Yes. Plus the turing machine has an infinite memory tape to write and read. Something that is in scope of mathematics, but we don’t have any infinite tapes in reality. That’s why we call it a mathematical model and imaginary… and it’s a useful model. But not a real machine. Whereas an abacus can actually be built. But an Abacus or a real-world “Turing machine” with a finite tape doesn’t teach us a lot about the halting problem and the important theoretical concepts. It wouldn’t be such a useful model without those imaginary definitions.

(And I don’t really see how someone would confuse that. Knowing what models are, what we use them for, and what maths is, is kind of high-school level science education…)

It’s a long article. But I’m not sure about the claims. Will we get more efficient computers that work like a brain? I’d say that’s scifi. Will we get artificial general intelligence? Current LLMs don’t look like they’re able to fully achieve that. And how would AI continuously learn? That’s an entirely unsolved problem at the scale of LLMs. And if we ask if computer science is science… Why compare it to engineering? I found it’s much more aligned with maths at university level…

I’m not sure. I didn’t read the entire essay. It sounds to me like it isn’t really based on reality. But LLMs are certainly challenging our definition of intelligence.

Edit: And are the history lessons in the text correct? Why do they say a Turing machine is a imaginary concept (which is correct), then say ENIAC became the first one, but then maybe not? Did we invent the binary computation because of reliability issues with vacuum tubes? This is the first time I read that and I highly doubt it. The entire text just looks like a fever dream to me.

I have it role-play for my amusement, bounce ideas and regularly ask it to give me 5 ideas about something, if I’m not creative myself. I also use it for translation. Either copy-paste an entire text or help me phrase things.

I’ve also fooled around with image generation, music, tried to program a website, had it rephrase my emails… But all of that wasn’t super useful to me. At lest not for my everyday tasks.

Other AI things I use (occasionally) are: Text-to-Speech, Speech-to-Text and text recognition.

Yeah, seeking support is notoriously difficult. Everyone working in IT knows this. I feel with open-source, it’s more the projects which aren’t in a classic Free Software domain, who attract beggars. For example the atmosphere of a Github page of a Linux tool will have a completely different atmosphere than a fancy AI tool or addon to some consumer device or service. I see a lot of spam there and demanding tone. While with a lot of more niche projects, people are patient, ask good questions and in return the devs are nice. And people use the thumbsup emoji instead of pinging everyone with a comment…

I feel, though… I you’re part of an open source project which doesn’t welcome contributions and doesn’t want to discuss arbitrary user needs and wants, you should make that clear. I mean Free Software is kind of the default in some domains. If you don’t want that as a developer, just add a paragraph of text somewhere prominently, detailing how questions and requests are or aren’t welcome. I as a user can’t always tell if discussing my questions is a welcome thing and whether this software is supposed to cater for my needs. Unless the project tells me somehow. That also doesn’t help with the beggars… But it will help people like me not to waste everyone’s time.

There are no tags as of now. Everyone is running the main branch from Git. I suppose that’s going to change at some point when PieFed deems itself ready.

Do you have some source for this IP thing in the EU? I wasn’t aware of any new privacy laws.

I’d go with the Full Disk Encryption. You can be sure everything is encrypted that way. Any additional complexity adds ways to mess up and compromise security. Entering the password is a bit cumbersome. But that’s part of the deal. I just carry my computer keyboard to my NAS and enter the password each time I need to reboot. Which doesn’t happen that often. There also used to be some tutorial somewhere on how to put a Dropbear SSH server into the initrd so you can enter the password over network.

I think Nobara is the other most(?) popular choice by gamers.

I don’t have much experience with gaming distros. I just think whatever it is, a computer shouldn’t bee too locked down for a kid so they can also install other things, try other tools like an office suite, video editing or content creator stuff and maybe even have the experience of messing up. Within limits of course.

That’s right. I don’t think there is a good way to do it. I just take whatever link is provided by the small Fediverse icon. But I don’t think it matters that much for your audience, they’re spread over several instances and it’ll be an external link for some of them, no matter what you do. I’m not sure whether we have the ambition to solve this. I don’t see anything the user could do. Either this gets handled in some way by the software, or it is how it is.

Btw, I think a great thing would be tariffs, or more taxes on private data. Make companies pay for hoarding them and for exporting them into other countries. Maybe not in every case, but if it’s not just processed directly for the user themselves.

Right, I must have scrolled past it. I think linking posts is fine. I mean they’re public and this is social media… At least I do it somewhat regularly and I’ve never had anyone complain.

