harrisoned

I'm not particularly into VTubing (not a fan of social streaming or the japanese style), but i find the tech and creativity amazing. I'm very into mocap and started building an experimental facial capture system, which overlaps a lot with the VTubing concept, but using different protocols (most VTubing models/solutions uses VMC) and more varied custom cartoon models.

If you are into the tech behind mocap this community is a fun place to explore.

Mandatory ID verification at software level for local LLMs is clearly the solution here. /s

I hope I'm being pessimistic about that. If I'm wrong, great.

I did saw it was an extension of the 1.0. To my understanding, "allows to scan" can also mean enabling a pipeline for accepting requests of scanning lawful content because, well, they can. In practice, it creates a mechanism to crawl people's information because when they feel like. While the law do make it explicit that this 'allowance' should not be used for anything else outside of the scope, i can't trust they won't. Once the mechanism is there, and that is valid for other countries, it might be used for stuff outside this scope since it's possible.

I work in the ISP field, and this happened in another context. First, a pipeline was built to block sites without a judge, because doing it only with court orders made it to difficult. After a few months of many ISPs complying the scope grew, now they can target piracy websites at will, and you must comply. Why stop there?

My fear is that this sets an example. I hope I'm wrong, but i don't trust them. There's a reason it was rejected before and it is being passed like that now.

Centralized messaging services won't last long, their capture is sadly inevitable. In the long run, only self-hosted/decentralized protocols can resist what's coming.

Very much. I also fear they coming for this, we already have instances of where using secure alternatives tags you as a criminal[0], so i don't doubt a future where non-approved applications will get you in trouble. With everything happening around Android locking itself down[1] and Windows being a spyware[2] anybody who wants privacy will be 'different', and can be tagged and excluded from parts of society for not using the same services.

[0] https://x.com/GrapheneOS/status/1940440326830989549

[1] https://news.ycombinator.com/item?id=48801059

[2] https://news.ycombinator.com/item?id=48815196

Even if you are not in the EU, this will affect you. Some countries really like to copy such regulations from others. Once services starts complying, other governments will go like "if you did for them, you can do it for us, right? so it's not technically impossible", and things only get worse from there. Not all services will simply block the EU as well, which would be better to send a stronger message if approved.

I really fear where this is headed.

"PIX will be discontinued today"

In the late 2000's/early 2010's that started to pick up where i live, and it was referenced (even by the government itself with it's social programs) as 'digital inclusion', where technology products where subsidized and made more accessible (dumbed down) to the general public, for the reason you mentioned on your third paragraph.

While this was in a time when PCs where the main and only thing for most to access the internet, in my opinion this problem really took shape as smartphones got popular. They where shaped to adapt to a low-level of tech literacy, and lock you in instead of actually providing knowledge to the user. Facebook then went in, and then Whatsapp. Whatsapp was a fun one because carriers made it free of charge, the push to the masses was insane, and nowadays you are questioned and looked at weird if you say you don't have a Whatsapp account.

That trend is what broke most of the social interactions and services on the internet for me.

I have been working for some time on a budget body/facial mocap solution with Unity. Mocap is hard, and what exists is locked behind subscriptions or is just very expensive.

With Unity I'm trying to bundle a bunch of different free, cheap or open source solutions together. For facial, that includes a custom converter from the output of Deadface (based on Mediapipe) with ARKit blendshapes, and also eye movement. For body it's a custom hook to SlimeVR that allows you to mocap with cheap-ish IMU-based DIY trackers, and all that on top of a custom made (not free but open source) physics rig solution that gives you accurate rigid body real time collision, saving on cleanup work.

It's being going really nice despite being an unusual workflow. Hope to release it as a plugin for a in-development sandbox game in the near future. Mocap and animation has been my passion long before i started with tech stuff, and finally I'm able to pursue it.

"Ah, smoking is not good for you, and it's been deemed that anything not good for you is bad; hence, illegal" — Demolition Man, 1993

For you, no. For the services you depend on and will continue receiving your data, and may jack up the prices/add limitations knowing that your dependency won't be easily broken, yes.

All this look so dystopic to me. Even without assuming all those are real (which i doubt), i have heard similar stories from friends and others. The level of dependency people are getting from those services is surreal.

