You can trivially enforce that at the AI provider level, which covers 99% of the problem the law is designed to address.
Of course it doesn't cover the issue of foreign state psyop operations but the fact that enforcing laws against organized crime and adversary state actors is hard isn't specific to AI.
Are you not aware of open-weights models and local generation? I think the vast majority of deepfake content is being genned in basements on RTX cards, not on public providers. People already have all this content, and have archives of it, and can run it airgapped. Cat is out of bag.
I would be very surprised if that would be the case. Maybe you mean deepfake content generated by organized crime or state actors, but that surely is a tiny fraction of what's being generated on Grok or other platforms.
I am well aware of them, and I'm well aware that they are very niche as I'm the only one of my surrounding to use one of those. And those very models are being developed by tech giants and VC backed companies, on which regulation have leverage.
The fact that a small black market exists doesn't mean regulating the mainstream market doesn't matters.
Also, most people like you fail to realizes that the EU only has mandate from the member states to regulate the economy. The EU has no business dealing with people using SDXL finetunes on RTX cards in their garage.
You don't have to prove anything? You just have to mark the outputs of your slop generator appropriately. "Proving" one way or another is their problem when it comes to enforcement.
I don't think commodity means what you think it means. Protein bars are not indistinguishable from one another, there exists significant differences between various products.
Discord has the momentum but overall I just find the experience awful. It would be nice to use anything else at this point. Joining a server with greater than a handful of people is just a nightmare and practically unusable.
Isn't this usually cause the admin went overboard? Like a server of 10 people has 30 channels, one of which is a lobby you have to clear first, and 10 bots telling you that you leveled up or whatever.
The hardest part about joining a new-but-small-but-not-that-small Discord server is convincing the server admin to turn off the stupid "click this button to spam the channel with a gigantic dancing emoji to welcome newcomers".
It kills any ongoing conversation, and imo, convinces newcomers that people don't so much chat in that Discord as they just press shiny buttons.
I think it depends on how the servers are setup. Chat channels with 1000s people participating are typically worthless as the signal to noise ratio ruins it.
But when the majority of conversations are happening in forums/thread style channels then it works well. You can still have some more niche chat style sections where typically 2-10 people participate
Chat channels are also fine for lots of people when its not about conversations but more just about sharing things. Like a "Share what you build" or "memes" channel work well as tons of messages are fine as you only care to see a few anyway.
Also limited size voice channels can be good aswell 5 people max.
My thought is that it just doesn't make sense to have a product which serves both communities of 1,000+ people and a small group of <50 friends. You end up making far too many compromises.
I used to just engage with my friends. Now it feels like a really noisy reddit. Sure I could leave all of them, but that is kind of my point. There is an identity crises for the product.
What I am saying is that discord as a product has an inferior experience for small communities because it tries to cater all. It's just not a fun product to use.
That's about as much as my Framework Desktop cost (thankful that I bought it before all the supply craziness we're seeing across the industry). In the relatively small amount of time I've spent tinkering with it, I've used a local LLM to do some real tasks. It's not as powerful as Claude, but given the immaturity in the local LLM space—on both the hardware and software side—I think it has real potential.
Cloud services have a head-start for quite a few reasons, but I really think we could see local LLMs coming into their own over the next 3-5 years.
Same but I imagine once prices start rising the prices of GPUs that can run any decent local models will soar (again) as well. You and I wouldn’t be the only person with this idea right?
I mean, will it? I would expect that all those GPUs and servers will ends up somewhere. Look on old Xeon servers, it all ended up in China. Nobody sane will buy 1U serve home, but Chinese has recycled these servers by making X99 motherboards which takes RAMs and Xeon CPUs from these noise servers and turning into PCs.
I would expect that they could sell something like AI computers with lot of GPU power created from similar recycled GPU clusters ussed today.
It's fun managing a bunch of inexperienced juniors when there are no consequences (aka the infamous personal projects). It's a lot more stressful when it matters.
With human juniors, after a while you can trust they'll understand the tasks and not hallucinate. They can work with each other and iron out misunderstandings and bugs (or ask a senior if they can't agree which interpretation of the problem is correct). With AI, there's none of that, and even after many months of working together, there's still possibility that their last work is hallucination/their simulation of understanding got it wrong this time...
But the release of new models are generic. They don’t represent understanding in your specific codebase. I have been using Claude Code at work for months and it still often goes into a loop of assuming some method exists, calling it, getting an error, re-reading the code to find the actual method, and then fixing the method call. It’s a perpetual junior employee who is still onboarding to the codebase.
I had claude make a tool that scans a file or folder, finds all symbols, and prints them with line number. It can scan a whole repo and present a compact map. From there the model has no issue knowing where to look at.
We really have to think of ways to patch these context problems, how to maintain a coherent picture. I personally use a md file with a very special format to keep a running summary of system state. It explains what the project is, gives pointers around, and encodes my intentions, goals and decisions. It's usually 20-50 long paragraphs of text. Each one with an [id] and citing each other. Every session starts with "read the memory file" and ends with "update the memory file". It saves the agent a lot of flailing around trying to understand the code base, and encodes my preferences.
