@prologic@twtxt.net @carsten@yarn.zn80.net

There is (I assure you there will be, don’t know what it is yet…) a price to be paid for this convenience.

Exactly prologic, and that’s why I’m negative about these sorts of things. I’m almost 50, I’ve been around this tech hype cycle a bunch of times. Look at what happened with Facebook. When it first appeared, people loved it and signed up and shared incredibly detailed information about themselves on it. Facebook made it very easy and convenient for almost anyone, even people who had limited understanding of the internet or computers, to get connected with their friends and family. And now here we are today, where 80% of people in surveys say they don’t trust Facebook with their private data, where they think Facebook commits crimes and should be broken up or at least taken to task in a big way, etc etc etc. Facebook has been fined many billions of dollars and faces endless federal lawsuits in the US alone for its horrible practices. Yet Facebook is still exploitative. It’s a societal cancer.

All signs suggest this generative AI stuff is going to go exactly the same way. That is the inevitable course of these things in the present climate, because the tech sector is largely run by sociopathic billionaires, because the tech sector is not regulated in any meaningful way, and because the tech press / tech media has no scruples. Some new tech thing generates hype, people get excited and sign up to use it, then when the people who own the tech think they have a critical mass of users, they clamp everything down and start doing whatever it is they wanted to do from the start. They’ll break laws, steal your shit, cause mass suffering, who knows what. They won’t stop until they are stopped by mass protest from us, and the government action that follows.

That’s a huge price to pay for a little bit of convenience, a price we pay and continue to pay for decades. We all know better by now. Why do we keep doing this to ourselves? It doesn’t make sense. It’s insane.

@carsten@yarn.zn80.net

I have to write so many emails to so many idiots who have no idea what they are doing

So it sounds to me like the pressure is to reduce how much time you waste on idiots, which to my mind is a very good reason to use a text generator! I guess in that case you don’t mind too much whether the company making the AI owns your prompt text?

I’d really like to see tools like this that you can run on your desktop or phone, so they don’t send your hard work off to someone else and give a company a chance to take it from you.

@prologic@twtxt.net @carsten@yarn.zn80.net

(1) You go to the store and buy a microwave pizza. You go home, put it in the microwave, heat it up. Maybe it’s not quite the way you like it, so you put some red pepper on it, maybe some oregano.

Are you a pizza chef? No. Do we know what your cooking is like? Also no.

(2) You create a prompt for StableDiffusion to make a picture of an elephant. What pops out isn’t quite to your liking. You adjust the prompt, tweak it a bunch, till the elephant looks pretty cool.

Are you an artist? No. Do we know what your art is like? Also no.

The elephant is “fake art” in a similar sense to how a microwave pizza is “fake pizza”. That’s what I meant by that word. The microwave pizza is a sort of “simulation of pizza”, in this sense. The generated elephant picture is a simulation of art, in a similar sense, though it’s even worse than that and is probably more of a simulacrum of art since you can’t “consume” an AI-generated image the way you “consume” art.

@carsten@yarn.zn80.net @lyse@lyse.isobeef.org I also think it is best called fake. Art is created by human beings, for human beings. It mediates a relationship between two people, and is a means of expression.

A computer has no inner life, no feelings, no experience of the world. It is not sentient. It has no life. There’s nothing “in” there for it to express. It’s just generating pixels in patterns we’ve learned to recognize. These AI technologies are carefully crafted to fool people into experiencing the things they experience when they look at human-made art, but it is an empty experience.

@carsten@yarn.zn80.net Who says you need to use anything like that? Where’s the pressure coming from?

@carsten@yarn.zn80.net yeesh, it’s a for-pay company I wouldn’t give them the output of your mind for free and train their AI for them.

@xuu@txt.sour.is this is alarmingly catchy

@xuu@txt.sour.is everyone’s moving to gated communities!

@prologic@twtxt.net ack, I didn’t see this before. Get well soon!

ChatGPT and Elasticsearch: OpenAI meets private data | Elastic Blog

Terrifying. Elasticsearch is celebrating that they’re going to send your private data to OpenAI? No way.