I was thinking the other day, "since social media is kinda wearing off, could 'LLM As A Service' be the new addictive thing for the masses?" because i'm hearing horror stories of people who are outsourcing their brains, in some cases their feelings, to those services, and i personally saw a case of a 'high level professional' asking an LLM how it should respond to somebody in real time during a Whatsapp conversation. It is in fact a drug, and it tricks you very well into thinking you should rely on it.

Also when reading this piece (https://news.ycombinator.com/item?id=47790041) earlier, i thought about it again. Nowadays instead of searching for something and being forced to learn, those services spoon-feeds contents of dubious accuracy for everybody, which will not only cause trouble for them eventually, but also creates a stream of revenue based on people's cognitive laziness, to not use harsher words.

Social media is/was bad and it relied on a similar mechanism, but i feel this is much worse. People crying as if their brains where took away is proof of that.

This is tiring. The text is so vague, and if a big country adopts it software companies will comply, and there's no reason to why smaller ones wouldn't, since 'the work is already done'.

I wonder if it would be illegal for an user to use an outdated system without those functions when they roll out, or to use outdated applications, or to distribute outdated applications, or to keep mirrors of multiple versions of operating systems. I doubt they thought that far, or if they care at all.

There should be more noise about this here, but to whoever you talk about that issue they don't seem to grasp the situation, or simply don't care, and call you crazy/paranoid. I have been told you also need the GOV app for certain things related to companies.

Those require a phone for you to send messages and interact. It will ask you to 'Verify phone', but you can chose not to and stay on the server as read-only, Discord itself won't bother you about it. I am on a few like that for quite some time.

Same for me, and my account is almost a decade old. I think it depends a lot where are you from and the kind of activity, as i read stories of people being asked to register a number out of nowhere. Many servers requires you to have it tho, due to spam protection. I just don't talk on those.

This is very good, but I'm surprised the term "game-changer" is not mentioned there. From my observations this is used a lot in LLM texts.

They use Perforce.

https://developer.valvesoftware.com/wiki/Using_Source_Contro...

This is something that worries me. I know that the laws/constitution that guarantees the rights of somebody may vary from country to country (and may not even be enforced by the letter), but lets say: All commercial companies will have a ToS, data sharing agreements, etc. You, as a user, i assume is not obligated to agree to that ToS at the expense of not using the service. If a government body requires you to use their service to access basic services (and offers no 'offline' alternative) required by law, are they, by proxy, coercing you to accept a commercial ToS? I would very much like to hear a lawyer opinion on this.

I know some government may do this with intent, but i imagine many governments simply never thought about it, or no citizen ever didn't accepted a "popular smartphone OS provider's ToS" and challenged that government requirement. I know some make offline alternatives very inconvenient, but that still technically legal.

That's exactly what i think it is. Have a legion of users willingly scraping the internet for you, going behind captchas, logins, and all the mechanisms that was put there to stop bots. With this every user user of the browser is also a bot. Now i wonder if the agent string will be something unique, and certain places will just block those browsers from their websites by it.

That's something that drives me crazy as well. I don't actually use the big 'algorithmic social media' sites, only Telegram and Discord mostly, and seeing screenshots/memes with those words censored there made me wonder why, at first. Then i saw people auto-censoring themselves in those places where there's no such thing as algorithmic de-ranking. The social media generation already find it normal, acceptable, and is specially ironic to me that a lot of people who are vocal against those services have conformed to what they say to stand against.

That behavior also highlights how people within those services care so much about reach, clout, 'going viral', instead of communicating with other people.

At present the project is focused on mobile platforms, specifically Android and iOS, as they cover the vast majority of users and real-world use cases. (..) Desktop support is not currently within the project's scope.

This is the equivalent of a "Do you guys not have phones??"[1] but on a way larger scale.

At least where i live i am able to use the bare minimum of phones, even working with tech. The friction is increasing though, which worries me a lot, and day after day there is a new attempt to shove it down your throat if you want to be considered a member of society. Seeing that a lot of countries (including mine) are pushing for age verification, and the whole thing about Android blocking 'sideload', by the end of 2026 you won't be considered a human being without a government certified smartphone.

[1] https://www.youtube.com/watch?v=ly10r6m_-n8

So this answers my question about not actually having social media. In theory you wouldn't be denied. But as a professional in the field who cares for privacy, and simply has no use for such services, i wonder if they could just assume you are lying and has bad intentions.

