“The potential here is absurd,” wrote app developer Nick Dobos in reaction to the news. “Why write complex rules for software by hand when the AI can just think every pixel for you?”
“Can it run Doom?”
“Sure, do you have a spare datacenter or two full of GPUs, and perhaps a nuclear powerplant for a PSU?”
What the fuck are these people smoking. Apparently it can manage 20 fps on one “TPU” but to get there it was trained on shitload of footage of Doom. So just play Doom?!
The researchers speculate that with the technique, new video games might be created “via textual descriptions or examples images” rather than programming, and people may be able to convert a set of still images into a new playable level or character for an existing game based solely on examples rather than relying on coding skill.
It keeps coming back to this, the assumption that these models, if you just feed them enough stuff will somehow become able to “create” something completely new, as if they don’t fall apart the second you ask for something that wasn’t somewhere in the training data. Not to mention that this type of “gaming engine” will never be as efficient as an actual one.
I mean, you’ve never seen a purple elephant with a tennis racket. None of that exists in the data set since elephants are neither purple nor tennis players. Exposure to all the individual elements allows for generation of concepts outside the existing data, even though they don’t exit in reality or in the data set.
Try to get an image generator to create an image of a tennis racket, with all racket-like objects or relevant sport data removed from the training data.
Explain the concept to it with words alone, accurately enough to get something that looks exactly like the real thing. Maybe you can give it pictures, but one won’t really be enough, you’ll basically have to give it that chunk of training data you removed.
That’s the problem you’ll run into the second you want to realize a new game genre.
There are more forms of guidance than just raw words. Just off the top of my head, there’s inpainting, outpainting, controlnets, prompt editing, and embeddings. The researchers who pulled this off definitely didn’t do it with text prompts.
But at what point does that guidance just become the dataset you removed from the training data?
The whole point is that it didn’t know the concepts beforehand, and no it doesn’t become the dataset. Observations made of the training data are added to the model’s weights after training, the dataset is never relevant again as the model’s weights are locked in.
To get it to run Doom, they used Doom.
To realize a new genre, you’ll “just” have to make that game the old fashion way, first.
Or you could train a more general model. These things happen in steps, research is a process.
I know the input doesn’t alter the model, that’s not what I mean.
And “general” models are only “general” in the sense that they are massively bloated and still crap at dealing with shit that they weren’t trained on.
And no, “comprehending” new concepts by palette swapping something and smashing two existing things together isn’t the kind of creativity I’m saying these systems are incapable of.
Bloated, as in large and heavy. More expensive, more power hungry, less efficient.
I already brought it up. They can’t deal with something completely new.
When you discuss what you want with a human artist or programmer or whatever, there is a back and forth process where both parties explain and ask until comprehension is achieved, and this improves the result. The creativity on display is the kind that can unfold and realize a complex idea based on simple explanations even when it is completely novel.
It doesn’t matter if the programmer has played games with regenerating health before, one can comprehend and implement the concept based on just a couple sentences.
Now how would you do the same with a “general” model that didn’t have any games that work like that in the training data?
My point is that “general” models aren’t a thing. Not really. We can make models that are really, really big, but they remain very bad at filling in gaps in reality that weren’t in the training data. They don’t start magically putting two and two together and comprehending all the rest.
You keep moving the goal posts and putting words in my mouth. I never said you can do new things out of nothing. Nothing I mentioned is approaching, equaling, or exceeding the effort of training a model.
You haven’t answered a single one of my questions, and you are not arguing in good faith. We’re done here. I can’t say it’s been a pleasure.
My argument was and is that neural models don’t produce anything truly new. That they can’t handle things outside what is outlined by the data they were trained on.
Are you not claiming otherwise?
You say it’s possible to guide models into doing new things, and I can see how that’s the case, especially if the model is a very big one, meaning it is more likely that it has relevant structures to apply to the task.
But I’m also pretty damn sure they have insurmountable limits. You can’t “guide” and LLM into doing image generation, except by having it interact with an image generation model.
