Open link in next tab
Uncovered EA patent would use AI to replace professional voice actors with (fake) you
https://www.tomshardware.com/news/ea-player-ai-va-patent
Even the best version of this sounds like an unmitigated disaster
https://www.tomshardware.com/news/ea-player-ai-va-patent
Even the best version of this sounds like an unmitigated disaster
What an incredibly godawful, dumbass idea.
Look, I'm going to skip over issues of voice actors, quality, etc, and go directly to:
Your voice doesn't sound to you like it sounds to other people. Almost everyone hates playback of their voice. This is a great way to get people to haaaaaate playing your game.
In theory, the idea is okay. I wouldn't mind an avatar having my own voice.
In practice, fuck no. It's EA. I don't want my voice to live on producing new dialogue long after I'm dead in order to fill some corporation's pockets.
Idk, my kids love it, and I wouldn't be surprised if it's a lot more common than not.
I personally hate it, but then again I would also hate an avatar of myself as well in a game. I prefer to RP as someone else in games, not as myself (I like myself, I just don't like to be myself in games).
I'm sure they could change the synthetic voice so that it sounds like your perception of your own voice
Interesting idea, but I doubt that the modulation is the same for everyone, just cause our heads aren't identical in shape or size. I suspect that my head distorts my voice differently than most other people's heads.
I don't particularly hate my voice when played back.
I just don't recognize it as my own voice at all.
Is the AI gonna play a voice I pick, or am I just going to talk to the game and hear the responses of the other, non-player characters like I was having a real conversation with people? Because one of those is an awesome idea.
If I read the article correctly, instead of the protagonist character having a voice actor, they want to take a bunch of voice samples of you and then use that to generate their dialogue.
I wouldn't be surprised, in the slightest, to learn that agreeing to play this game, gives them full rights to the voice generator they build from your samples, and they pick and choose from those instead of having any voice actors in future games.
i don't talk while playing, that character gonna sound like family talking in the other room lmao
The way it's written in the article, with the mention of the detective game and how it is using it, it sounds like it's not generating a voice based on your voice, but dialogue based on what you are saying. But then the article author's use of "dubbing" throws that into question. I'm pretty sure I've seen the detective game they mentioned in the article, though. And it just generates the NPC responses on the questions (or statements) you give to them. And you're talking to them entirely with a microphone, not typing anything or selecting pre-made choices.
Setting the shittiness aside, once again a horseshit patent award. This is not a novel or innovative idea. It's a stupid fucking limitation on others if entertained.
Exactly. I'm completely fine with patents over something like a prosthetic or manufacturing equipment. I'm not okay with patents over software or business methods.
As a kid, I liked law and computers, so I thought I wanted to be a software patent attorney. Midway through my CS program, I decided software patents are completely awful and decided to work on FOSS instead of go to law school. Software patents should all be invalidated.
That said, I think patents should have a much shorter duration. I'm thinking something like 2-3 years, with an extension to 5-7 years if the patent holder can prove they need the extra protection to bring the product to market (i.e. they can demonstrate active work on it). Maybe certain types of patents can have another extension if it's a long lead-time product, but definitely not longer than 15 years. Most patents should expire within 7 years.
Software patents need to be shorter.
Hardware ones I think can be as long as they are, but need loopholes and tricks closed that allow for extending patents on the same thing artificially.
Best would be to have many different categories with vastly different duration and the durations need to be reviewed periodically.
Like the fact large parts of x86 is still patent protected is an obscenity.
I agree with different durations based on the type of product, but I really need to see some evidence that the current patent length is needed by anyone. First mover advantage is a real thing, so they only need enough protection to get a head start. Patents are just a license to be lazy, so they should only exist as long as necessary to get to the market first.
Seriously. Patent length was chosen before the industrial revolution. It only coincidentally made sense through that period of mechanization. But in computing, twenty-odd years is an eternity.
In 2000 there were no shaders.
In 1980 there were no IBM PCs.
In 1960 there were no microchips.
Why the fuck would any idea from when Pong was fresh and new deserve absolute control until after the Super Nintendo? There could be Dreamcast games with features that that nobody was allowed to do again until last year.
An AI image is a result of a technology, which falls under copyright.
A method to generate AI audio is the technology itself, which falls under patent law.
These are two entirely different things that should never be conflated.
Exactly. You can patent the way you integrate this tech into a game, as well as the process for how the audio is generated, but you cannot patent the audio itself or the implementation since those fall under copyright. Use of an implementation would need a patent grant, but use of the audio does not.
Patents and copyright are two sides of the same coin, and as such are related but are completely separate entities. Patents cover ideas and processes, copyright covers implementations and products.
Agreed on the general principle, but I'm kind of glad a company that everyone already thinks of as shit will hold the patent on this. It's absolutely not an idea that I'd want to spread throughout the industry and at least now it's limited to use in games I'll never play.
We already fucking hate hearing our own voices on recordings or in Discord with feedback. Why does anybody think this is a good idea?
Not to dismiss people who are fine listening to their own voice but this feels like mega narcissistic at the very least and just plain weird to everybody else.
I'm okay with it if it's optional (some people love to hear themselves), but I am absolutely not interested. Then again, I'm pretty much the polar opposite here since I don't even like public recognition. My kids would love it though, since they love hearing and seeing themselves in videos and whatnot.
But if there's no way to disable it, I'm not buying the game, simple as that.
One, FUCK software patents.
Two, this is what AI means for voice acting: there won't be any voice acting. Nobody has to get cloned. There doesn't need to be an original performance to modify. We have the technology for artists to make up how a character sounds, the same way artists can make up how a character looks. Like how nobody plays a cartoon. There's still live-action, and mocap, and there will still be people's real voices behind imaginary faces. But that is now optional.
Three, nobody wants to hear their own voice. Morgan Freeman probably thinks he sounds weird and nasal onscreen.
If it were entirely local and respected privacy, sure.
Since there's no shot of that, it's a nightmare.
Yup, my kids would love it (I would hate it though), but I'm not buying it if it's cloud-based.
That said, I avoid EA games in general, so it wouldn't impact me either way.
This is the best summary I could come up with:
The patent itself spends a considerable amount of time documenting the technical features of how input data is processed, but I'll be focusing on the game design implications here since they're more interesting.
Inworld Origins hinges around the player roleplaying as a detective and asking AI NPC characters questions, with answers generated on-the-spot based on that input.
It's so possible that recently, official English anime voice actors were openly speculating that they had been dubbed over with AI by Namco Bandai in the latest Naruto Ultimate Ninja Storm game.
Most video game playable characters already function primarily as self-inserts from a narrative perspective, but making that connection literal seems like a quick way to break immersion.
Considering the sheer scope of Electronic Arts and the funding available to them, I think they should just stick to hiring actual people for video game voiceover, especially for plot-critical characters.
Adopting technology like this seems like a pretty sharp step backward from an industry that's been selling video games to us with face-scanned screen actors for the better part of the past decade.
The original article contains 473 words, the summary contains 180 words. Saved 62%. I'm a bot and I'm open source!