I don’t get the reference. This is the first time I’ve read that claim. But I’d certainly hope people know there is a difference between End Of Line and End Of File… I mean they’re alike, they both end something. But it’s not the same thing. The article explains the details how it’s handled.

inwx.de

You’re right, the private data is a bit of a construed example. I just wanted to make the argument that this isn’t just about copyright law. Something could be fair use under copyright law, but still illegal to use for other reasons. Which is problematic when doing unsupervised web scraping, for example. It’s definitely an issue, but out of scope if we limit the discussion to copyright only.

I’ve just skimmed the last article, I’m going to read it tomorrow. But I don’t think I’d like to argue for extending copyright. I think that would be bad. But I think it’s debatable whether AI training falls into that category. I’m not sure how it is in different jurisdictions… Maybe it’s clear in the US? I always struggle to read American legislation. I can just say it’s not clear where I live. And that comes with consequences: Companies do AI in other countries like the USA or China, rather than in the EU. Which is an issue for our economy and scientific progress. And everywhere where law isn’t clear enough, that’s a disadvantage for smaller companies/institutions or individuals, since it’s the big companies who can afford lawyers easily. And it has consequences for me personally. For example Meta’s use policy for the newer Llama models excludes Europeans. I’m not allowed to use it. That might not be about copyright either, but definitely due to unclear regulations.

So I don’t advocate for extending copyright. My stance is, we don’t have clear regulation in the first place. I’d leave all exemptions and specifics in place. We can leave libraries, music, research and reverse engineering as is. But the current warfare is super unhealthy. We have some companies scraping everything, meanwhile other people come up with tarpits and countermeasures like Cloudflare with their AI labyrinth last week… One newer business model is introducing walled gardens so companies can make sure they’re the only one selling their userdata… I think that’s all very unhealthy. And it favors large companies doing that “research”. Meanwhile the internet gets flooded with slop, half the internet services are barely usable and we might end up with dystopian Skynet corporations dominating information flow anyways. And I think that’s the bigger issue than copyright. If AI proves to be disruptive, it needs to be used somewhat ethically. And I think the only way to do it is regulation. We need to even the field so research and non-profit gets a chance. We currently have “smaller” startups participating, and several companies release open-weight models. But we can’t rely on their altruism. My prediction is they’ll all stop once this starts to interfere with their business or those models get really useful. And then it’s going to be OpenAI and Anthropic & Co who get to decide what kind of information the world has access to. Which would be very bad. And they also offer little transparency. More and more people rely on these services and AI is very much a black box. And the large companies have stopped telling what went in a few years ago when all the copyright lawsuits started. The first Llama model still came with a scientific paper, detailing all the dataset that went in. But as far as I understand, they stopped soon after, when copyright lawsuits started. And the rest are trade secrets. So if someone were to use ChatGPT (which lots of people do), they’re completely at mercy of OpenAI. OpenAI get to decide in which ways the model is biased (or not), what it can and can not answer, what is fed to the users. I think that’s the main issue with it. (Along with slop.) Copyright of training data is some sort or sideshow. But I still think we have a lot of unaddressed issues with AI. And leaving that open is just going to help the big players. I think we need more clear regulation so a small company who can’t afford a lot of lawyers can also be 100% sure whether someting is fair use or whether it isn’t. And personally I think we need to hold them all accountable and force them to be more transparent with everything. Like a rough estimation of the datasets. And I’d force service of generative AI services to implement watermarking to at least try to tackle slop and people doing their homework with ChatGPT. Sure this can all be circumvented, but we can at least try to do something about it. And I’d also like if big companies bought at least one copy of a book they use to train their AI. Meta or OpenAI can afford to pay a few millions. Otherwise they just leech on people’s content. I think it’s unfair that some people take quite some time to write books, Reddit comments, Wikipedia articles and then someone else gets to make big profit from that. It’s not very straightforward to solve it, but I don’t think it’s very healthy for humanity to just hand over everything to greedy companies. And I also don’t think it’s healthy to embark in a warfare, which seems to happen right now. That way we’re likely to all lose access to free information.

Private conversations are something entirely different from publically available data

But that’s kind of the question here… Is data processing Fair Use in every case? If yes, we just also brought private conversations and everything in. If not: What are the requirements? We now need to talk about which use cases we deem legitimate and what gets handled how… I think that’s exactly what we’re discussing here. IMHO that’s the point of the debate… It’s either everything… or nothing… or we need to discuss the details.

Compensation for essentially making observations will inevitably lead to abuse of the system and deliver AI into the hands of the stupidly rich, something the world doesn’t need.

I’m not sure about that. I mean I gave some examples with licensing music on events and libraries (in general). Does that also get abused by the rich? I don’t think so. At least not that much, so that makes me think that might be a feasible approach. Of course it get’s more complicated than that. Licensing music for example brings in some collecting societies, and all those agencies have proven to be problematic in various ways and all the licensing industry isn’t exactly fair, they also mainly shove money into the hands of the rich… So a proper solution would be a bit more complicated than that.

I mean I’d like to agree with you here and have a straightforward parallel on how to deal with AI training datasets. But I don’t think it’s as easy as that. We can’t just say processing data is Fair Use, because there are a lot of details involved, as I said with privacy. We can’t process private data and just do whatever with it. We can’t do everything with copyrighted material, even if it’s in the public. If a use is legitimate already depends on the details. And I think the same applies to AI. It needs a more nuanced perspective than just allow or prohibit everything.

I’m not that educated on US law and if everything is subsumed under Fair Use. I believe in Germany, we have a seperate rule for ephemeral copies during data processing and network transfers ($44a UrhG). So we don’t have to deal with that using a law that was more concerned with someone photocopying a book. And I believe some countries distinguish between commercial interests and non-profit research. Plus we have exemptions for example allowing someone to play music on their non-profit events, even without consent of the copyright holder. They still need to pay them a “fair” amount, but it’s not up to the copyright holder to decide… We specify under what circumstances libraries can use content, again differentiate between interests, and we have a rudamentary law concerning data mining for research since 2017.

I think some specific laws like that would be more suited to guide the issue with AI towards a healthy solution, than use one blunt tool for everything. Why not say AI training is allowed, but it requires a fair compensation? We could even have a standardized way of opt-in or opt-out… I’m not sure if we need that. But I’m fine with my blog posts and Free Software projects end up in some AI. But I don’t want them to listen in on my private conversations, like for example an Alexa could do… I believe that requires a law distinguishing between that. If everything is Fair Use, I can say goodbye to privacy, but at the same time cancel my Netflix and Spotify subscription, since I’m going to claim, I’m just collecting all of that for future AI training.

I -personally- think we can’t allow Amazon to spy on me and just claim it’s fair use. So context matters. And I also think the goal and nature of the AI matters. Research needs to be less strict than commercial interest. And I don’t think networking of digital devices can be handled the same way as AI training, I strongly believe that requires separate laws and also needs to factor in if there is some legitimate interest to begin with.

Those are great links. I think I already read Cory Doctorow’s post.

I think I already struggle with the premise. I think Google, Facebook, etc using my data is NOT Fair Use. They can not just publish my full name, pictures and texts without my explicit consent.

And this is kind of lumping everything together again… For-profit AI and open-weight models to the benefit of humanity aren’t the same thing. And I think we should give open-weight models some advantage by applying different rules. I.e. let people use data more freely if they contribute something back and the resulting product can be used freely as well. And make the rules more strict for big and closed for-profit services. And demand more transparency as well.

I mean realistically, we don’t have any proper rules in place. The AI companies for example just pirate everything from Anna’s Archive. And they’re rich enough to afford enough lawyers to get away with that. And that’s unlike libraries, which pay for books and DVDs in their shelves… So that’s definitely illegal by any standard.

But I agree that learning something from a textbook is a different thing than copying it. The resulting knowledge escapes the copyrighted material. I believe that’s the same no matrer if it’s machine learning or me learning computer programming with textbooks… The thing is just that you can’t steal in the process. That’s still illegal. IMO.

One of my fears is that AI is as disruptive as people think. And that the market is going to be dominated by unsympathetic big-tech companies, due to the nature if it. I think we need some good legislation to push AI in the right direction, or we’re going to end up in some scifi dystopia where big companies just shape the world and our lives to their liking.

Actually there is a small but important difference. Libraries usually don’t generate (much) revenue. They’re funded by public money, institutions… And a subscription is like 25€ annually or it’s free… While AI companies have billions of dollars turnover and they’re very much for-profit. And Fair Use has other applications as well. It allows science, allows me to record television or listen to music in the car or together with friends. I don’t think we can lump all of this together.

Now build a robot / autonomous (remote controlled) car / robo vacuum with that.

I went with the option with keyfiles. I keep backups of LUKS headers and keyfile in case a harddisk gets damaged. Currently have that stored as a file attachment in my password manager.