@prologic@twtxt.net I’m a bit of a GPU junkie (😳) and I have 3, 2019-era GPUs lying around. One of these days when I have Free Time™ I’ll put those together into some kind of cluster….

@darch@neotxt.dk yes!

@prologic@twtxt.net yeah. I’d add “Big Data” to that hype list, and I’m sure there are a bunch more that I’m forgetting.

On the topic of a GPU cluster, the optimal design is going to depend a lot on what workloads you intend to run on it. The weakest link in these things is the data transfer rate, but that won’t matter too much for compute-heavy workloads. If your workloads are going to involve a lot of data, though, you’d be better off with a smaller number of high-VRAM cards than with a larger number of interconnected cards. I guess that’s hardware engineering 101 stuff, but still…

@prologic@twtxt.net I would politely suggest again that we not react to people with bad attitudes who talk shit about yarn. If twt is forked, it should be forked to add features that are otherwise not possible. Not to appease people who will probably never be appeased.

On LinkedIn I see a lot of posts aimed at software developers along the lines of “If you’re not using these AI tools (X,Y,Z) you’re going to be left behind.”

Two things about that:

1. No you’re not. If you have good soft skills (good communication, show up on time, general time management) then you’re already in excellent shape. No AI can do that stuff, and for that alone no AI can replace people
   
2. This rhetoric is coming directly from the billionaires who are laying off tech people by the 100s of thousands as part of the class war they’ve been conducting against all working people since the 1940s. They want you to believe that you have to scramble and claw over one another to learn the “AI” that they’re forcing onto the world, so that you stop honing the skills that matter (see #1) and are easier to obsolete later. Don’t fall for it. It’s far from clear how this will shake out once governments get off their asses and start regulating this stuff, by the way–most of these “AI” tools are blatantly breaking copyright and other IP laws, and some day that’ll catch up with them.
   

That said, it is helpful to know thy enemy.

@movq@www.uninformativ.de Cheers! I’m happy to agree to disagree too of course! Thanks for engaging!

@xuu@txt.sour.is That has no relevance to the point!

Like, check it out. That link to DRY? It doesn’t render as a link in the webapp. However, it does render as a link, and works fine, in Goryon. I’ve seen before that Markdown tables render fine in Goryon but not in the webapp. They ought to behave as similarly as possible, right? So just in this small interaction there are three discrepancies between how the mobile app and webapp render Markdown.

@prologic@twtxt.net @movq@www.uninformativ.de this is the default behavior of pass on my machine:

Image

I add a new password entry named example and then type pass example. The password I chose, “test”, is displayed in cleartext. This is very bad default behavior. I don’t know about the other clis you both mentioned but I’ll check them out.

The browser plugin browserpass does the same kind of thing, though I have already removed it and I’m not going to reinstall it to make a movie. Next to each credential there’s an icon to copy the username to the clipboard, an icon to copy the password to the clipboard, and then an icon to view details, which shows you everything, including the password, in cleartext. The screencap in the Chrome store is out of date; it doesn’t show the offending link to show all details, which I know is there because I literally installed it today and played with it.

@xuu@txt.sour.is yeah, I know less about ISO27k (in part because you have to pay for access to the complete standards documents!!!), but I figured it was similar.

And in the latest “don’t store your passwords in the cloud” news, NortonLifeLock warns that hackers breached Password Manager accounts

@xuu@txt.sour.is that doesn’t seem to fit the spirit of the spec, at least by my read (I could be wrong obv). The example on Wikipedia’s webfinger page,

