you sure this was a real email from Amazon and not a phishing email? I’d contact them again to verify, first. Then if you can’t resolve it, go to your bank and ask for a reversal of the charge.

Not too surprising. Data backups need to be with different providers. The article seems to think it’s not “putting all your eggs in one basket” because the provider had redundancy. But that’s not much different from storing physical backups locally because they were stored in a fire-proof safe. Sure you made backups, but by storing them in the same building as the servers means the same disaster that could take out the servers could take out the backups. A “fire-proof” safe will protect it from some things that won’t protect the servers, but there are still types of disasters that could take out both, like a big enough bomb rather than just a fire.

What if AWS went bankrupt and the servers were repossessed and sold off with the data spread across all the different new owners of the disparate data centers? What if Amazon just decided AWS was no longer profitable and shut it all down.

Sure that’s not going to happen to AWS right now because it’s hugely profitable, but a serious US market crash combined with a major escalation by the current administration in the increasing surveillance state in the US which could kill the trust in the company, cause a massive migration to EU based companies and cause the subsidiary company that holds the data to go bankrupt without necessarily killing Amazon as a whole. Those subsidiaries often “run at a loss” even with extremely high income in order to divert profit to shareholders, claim tax breaks on “losses”, and eliminate liability to the main company.

The legal proceedings of bankruptcy or other events could put the data in legal limbo for years before it’s accessible again.

I’ve seen some, but it’s way, way less common than the major corporate platforms, but since there’s no pay and no ads to make money off of, it’s not too surprising. As the article briefly mentions, there’s just too much monetary incentive to make junk vs quality content on most for-profit platforms.