To be fair, half of the AAA gaming industry is all about trying to clone the latest successful game with a new coat of paint. Maybe using AI to make these clones will mean that the talented people behind the scenes are free to explore other ideas instead.
Of course in reality, it just means that the largest publishers will lay off a whole lot of people and keep churning out these uninspired games in the name of corporate profits, but it’s nice to dream sometimes.
A lower cut. 30% revenue cut means we pay more than necessary for games and we also miss out on some indie games that cannot be profitable with such a large cut.
We already know lowering the cut doesn’t make us pay less. All it does is put more money into the pockets of the publisher.
And I very much doubt Valve’s cut is a reason indie game can’t be profitable. There are asset flips going up on Steam on a daily basis. If asset flipping wasn’t profitable we wouldn’t see them propping up like mushrooms after rain. When asset flips are more profitable than an indie game there’s something wrong with that game.
Hell, Epic does not have any social features, didn’t have cart, refund process through support only, very basic search, I am not sure about cloud saves and if they don’t break completely when you play offline (is there even offline mode?).
Steam, on the other hand, is constantly adding and improving features - like the new beta family sharing which is finally what an easy way to share with my GF and sister.
The only things that Epic has are free games, exclusivity, and lower fees - and that’s about it. All three, as you can see, are not really hard to implement for the developer team, but easy to throw large sums of money at for a quick boost so they can boast numbers.
Fuck Epic, seriously. Money can solve lots of stuff, but not by throwing it at the wall. Meaningless.
Oh, completely forgot about my Steam Deck, it is just that seamless.
I also hate the other side of the coin that is against both Steam and EGS. Citing Steam doesn’t “deserve your loyalty”. Why not? I can’t really pinpoint any particular fuckup in the 15 years I’ve been using it. Sure, some delays in games, updates, and other minor shit - but imagine if like game ratings broke, I am sure they’d get fixed in an hour.
I like the fact they tried to compete with Steam from the begining. I have a large library of games and some real gems that I wouldn’t normally look at.
EGS is ok, GOG is ok and also Steam is just ok too for what I want from a store/launcher.
It would hold water if their solution was proprietary and closed source. But it isn’t, and anyone else, literally anyone, can take Proton and use it in their project for profit.
Even if they closed shop tomorrow, or even just gave up work on Proton itself, we’d all still reap the benefits at no cost to us.
Steam is just ok to for what I want from a store/launcher.
It’s not just ok, compared to the alternatives. A games library that cannot be matched, regular sales, easy no-frills refunds, cloud saves, beta support, family mode, big picture support, seamless integration with the Steam deck, which in its own right, has pushed right-to-repair and Linux gaming to new heights. The competition doesn’t even have any of this stuff, including the console market, and if they can’t compete, they don’t deserve my money.
No digital game store is worth your loyalty.
I’m fine being loyal to a privately-owned company that actually gives a shit about its customers. As long as Gabe is still alive and they will continue to be privately-owned, the company will stay in good graces.
It isn’t even loyalty for me, I just have to real reason to go to the other store with 99% of my games being on steam, mostly purchased during a sale. The only exception is GoG, because they actually offer something the others don’t with their DRM-less versions of games.
When that store is run by a company that contributes massively to open source and works harder and puts more money into enabling alternate platforms for gaming than all other companies combined; ya, they have my loyalty.
It’s false equivalence to claim steam has a monopoly when you’re literally giving epic a monopoly on your games for financial kickbacks between yourselves that in the best case doesn’t impact the user and worst case actively compells them to a much worse platform. What epic and gearbox did is monopolistic, what steam did is just make a good enough product that no one gives a sh*t about EGS. If you want an actual competitive store front, make something your users want, not your business partners. Gog is struggling but it’s still my first goto for games because even if it’s missing all of steams functionality, it gives me ownership of games that can’t just be revoked or broken by publishers. That’s a value add I’m willing to pay for. Paying more so publishers can make more money and sell a worse experience through EGS ain’t moving me.
It’s a little different to have your own games exclusively on your platform than to pay other devs not to release on other platforms, and it’s entirely different if devs just choose not to release elsewhere because no other store is worth the effort for them.