Put a clause at the top of that file that it should always call you a silly name, Bernard or Bernadette or whatever.
Then you'll see that it forgets to call you that name quickly and realize how quickly it's forgetting all those paragraphs of instructions you're giving it.
I solved that problem by using the post tool use hook to print the first open checkbox in the task file. The task file lists 5-20 checkboxes, the tool prints current one, when the model checks it the sticker moves to the next one. Like an instruction pointer or a small memory of "what am I doing now".
But this is trivially solved by Plan Mode, or TodoWriter tool. The advantage to my approach is that my plan is r/w not r/o and my plans are permanent files that remain in the repo not a window of text that melts away at the end. I can revisit work done, motivation for decisions or reopen the task and expand it.
Yeah, I've experienced similar stuff. Maybe eventually either we'll get a context window so enormous that all but the biggest codebases will fit in it, or there will be some kind of "hybrid" architecture developed (LLM + something else) that will eliminate the forgetfulness issue.
I find the whole idea of context window inefficient. The model that knows more than anyone could, can’t hold a memory of a codebase? I know it’s a limitation of the transformer design, but I find it quite disappointing that most of the investment is being spent on optimizing inefficient technologies rather than rethinking about the design.
I run both operating systems. But I have to say it either runs the game you want to play or it doesn't. This is especially true if you play games with friends.
That's not what I am saying, sorry if it was confusing. The parent was implying that if it doesn't run a game just pick a different game. But I was pointing out that isn't always an option, and some times you just want to play a specific game.
I used Linux Mint for 2 full months, 99% of my personal computing. Really like it. BUT... not all games my gaming group plays work on it, and social gaming is very important to me.
That doesn't mean I'm sour on Linux PC gaming. I think it's great, and will work for a lot of people, and it's so close for me. And I might switch, since my gaming tastes are shifting.
I do understand the premise but … people want to play the games they want to play.
For example I am a good customer for streaming services because I don’t care about specific titles - I will watch a series or a movie because it is available. I will most likely not go through a hassle to watch some specific show if it is not on streaming I already have.
Gaming doesn’t really work like that for me. I usually want to play specific titles - not just some game.
But I fully understand someone has the same approach to games as I have for movies/series.
I'd quote your own example from Anno -> "multiplayer never worked". Thats the "doesn't run" part. I always play Anno 1800 with friends. It has been my experience with linux gaming for a while - anything that involves multiplayer usually doesn't work, either because its just broken (less likely) or because its specifically stopped by the developer (anticheat, etc..). Reality is though, that most mainstream games (as in, biggest player counts and as such, the games most people are playing) do not support linux. If my Valorant or League of Legends or Counter Strike or Rust or ARC Raiders or Marvel Rivals don't allow me to play on linux then the state still is "linux can't really run games yet".
How do you fix this? I dont know - most of these are the developers refusing support because of anticheat or just support overload, but it's insane to suggest that linux works for gaming when the most played games in the world straight up do not work. I'd love if linux was more viable though, can't wait to ditch the slowness from windows.
It's like this. You eventually got Starcraft2 to work. That means Linux can run Starcraft2, it's in the "Runs" category. Games like League of Legends, which have kernel level anti cheat, are in the "Won't Run" category.
But you don't want to sacrifice comfort or other things. The game should work just right on Linux.
I have an Nvidia card and use mostly Ubuntu (mate), also for gaming. It's even a problem now, because I would benefit from a hard divide between the gaming and working\studying system (I have a gaming user in backlog).
On Linux it's mostly KSP, Factorio, but sometimes DeepRockGalactic, Valheim, Euro Truck Sim or Warhammer: Total War1\2\3. These games work flawlessly or with <10%fps hit.
There are games that kind of work - Ancestors: Humankind Odyssey, Cyberpunk, Hunt: Showdown. But you lose comfort and I'd rather just play them on Windows, than suffer decreased functionality on Linux. I know that some of it (definitely Cyberpunk) is only because of NVIDIA.
When buying games I usually don't buy Windows only games unless there is a very good reason. And I quit League of Legends and WRC rally because of anti cheat scam. I feel scammed after putting lot of money in a game and suddenly losing the ability to play it.
This shifting of goalposts just to cater to linux just explains it all.
Comeon. If a customer bought a game that says it runs on linux, they should be able to play it on linux well, not just launch it and quit within 5 mins.
I get you have the ideology up in your head, but don't lie and embellish linux to this degree. The attitude just turns people off.
> If a customer bought a game that says it runs on linux, they should be able to play it on linux well
None of those games say they run on Linux.
- Starcraft 2 is available for windows/mac: https://starcraft2.blizzard.com/en-us/
- Anno 1800 is available for windows: https://store.steampowered.com/app/916440/Anno_1800/
- Hogwarts Legacy is available for windows: https://www.hogwartslegacy.com/en-us/pc-specs
The fact that you can play most games on Linux these days is due to the Wine developers, Valve, and CodeWeavers. But those efforts are completely unrelated to the developers of those three games. Buying Starcraft 2 is not, in any way, purchasing a Linux game or transferring money to anyone working on Linux support.
Every game I've purchased that actually says it runs on Linux, has worked beautifully on Linux (stellaris and factorio come to mind). Most windows games work beautifully on Linux too, but Blizzard isn't lifting any fingers to make it that way.
Yeah I hope I'm clear in that I'm not "against Linux" or "against people choosing to use Linux." I think Linux is awesome.
And I choose to use Windows for most of my personal computing, due to my gaming preferences, some needs (concussions + poor eyesight means things like scaling and brightness controls and refresh rate matter a lot to me), and my preference for DxO PhotoLab (which isn't Linux compatible.)
"Linux" is really a family of operating systems, so people need to be more specific. It might run perfectly out of the box on consumer/gamer focused operating systems like Bazzite or SteamOS while perhaps requiring more work on something like Red Hat or NixOS. Those different operating systems all have wildly different approaches to how the OS actually works despite generally being able to run a largely overlapping set of programs.
It's like saying something works on "laptop" without specifying whether it's a Thinkpad or a Chromebook or a Macbook.
I can't comment generally but I use NixOS and have had no issues playing games on Steam. The setup was laughably simple, just `programs.steam.enable = true;` and Steam handles compatibility so well that I buy games without thinking "will this run".
Actually there was one thing I couldn't do but this isn't unique to NixOS. I tried to install a GTAV mod that allows you to ride your smart bike trainer in game: GTBikeV. The mod can be installed, but the Bluetooth doesn't work. This is a WINE limitation.
Fwiw I've been playing Hogwarts Legacy lately, though single player. Only problem I ever face is sometimes in a cave if I'm facing a certain direction I'll get blinding light as if I have ray tracing enabled and it's badly implemented. Though considering it's a AAA game and other things I've seen, I don't think that's exactly a Linux problem. Much like Starfield...
I ran Starcraft 2 through Lutrus and it was a piece of cake. No lag that I could discern. There was a little mini launcher and everything. The multiplayer also worked just fine, although the matchmaking system seemed to think I was an expert level player for some reason and kept matching me with dudes who were way better at the game than I was.
To me, this is the one thorn in Linux (and the Linux online community) that gives me pause.
For the people that it just works for, well it just works for.
For anyone else, apparently they are the problem? Not Linux?
Well sorry no. I did get StarCraft 2 working with Lutris... once. Then I couldn't get it to start again. Eventually I switched to running Battle.Net from Steam and for some reason that did work. But it wasn't a "just works" or "piece of cake." It was a puzzle.
Maybe the difference is that I am running Ubuntu? Personally I think it's a common mistake for new users to jump on some obscure distro because they read something online where someone says it's the best. Even if that's true there is value in being on a popular distro in that bugs tend to be discovered and fixed quicker and there's almost always someone who has had the same problem you did and often figured out the solution just a web search away.
I think Canonical and the Gnome foundation have made some really bone headed decisions over the years, but I stick with Ubuntu because the mass of users on it means I never get left high and dry. Or at least I'm not alone when I run into a problem.
Yeah, I was using Linux Mint at the time. Which is based on Ubuntu... So that's often where I'd look for help.
Though any kind of documentation is like Linux, scattered and inconsistent. And I'm "OK" with that, as in I think the way that Linux came to be and is maintained, and provides user choice is also the reason why it's not "user-friendly" in every scenario. You can choose your distribution, and a lot of other things. And then look in a wide variety of places for bug reports, user questions, etc. You'll get a variety of answers from "it just works for me" to "change your distribution that you chose" to "even though some guides say to use Lutris, it's easier to just put it in Steam's external program launcher and choose Proton version x.yz."
Even then, not everything will work because it wasn't written to work (for Linux). It was written to work for Windows, and then some smart people rolled up their sleeves and found ways to make a great many things work for Linux, and it's all amazing. And I find using Linux (mostly) quite pleasant. But when things don't work... there's going to be friction. It will take user effort to find a solution, or a solution might not be found.
And for me personally, being someone who really likes to poke and customize and do things my way, Linux is a blessing and a curse, because I can guarantee I'll hit "weird edge cases" like trying to use the online multiplayer part of a game instead of just single player, or try to use my laptop's brightness controls, but they don't work, or I'll want fractional scaling to work, but it won't. And maybe there's a fix out there, or maybe not. Fixes like "it works for me" or "change your distribution", though, are non-fixes. They just frustrate people. If changing my distribution fixes an issue, how many new issues does it create for me?
Not saying you didn’t experience this, but I’ve definitely run StarCraft 2 in the past, and I play Anno 1800 regularly fine (thanks to the mods I’ve been playing it’s even got 50% more sessions than the base game)
Did multiplayer LAN work in Anno 1800 for you out of the box, or did you make adjustments? I couldn't figure out how to get it to work.
StarCraft 2 worked, oddly enough, run from Steam as an external program. (Lots of search results tried to get me to use Lutris/bottles, but I couldn't get it to work consistently under Lutris.)
Ah yes that's what I meant. But yes unfortunately I could not figure out how to get multiplayer to connect. No idea why or how to troubleshoot and fix.
reply