{
	"subject": "acct:bob@example.com",
	"aliases": [
		"https://www.example.com/~bob/"
	],
	"properties": {
		"http://example.com/ns/role": "employee"
	},
	"links": [{
			"rel": "http://webfinger.example/rel/profile-page",
			"href": "https://www.example.com/~bob/"
		},
		{
			"rel": "http://webfinger.example/rel/businesscard",
			"href": "https://www.example.com/~bob/bob.vcf"
		}
	]
}

and then the comparison with how mastodon uses webfinger,

{
    "subject": "acct:Mastodon@mastodon.social",
    "aliases": [
        "https://mastodon.social/@Mastodon",
        "https://mastodon.social/users/Mastodon"
    ],
    "links": [
        {
            "rel": "http://webfinger.net/rel/profile-page",
            "type": "text/html",
            "href": "https://mastodon.social/@Mastodon"
        },
        {
            "rel": "self",
            "type": "application/activity+json",
            "href": "https://mastodon.social/users/Mastodon"
        },
        {
            "rel": "http://ostatus.org/schema/1.0/subscribe",
            "template": "https://mastodon.social/authorize_interaction?uri={uri}"
        }
    ]
}

suggests to me you want to leave the subject/acct bit as is (don’t add prefixes) and put extra information you care to include in the links section, where you’re free to define the rel URIs however you see fit. The notion here is that webfinger is offering a mapping from an account name to additional information about that account, so if anything you’d use a "subject": "acct:SALTY ACCOUNT_REPRESENTATION" line in the JSON to achieve what you’re saying if you don’t want to do that via links.

Fomu | Crowd Supply

An FPGA board that fits inside your USB port

Whoa 🤯

@prologic@twtxt.net I started to write a snarky twt about Kafka and then deleted it because I didn’t want to be too negative 😆

git-bug

Distributed, offline-first bug tracker embedded in git, with bridges

Interesting!

What is NATS - NATS Docs

Might play with this at work next week.

I guess Google Hangouts is finally dead.

Why is Google such a mess at making messaging apps? This has more or less been a solved problem for decades. Google Talk worked well enough, and since it was based on XMPP and Jingle it was perfectly suited to become a large-scale text/voice/video messaging system. If they’d run with that they’d have been able to dominate that space, I think. Instead, they’ve created and shitcanned half a dozen messaging apps and platforms, flailing around copying someone else’s app (now they’re trying to copy Slack I guess).

@jlj@twt.nfld.uk @xuu@txt.sour.is hello! @prologic@twtxt.net and I were chatting about the question of globally deleting twts from the yarn.social network. @prologic@twtxt.net noted that he could build the tools and endpoints to delete twts, but some amount of cooperation from pod operators would be necessary to make it all work together. He asked me to spawn a discussion of the subject here, so here we are!

I don’t have enough technical knowledge of yarn.social to say with any credibility how it all should work, but I can say that I think it ought to be possible and it’d be good to do for those rare times when it’s needed.

I was inclined to let this go so as not to stir anything up, but after some additional thought I’ve decided to call it out. This twt:

Image

is exactly the kind of ad hominem garbage I came to expect from Twitter™, and I’m disappointed to see it replicated here. Rummaging through someone’s background trying to find a “gotcha” argument to take credibility away from what a person is saying, instead of engaging the ideas directly, is what trolls and bad faith actors do. That’s what the twt above does (falsely, I might add–what’s being claimed is untrue).

If you take issue with something I’ve said, you can mute me, unfollow me, ignore me, use TamperMonkey to turn all my twts into gibberish, engage the ideas directly, etc etc etc. There are plenty of options to make what I said go away. Reading through my links, reading about my organization’s CEO’s background, and trying to use that against me somehow (after misinterpreting it no less)? Besides being unacceptable in a rational discussion, and besides being completely ineffective in stopping me from expressing whatever it is you didn’t like, it’s creepy. Don’t do that.

@xuu@txt.sour.is proof of stake = people with money get more money. It accelerates the wealth inequality problems that are already plaguing us. Crypto has even worse wealth inequality than fiat currency systems, which is 100% predictable.

@brasshopper@twtxt.net bitcoin 🤮 that’s a hard no from me

I just resurrected a laptop that had been off for about six months and it needs 585 package updates.

Stumbled on WebID today. Besides being confusing, it doesn’t work on the site I tinkered with and it seems to be mostly abandoned?