I noticed that same behavior across older Gemini models. I build a chatbot at work around 1.5 Flash, and one day suddenly it was behaving like that. it was perfect before, but after it always saluted the user like it was their first chat, despite me sending the history. And i didn't found any changelog regarding that at the time.

After that i moved to OpenAI, Gemini models just seem unreliable on that regard.

I agree its an amazing payment method, it worked for me for most of the time. Still, we depend on bank's stability and technical availability for it to work. Once i needed to pay for something and forgot my card at home, at that same time my bank was going trough technical issues and i couldn't pay.

Despite rare reliability issues, my fear about it is that it requires a phone. Being so popular, i fear when places will refuse any other form of payments and accept only PIX, making anybody not using a phone unable to buy their products, with the common assumption that everybody uses it ("don't you guys have phones???"). You can't install banking apps on rooted phones or alternative mobile OSs (or is very very hard), so you are trapped with Android or IOs to use it.

I hope it doesn't come to that, but it seems it's going that way.

It is. A friend of mine has this. It took her so much time to actually seek help, despite the excruciating pain, and she got part of her digestive system removed because it spread too much.

Nowadays, after 2 years of the surgery, she manages it with a restrict and healthy diet, and the pill. It takes a big toll on the well being even after the pain is gone, and she is almost always tired because of it, the body is constantly fighting the inflammation.

I have a big issue with this, and the truth is that the majority of people simply do not care and/or do not understand the implications.

By tying your service to a smartphone your are basically refusing to provide service if the costumer doesn't agree to Apple's or Google's TOS. If the app doesn't complain about emulation or something different than Android or IOS you are in luck, but that's not the case with most banking apps. And that's only talking about people who don't have it by choice and have money to buy one.

For me, once, it went beyond: I took my first dose of the Covid vaccine, and the second dose's date would still be announced. I asked where it would be available to the nurse, "On the Instagram page of the <local health body>". "But i don't have Instagram" i said, and the nurse shrugged. It requires both a phone and a social media account with your real info, but since absolute nobody complains about it they just do because it's easier.

This will continue as long people are complacent with it. In some places the government is required to provide you services, by law, by any means available and not depending on 3rd party service, but they do require apps anyway and people stay quiet. Phones as an alternative is fine, it's a tool, but should not be an obligatory device for you to be considered an human being.

I agree with that. At work, we are about to implement a decent LLM and ditch Dialogflow for our chatbot. But not to talk directly to the client (it's asking for a disaster), just to recognize intentions, pretty much like Dialogflow but better.

Right now there are many small but decent models available for free, and cheap to use. If it wasn't for the hype, it would never have reached that level of optimization. Now we can make decent home assistants, text parsers and a bunch of other stuff you already mentioned.

But someone paid for that. The companies who believed this would be revolutionary will eventually have a really hard reality check. Not that they won't try and use it for critical stuff, but once they do and it fails spectacularly they will realize a lot of money went down the drain.

It certainly has use cases, just not as many as the hype lead people to believe. For me:

-Regex expressions: ChatGPT is the best multi-million regex parser to date.

-Grammar and semantic check: It's a very good revision tool, helped me a lot of times, specially when writing in non-native languages.

-Artwork inspiration: Not only for visual inspiration, in the case of image generators, but descriptive as well. The verbosity of some LLMs can help describe things in more detail than a person would.

-General coding: While your mileage may vary on that one, it has helped me a lot at work building stuff on languages i'm not very familiar with. Just snippets, nothing big.

I'm curious about the ToS. As far as i know, their version of ChatGPT on Copilot don't like +18/sexual stuff, so how would the Copilot+ react to a lot of it? Would you get banned? would it work? Would it simple ignore all the stuff when asked for it?

I did that with Llama 3 8B with some stuff i could think of, and it did very good. It was on par with GPT4. I prompted some scenarios and asked it to use CoT. Scenarios like "i was standing and eating chocolate, and it melted. Will i find chocolate at my feet?", and the reasoning was pretty good.

But there was something it did way better than GPT4. I asked to create 10 phrases where the last word was an animal, excluding equines, and in alphabetical order. GPT3.5 and GPT4 aren't able to follow such instructions, but the 8b model did it with maestry.