Tim Sweeney claims it’s a “Scarlet Letter” which makes players “try to kill the game”
Epic Games CEO Tim Sweeney has criticised rival Valve for forcing studios to disclose when they use AI in game development.
Epic recently showed how it was integrating AI into Unreal Engine 6.
Time Sweeney said:
“If you want to launch a game, and get it as widely publicized as possible, you’ve got to put it on Steam so people can wish list it, and if you want to play it on Steam, then you have to get this Scarlet Letter of AI attached to your product, and now there is a hater community trying to kill the game.
“I think it’s really irresponsible of Valve. They shouldn’t do it, because it makes it much, much, much harder for a game developer to have a chance of success. You have to choose from either not using tools that can make you way more productive, and probably failing due to competition that does.”
Which is totally ignoring the factor that the user should know about the purchase it makes and be able to decide for themselves. Transparency for the player is not a bad thing.
Because it does not matter if you use Windows or Visual Studio. These are not generative tools based on stolen data. Ai is. If you don’t understand the difference and why people have an issue with Ai, then no wonder we don’t agree on.
What data is missing? It’s stolen in the same way an artist “steals” art by watching a Disney movie. Or a musician “steals” it by listening to an album. Nothing was stolen.
And Visual Studio is very much so generative. You generate programs with it using code; code that by your definition was “stolen” from a textbook, a blog, or a StackOverflow post.
No, it is not. Data is extracted and copied and analyzed with machines. As said before, you seem not to understand the different of a Large Language Model dataset/program and a human. This is not the same, nor should it be treated like the same. It’s like suggesting that a Bash script you wrote can download, extract data from books and then you claim its the same as how humans watch films. It is not.
No. Stolen means not respecting the license and not even giving credit.
It is only different because you say it’s different. Whether human or machine, in both cases it’s learning from the material. There’s no law in any county against a machine learning from content. Whether there should be is an entirely different discussion.
Stealing has a definition: “Stealing is the act of taking, using, or appropriating another person’s property, ideas, or services without right, permission, or legal consent, with the intention of permanently depriving the true owner of their use or value”
Permanently depriving the owner. It means something is taken and is now gone. Nothing is missing. Nothing was stolen.
Ai is not “learning”. These are autocomplete algorithms. They don’t understand it. As said copying into a database and then just autocompleting based on probability is not learning and thinking. There is no reasoning. Its fundamentally different to how humans learn.
Whats the source for the definition and why is that definition correct and mine wrong? Did you ask an Ai? That’s probably you don’t have a source.
That’s when you talk about real goods, not code. You forgot your Ai to tell that. In software and art stealing is if you take it and use it without permission, and possibly give no credit.
In example, if you see a photo on the net that is sold to you, but you copy it and use it without permission, without respecting the license, then its stolen image. The owner does not know, nor did you rob the ability for him to use the image. Another example is to take an Open Source project that is licensed under GPL, but then use it without respecting the license and sell the product on Steam for money. That is stolen Open Source project.
You literally don’t understand what you’re talking about. It’s not a database, no copies are being made. The weights of digital synapses are adjusted in a neural network, the same as happens in our brains when we learn something. For all intents and purposes it is indeed learning from an example, not making a copy. There’s no “auto-complete” involved, anymore than there is when a human creates something using something else as inspiration.
No copies are being made, no money is being made by something being sold. You can download the models and use them for free. You are only paying for access to the hardware to host the neural network, not the model itself. Again, nothing has been stolen, and no copies have been made. It’s learning from data, same as people do from open source projects.
If there was a license that said you are not allowed to learn from reading the code, then you might have a point, but there are none of those that exist to my knowledge. If you weren’t meant to learn from them then they wouldn’t be publicly available.