Steam did exactly that for years under the “Steam Greenlight” prism where users voted for games to be released on steam with the condition that they would be exclusive. They only stopped it when they decided to go the Amazon route and sell any old shit with zero curation instead.
And Tim Sweeny made the offer to stop offering Epic exclusivity and even sell their games on Steam if Valve offered to provide their service to developers at the same rate as Epic.
But Steam charges nearly triple what Epic does and can depend on gamers to defend them for some reason.
The cut taken by stores is of little concern to me as a consumer. Greenlight was a mess for a lot of reasons, but they discontinued it years ago, while Epic continues to pay for exclusivity deals. Steam provides lots of services to me that Epic doesn’t, though, as others have listed here. That said, I also like GOG and itch.io.
It makes the cost of developing games more expensive. They have to charge nearly 20% more for games on Steam to make the same money they do on EGS.
It’s also why Valve hardly makes games anymore. They sell 4 games made with other people’s money and they’ll have the same gross income as selling a game they paid to develop. Throw in the cost of development, and they just can’t justify game development as a major part of their business.
The last time they made a full-sized game was Half-life 2, which launched the same day as Steam.
It’s litterally a tunnel shooter with endless repetition to pad it out and pretend it’s a full game, when in reality it’s a tech demo to bundle with VR hardware and try and make Steam the default home of VR games.
This argument about cost of development would hold more weight if the game store savings were passed onto the users rather than just eaten up by the publishers. Borderlands 3 base game has the exact same price on steam vs EGS atm, £49.99. Clearly those 20% savings are just extra money the publisher wants to pocket rather than actual necessary costs to the game. If their happy to pass it off to steam when sold on the steam platform rather than raise the price to recoup the platform tax.
Yes, but with EGS more money goes to the company making the games. AAA games have never been more expensive to produce, and developers are shutting doors left and right. After the costs of marketing and overhead, more of the proceeds of the game are going to the fucking download service than the people making the game when it’s on Steam.
Blame the publishers then. They set the price and they dictate the bonuses of the devs based on sales. Choosing to believe more money from the game store is actually making its way to devs instead of shareholders is naive at best.
And Tim Sweeny made the offer to stop offering Epic exclusivity and even sell their games on Steam if Valve offered to provide their service to developers at the same rate as Epic.
Tim Sweeny didn’t make an offer, he tried to make positive PR to EGS while trying to paint Valve as the bad guys; Valve obviously wouldn’t charge the same rate as Epic because they include a lot more value for both user and developers than Epic does: to list a few of Valve services that Epic doesn’t have:
Steam Workshop (hosting terabytes of content for absolutely free);
Family sharing;
Steam Link for game streaming;
Remote Play Together tech for all the major OSes;
Linux and Wine/Proton investments (which you could argue was an investment because of the Steam Deck, but that’s an investment that benefits everyone, regardless of whether they own a Steam Deck or not);
Cloud save hosting;
Universal controller remapping interface compatible with all the major gamepads;
That’s not to mention the benefits developers can get from Steam’s platform and SDK:
Steam Input (for not needing to deal with custom implementations);
Steam Voice API (for in-game voice chats);
Steam Inventory and Trading Cards, which can result in extra cash for the developers;
Multiple networking options: Steam Game Servers, Steam Matchmaking & Lobbies, Steam Peer-to-peer Networking, etc.
If you ask me, I think Epic is the one charging way too much
The entire greenlight catalog was exclusive. That’s over 100 third-party games, and they only reason it stopped is because they stopped curating products to become the Amazon of online gaming.
Sometimes I wonder if these people understand that no player ever wanted exclusivities on a game store. Instead of providing a decent service, they’re litteraly trying to kidnap customers with a choice between waiting for months for this big release or taking it on a subpar platform.
This is my current dilemma with the new Star Wars outlaws game. Epic has exclusivity on release (or can buy direct on ubisoft), but I have 29 other Star Wars games all on Steam. Do I really want one odd game on a different platform, or do I just accept that I won’t be playing it at release and wait the months for it to come to Steam?
arstechnica.com
Gorące