No, because until the dot-com era when “tech bros” first became a thing, a lot of the computing pioneers were leftist/hippie counterculture types. Think MIT AI Lab (RMS), The Homebrew Computer Club (Woz, Captain Crunch), Germany’s Chaos Computer Club, etc.
Further reading:
https://bigthink.com/the-present/1960s-counterculture-created-silicon-valley/
Surely if it’s that bad at least one of the reasons is simple enough to list in a sentence or two here.
But they said “DaTA cenTeRs wILL BrINg JObS”!
OnlyFans, maybe? (I have no idea if PizzaCake has an OnlyFans.)
cat /proc/cpuinfo gives me a bunch of copies of this (one per CPU core):
$ cat /proc/cpuinfo
processor : 0
BogoMIPS : 49.15
Features : fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdh
p cpuid asimdrdm jscvt fcma lrcpc dcpop sha3 sm3 sm4 asimddp sha512 sve asimdfhm
dit uscat ilrcpc flagm sb paca pacg dcpodp sve2 sveaes svepmull svebitperm sves
ha3 svesm4 flagm2 frint svei8mm svebf16 i8mm bti mte mte3
CPU implementer : 0x41
CPU architecture: 8
CPU variant : 0x1
CPU part : 0xd46
CPU revision : 1
Not sure how “MTE3” relates to “EMTE”…
Scraping more than necessary is so stupid that I just sort of dismissed it as a straight-up mistake that will eventually be corrected. I was arguing based on general principle, not specific current practice.
Obviously, yes, the AI companies should fix their (probably vibe-coded) scrapers so they stop misbehaving; that should’ve gone without saying.
I’ve been avoiding reading that because I still hold out hope. I think some community effort started up to do it; I wonder how that is going?
De-de-regulating. The double negative is important! 🤓
From https://grapheneos.org/faq#recommended-devices :
8th generation and later Pixels also have support for the incredibly powerful hardware memory tagging security feature as part of moving to new ARMv9 CPU cores. GrapheneOS uses hardware memory tagging by default to protect the base OS and known compatible user installed apps against exploitation, with the option to use it for all apps and opt-out on a case-by-case basis for the few incompatible with it. Pointer authentication, branch target identification and other ARM security features were also introduced at the same time as hardware memory tagging.
No, not “always has been!” Some of us remember the '90s, before the corporations found it and ruined everything.
That means using existing designs with a die shrink or something is often the right choice when you’re shooting for the low cost portion of the market. The technology referenced here, enhanced memory tagging extensions, is a hardware implementation of tagging that requires a specific revision of the arm design and it’s a relatively new change (mte, the precursor to emte, is flawed and not used).
I’m writing this on a Pixel 8 running GrapheneOS, which I’m pretty sure is a “recommended” device that supports all the security features and is also multiple generations old.
If new lower-end devices use die-shrink versions of previously flagship SoCs, as you say, then I simply do not buy the notion that there’s any good reason not to put a die-shrink version of this SoC in a new lower-end phone.
That’s not even all of the problem. The other part is: “Windows Update allows third-parties to install computer viruses on your PC.”
Okay, if efficient APIs existed and they weren’t incompetently failing to use them, it wouldn’t be a problem. Happy now?
(I should’ve addressed that in my previous comment, as I was aware of how one of the Lemmy instances was taken down by scrapers the other day despite the fact that they could easily get all the content simply by consuming ActivityPub directly. But I was naively hoping it wouldn’t be necessary because, as you can see from this text, it would’ve cluttered up my writing with double the words.)
It is the “selling shit back to us” specifically, not the “scraping,” that’s the unethical part. If the AI companies were doing the same scraping (and destructive rare book scanning, for that matter), but were using the data to populate archive.org, would it still be a problem? I would argue “no.”
So probably good that GrapheneOS skips this one. Maybe even more to a different line of devices that’s is not moving in such a proprietary direction.
GrapheneOS didn’t support only Pixels because they liked Pixels; they only supported Pixels because Pixels were the only phones that had the necessary security features. Now that Pixels don’t have those features either, the number of new phones on the market that are suitable for GrapheneOS has changed from one to zero.
In other words, there is no “different line of devices that’s is not moving in such a proprietary direction” for them to switch to!
(Ignoring the Motorola phone because it doesn’t exist yet and there’s no actual hard guarantee it ever will, BTW.)
We’re talking about security features here, not processing speed. There’s no excuse for even the cheapest SoCs not to support them. Withholding them from the bottom end of the lineup is pure product differentiation (a.k.a. profiteering, a.k.a. spite and contempt).
The scraping wouldn’t be a problem if Reddit simply provided an RSS feed or other data-efficient API. The “ramfucking” is caused by the attempt to block bots; it is entirely self-inflicted.
Remember, it’s all our content to begin with and Reddit does not have any right to try to lock it up for itself.
That doesn’t mean I like all the AI bullshit going on, BTW. But the problem is the generation of the slop, not the data accessibility.
The entire fucking point of the Web was to make information as easily-accessible as possible, structured and semantically tagged, and consumable by humans and further machine transformation alike. “Scraping” is facilitated by design!
Using Javascript to deliberately break that is evil and every programmer who participates it is a piece of shit. No exceptions.
Meanwhile, the real solution – de-deregulating the utilities and getting rid of the moral hazard of having them be for-profit corporations at all to begin with – wasn’t even on the table.
Fixed the meme: