It also means that his praise for and indeed positive experience with this “tool” is largely irrelevant to the broader social/economic discourse surrounding so called AI.

Yeah, but tbf all of the statements he’s making about it are in the context of organizing linux development and the most effective way to carry that out, and just happen to be public and so end up being widely republished and incorporated into discussions with a broader topic than that.

That does change things a bit

I don’t see how this means he’s wrong though, if people are hating AI as a proxy for hating their abusive relationship with their employer, that doesn’t seem like a strong argument that everyone who isn’t in that situation must also hate AI.

For now, age verification laws are being passed all over the world though

I feel like most people won’t want to buy it if they look at this page though lol

For streaming/discovery, custom cli application that scrapes Bandcamp ‘recommended’ albums to get associations between them and considers what I’ve liked and disliked for probability weights determining what to play next, playing them with mpv.

For actually downloading stuff I like enough to want to save, I have a dedicated button to take the url in my clipboard and save audio files from it to my music folder with a script that uses yt-dlp

I had some bad luck with buying a SAS harddrive on ebay, bought like 3 in a row from different sellers that turned out to be DOA. To be fair, I was going for the lowest cost I could find for the capacity I wanted and ignoring seller reputation, and I did get full refunds after returning them, it was just a pain to keep installing broken harddrives, documenting how they were broken, repacking them and making trips to the post office. The fourth I bought used from Amazon for a little more money and that one worked, I noticed it was also packaged better, in its own harddrive shaped styrofoam thing instead of crammed in a small box with a thin layer of bubble wrap. The whole elapsed process took months. If I had to do it again I’d probably try more non-Amazon sellers again and try to find one that does packaging better and doesn’t try to sell broken shit because I hate Amazon but that is how it played out.

That’s your own addition though, it isn’t entirely clear that the person you questioned was talking about issues separate from its power structure.

Company policies aside

That’s kind of by far the most important class of problem though. I could complain about the complexity of its notifications system, or how all the ‘servers’ are represented with image icons and it sucks to navigate and find the one you’re thinking of if you add a lot of them, but it’s all small potatoes compared to being run by an enshittifying company complying (often in advance) with terrible laws and moving to use their accumulated leverage against their users. By developing a reliance on their service you put them in charge of you, but they cannot be trusted with that.

Now what?

I was very disappointed that the rest of this paragraph didn’t go in a completely different direction

But piracy demand also indicates to companies an untapped market that they could tap if they create the right product

I agree with your conclusion of not watching it, but the reason that companies might observe and take a non-purchase as some kind of hypothetical market signal doesn’t seem like the strongest argument for that, especially if people are avoiding the purchase specifically to deny them money for ethical reasons, that’s not something they can resolve by making ‘the right product’.

If pirating a Harry Potter show provides support to its author, I think that support is going to be more in the form of cultural relevance and mindshare than anything directly financial. Being a fan of an IP gives other people a reason to engage with it because your interest signals things to other people, like it means that by not watching it they might be left out of the loop etc, and since the law enforces property rights over what can be done with it and who can profit, that adds to the power of its owner by influencing the choices and interests of other people in their favor.

It seems that accounts aren’t required to download models or participate in the torrent swarm and have a somewhat reasonable justification of needing a way to maintain model provenance, which afaik pirate sites also require accounts for uploads for similar reasons:

Hugging Face verification confirms that someone claiming a creator’s handle controls the matching HF account or organization. It helps prevent impersonation and preserves attribution as models spread beyond their original host.

It’s also reasonable to have a centralized API since a lot of software in the ecosystem has using the huggingface API as a default way of doing things, so that makes sense for the goal of resilience to huggingface getting locked down somehow in the future.

The gamification-ish features do seem a little contradictory with the stated “decentralized infrastructure” objective though:

Pirate Face is preparing community benefits for account holders, including free compute credits and exclusive model releases.

Also loses points with me for (as far as I can see) not publishing a code repo, and the writing on the website having a very AI generated style which combined with the goals and features being a bit all over the place makes it tough to get a read on the project’s ethos and intentions.

Cops in the employ of organized crime are not exactly the people you would want having backdoor access to everyone’s phones lol

I had a report for AI slop today. 1m LOC in 3 months. How do you even modify and distribute that?

That makes some sense but does “AI slop” here refer to code with this type of problem or all code made with AI assistance?

Even the low tier models have seen huge ongoing improvements, both much smarter and much faster, on the exact same consumer hardware. Great news for anyone who doesn’t want to see these big US corporations form an AI compute oligopoly.

Given the Snowden leaks, it seems inevitable that if that data is being recorded and collected into the servers of a small number of US companies, that it will end up being used to enable nationwide searches regardless

Not sure. For the sake of speculation, one possibility is that the maximum liability to Meta is billions of dollars, which they can clearly afford, but if it targeted an individual that individual could declare bankruptcy as is common in high award amount piracy lawsuits against individuals (and maybe they pay him off under the table afterwards). Maybe he just didn’t think about it too hard. Maybe the risk of such a lawsuit succeeding is actually low somehow. Strike 3 seems to really want to tie it in to their existing case against Meta, probably to bump up the number of infringing works they can include:

Last summer, the company took aim at a much larger target. Together with Counterlife Media, Strike 3 sued Meta, accusing the tech giant of downloading thousands of its films via BitTorrent to train AI models. With 2,973 films at stake, the case could be worth up to $446 million

Maybe, although it looks like the company is trying to say they didn’t do it

However, even if the subscriber is who Strike 3 claims, Meta argues that nothing ties the downloads to the company.

“They simply assume that the Subscriber is responsible for everything done on the IP address, and then leap to the additional, similarly unsupported assumption that he did it for Meta,” Meta writes.

So presumably the executive himself would not be able to use an “it was for AI” defense since they don’t even want to admit it

But it cannot detect heart disease. That requires an echocardiogram, a type of ultrasound scan, which patients often have to wait months for.

Now a team have developed an AI tool that can spot signs of heart failure and heart valve disease – two of the most common forms of heart disease – from ECG results in “the blink of an eye”.

The tech cannot be used on its own to definitively diagnose or rule out heart failure or heart valve disease, but it gives a very strong indication someone may have them.

Someone judged as highly likely to have either could be sent rapidly for an echocardiogram, rather than wait months on the standard waiting lists. That could mean a quicker diagnosis which would enable them to start treatment earlier.

Maybe it’s just pessimistic of me but I could also see this being used to make it harder to actually get the definitive test if the AI doesn’t notice your heart problems on the first pass, like the insurance won’t approve it or something like that

looks like types of food additives

E407 E471 E472