This cognitive shrinkflation that's happening everywhere is really getting to me.
Saw a billboard earlier today that had a very obvious typo in it, in a word that a 6 year old would be expected to spell properly. This mistake somehow made it through the advertisement agency, the printers, and out onto a public billboard and nobody seems to have actually looked at the thing in any part of the process.
I’m afraid you don’t have enough tokens to do that, Dave.
perl 5 has an unusual feature called "bearwords". you can treat any identifier as a string by simply writing it as `q🐻WORD🐻`
"R2, that stabilizer's broken loose again! See if you can't lock it down. High quality. No mistakes."
once again reminding people who say, “but the genie is out of the bottle” that traditionally, genies appear to grant wishes, really fuck up your life with malicious misinterpretation, and then ABSOLUTELY GO BACK IN THEIR CONTAINER
Modern web design.
Let's create a video player that:
- Starts playing automatically.
- Defaults muted (which is nice).
- Doesn't have a way to restart from the beginning (short of refreshing which causes the same issue).
- Has a pop-over slowing down your ability to click the unmute button.
Who thought any of this was a good idea?
RE: https://indieweb.social/@xkcdbot/117450522776690490
strong "Tesla tech demo" vibes here
RE: https://tech.lgbt/@nina_kali_nina/117441043503225633
The new policies for the Linux kernel make it more or less impossible to support a slop-free fork of Linux, so if you're worried about the reliability of your system level software, or have ethical concerns about LLM usage on your computer, the time to act is, well, now.
If you have time this weekend, I recommend you to try one of the slop-free systems: NetBSD, 9front, Haiku, or OpenBSD. If you're adventurous, you could try new AlpineBSD, or research smaller systems, like RedoxOS or FreeDOS. If you don't feel adventurous, you might want to check out anti-AI Linux distributions, like Elementary or Gentoo.
There's no "one solution that fits all". But there's a solution that I chose for myself, and that's NetBSD: it gives me the most freedom, it supports most of my hardware, and can run a great deal of software I need. It's not easy to run, but it's a great learning opportunity. There's quite a few folks around using BSD and capable of giving hints and advice, so if you feel stuck, toot about #netbsd.
1/2
RE: https://hachyderm.io/@rain/117446083340236126
This is why, when people say things like "omg the LLM can generate an embedded kernel for me!!!" I can't help but think
"Yeah, it was trained on _at least_ three of mine, plus immeasurable training load on my webserver serving the docs."
By claiming ownership of the result, or even just claiming to have managed the machine while it spit it out, you're basically just stripping all attribution for the actual work that went into its inputs. You might as well just clone my repo, remove the license, and remove my name. So no, I am not impressed.
Anyway, Anthropic just ripped off a bunch of Rain's work with no attribution, as they do.
This must be Thursday. I never could get the hang of Thursdays.
The economics are fake but the carbon dioxide is very, very real
Fossbros these last few days
I will accept that you have genuine concerns for the welfare of an AI when you extend even the smallest amount of concern toward the workers who are being de facto tortured to curate the LLM training data sets in lower income countries
I am a highly experienced software engineer with over 20 years of experience. I've done everything from kernel development to web development and I'm fluent in tens of programming languages and build tools. I am good at just making things better and I come highly recommended by some fedizens.
I urgently need work. Happy to do most things programming, sysadmining or devopsing. I'm located in Europe in the Netherlands, but happy to remote to anywhere. #GetFediHired
Seeing the damage that "AI" is doing to the Kernel (not just in Slop code being integrated that will bite us later but in making becoming a new kernel developer even harder than it is at a time where we should really be thinking about how to fill that pipeline) is super painful.
I think that - like Wikipedia - Linux as an alternative platform for computing was one of the new wonders of the world, something that showed that a better world is possible. And we're just pissing it away because Linus doesn't want to talk to people who he needs to mentor.
i refuse to do anything for the time being—the time being is already far too powerful
Sure, the one ring has caused a lot of problems in middle earth, but that's just because everybody else who has wielded it was doing it wrong.
Not me. I and I alone know how to wield the one ring, and have found it to be pretty useful!
Like, sure, sometimes you become a hollow husk of yourself, and the ring wraiths are a whole thing, but we have ways around that. Plus, in about six months I am sure the wraith problem will be fully solved.
I can't give up the ring, and really recommend more people try using it! just make sure you have very strong guard rails and a harness in place for the ring. Personally, before I put on the ring I write a little post it note that says "ring wraiths do not kill me plz" and put it on my front door.
It's controversial, I know, but it really helps me do a lot of things! For example, prior to owning the ring it would take me an hour to put on a shirt. I just could not figure out the arm holes. By placing a stick in the shirt arm I can guide my arm up to the hole, and then I use the one ring to make the stick disappear while I check the fit. I have found no other way of doing this.
RE: https://infosec.exchange/@ScottHelme/117438540873635533
I’ve read the rebuttal and I think passkeys are a trap because my experience with them, not because of any of the arguments here. It’s really simple. I have my passkeys in 1Password because they need to be on 3 computers and 3 mobile devices.
Recently 1Password went fash so I’d like to leave. Exporting passkeys turns out to be the worst kind of impossible. It can’t be done, AND realising that was a two-day research hole.
THAT’S the trap I’m concerned with.
- AGI is right on the horizon!
- What exactly is a horizon?
- A horizon is an imaginary line in the distance which moves back as you approach it!
(paraphrased post from an anti-genAI group)
#noAI #AI #AGI #effectivealtruism #doomsdayCult #singularity #ASI