• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

Unity stock falls alongside Take-Two, Roblox after Google’s Project Genie launch

Draugoth

Gold Member


Unity Software (NYSE:U) stock tumbled 12% Friday, while Take-Two Interactive Software (NASDAQ:TTWO) fell 7% and Roblox Corporation (NYSE:RBLX) dropped 8% following Google's announcement of Project Genie, a new AI-powered world creation tool.

The gaming and interactive software companies saw their shares decline as Google, a unit of Alphabet (NASDAQ:GOOGL), unveiled its experimental prototype that allows users to create, explore, and remix interactive virtual worlds. The new offering appears to have sparked investor concerns about potential competition for established game development platforms.

Source
 
I really don't get people's aversion here to AI. It is the future.

All those who wanted the most realistic GFX, this is it.

No more 3D engines as we know it.

Epic is prob working on something similar as we speak.
 
I really don't get people's aversion here to AI. It is the future.

All those who wanted the most realistic GFX, this is it.

No more 3D engines as we know it.

Epic is prob working on something similar as we speak.

AI means we won't have to wait 10+ years for a GTA/Fallout/Elder Scrolls/Witcher etc
 
AI means we won't have to wait 10+ years for a GTA/Fallout/Elder Scrolls/Witcher etc
It means more people come to the fold of gaming, and all those losing their jobs because of exec/purple hair decision can strike on their own. And Art Director with an idea will be able to create a game by himself and some others.

Unity opened up that space for indies, and AI will open it up even more. it'll create more indies. more diverse games and def more shit but that's why you have ratings.
 
I really don't get people's aversion here to AI. It is the future.

All those who wanted the most realistic GFX, this is it.

No more 3D engines as we know it.

Epic is prob working on something similar as we speak.
there won't be no future if ram prices keep spiking like this
 
You see I don't agree with this move, I think what Tim Sweeney said makes sense, that these features will steadily integrate with existing game engines and they'll naturally intersect, kind of like this tech does with Photoshop.
 
I really don't get people's aversion here to AI. It is the future.

All those who wanted the most realistic GFX, this is it.

No more 3D engines as we know it.

Epic is prob working on something similar as we speak.

The problem is that much of GenAI is hype, made possible only by trillion-dollar investments.

The volume of matrix calculations and computational power required to render Sydney Sweeney in a furry suit with Grok 4.0 is so absurdly high that it's only feasible thanks to the insane capital that companies like Tesla and Alphabet continue to pour into it. In a sane world, there would be no way they would allow people to generate an unlimited number of pictures and videos with GenAI.

The current Genie demo is interesting, but they can barely retain environmental information for object permanence beyond a few seconds, not to mention being riddled with artifacts. Also, what kind of PC would be required to run this locally? This is an extremely suboptimal way to handle 3D assets.

Always be skeptical of the tech you see, no matter how impressive it is; otherwise, people would still be pre-ordering Ubisoft games based on their trailers.
 
Last edited:
I really don't get people's aversion here to AI. It is the future.

All those who wanted the most realistic GFX, this is it.

No more 3D engines as we know it.

Epic is prob working on something similar as we speak.
I can see AI being integrated into game engines for things like auxiliary asset production and other stuff. This however? There are far too many issues with this approach, from the ridiculous amount of processing power needed to many well-estabilished limitations in these current AI models that can't be easily solved.

Think of it like this. What is more efficient? Have the AI generate these worlds on the fly. Every. Single. Time? Or have it generate the assets that can then be used in a 'normal' game world for a fraction of the processing power? Which also happens to be far more reliable in terms of persistency and reproductibility?
 
Last edited:
AI means we won't have to wait 10+ years for a GTA/Fallout/Elder Scrolls/Witcher etc


The only reason for delays and cancelations is bad management. Is AI going to fix that? No, sir. The industry is suffering from a talent drain which precisely will be aggravated by the massive use of AI.

I really don't get people's aversion here to AI. It is the future.
Nothing is "the future". This is just a marketing statement that is made upon the release of the next big thing. AI is already being used in the making of books and movies with atrocious results. Good works always come from EFFORT and talent. There are no magic shortcuts. Your "future" is a fast food production chain, not gourmet.
 
I can see AI being integrated into game engines for things like auxiliary asset production and other stuff. This however? There are far too many issues with this approach, from the ridiculous amount of processing power needed to many well-estabilished limitations in these current AI models that can't be easily solved.

Think of it like this. What is more efficient? Have the AI generate these worlds on the fly. Every. Single. Time? Or have it generate the assets that can then be used in a 'normal' game world for a fraction of the processing power? Which also happens to be far more reliable in terms of persistency and reproductibility?
I'm saying exactly this. I don't know how it'll embed itself in gaming, I just don't understand why people are taking a stance when it's clear that this is the future we're headed too.

The only reason for delays and cancelations is bad management. Is AI going to fix that? No, sir. The industry is suffering from a talent drain which precisely will be aggravated by the massive use of AI.
It'll move to less "management" and execs and more to the power of the creatives.

Nothing is "the future". This is just a marketing statement that is made upon the release of the next big thing. AI is already being used in the making of books and movies with atrocious results. Good works always come from EFFORT and talent. There are no magic shortcuts. Your "future" is a fast food production chain, not gourmet.
you can burry your head in the sand, or you can see what a huge change AI is bringing. That's not marketing speak that's just the way it'll be.

Talent will be able to unleash it's creativity better with AI.
 
I'm saying exactly this. I don't know how it'll embed itself in gaming, I just don't understand why people are taking a stance when it's clear that this is the future we're headed too.
The fear is devs relying too much on AI and having their games look the equivalent of asset flips, replacing good art and artists for cheap AI generated assets. I don't think this'll happen, for the same reason you don't see serious productions looking like amateur unity tech demos using store bought assets.
 
Last edited:
This is the next stage for indies, just like it became easy for anyone to make a 2D roguelite game with the random generation feature.
 
I really don't get people's aversion here to AI. It is the future.

All those who wanted the most realistic GFX, this is it.

No more 3D engines as we know it.

Epic is prob working on something similar as we speak.
I am not sure how you could build a full game in that. Looks more like small rooms with very limited interactions.
 
I don't care if devs use AI for productivity.
Using AI for a whole game right now is bound to have many issues though, I wouldn't spend money on that at all.
Unless the tech can work like your own dreams, generating content on the fly, I don't see how this can be useful beyond getting a vision for a game.
 
It's for sure very impressive as a technological achievement. It deserves utter respect.

However, I think that this kind of process is simply missing the point of what makes games worth it. The difficult part of (success) gaming development is not (or, not only) the technical layer complexity, which AIs may or may not solve/smoothen out, but actually finding and implementing the fun. And that can't simply be AI generated starting from existing material: it is an iterative, human, and very creative process. It is what makes some indies special. Gaming dev has an intristic human factor that I think it's very difficult to be reached by machines.
 
Seems like the next frontier. New tech that will shape the medium for decades to come.

I wouldn't be surprised if by 2050 we'll look back at the first AI games the same way we do at the first PC (DOS) games.
 
I really don't get people's aversion here to AI. It is the future.

All those who wanted the most realistic GFX, this is it.

No more 3D engines as we know it.

Epic is prob working on something similar as we speak.
It is the future. I agree
 
Very cool. However, I won't be paying any money for AI generated games

Oh, you will be doing that, as a software developer myself, I'm using it to write code every single day, everybody in the software industry is.

If you mean games that are using this kind of product, well, there will be a time that you won't be able to notice what was produced using AI tools.
 
Wait, peope are drinking the Kool aid?.
Danger 5 Laughing GIF

Its even more funny when AI booty lickers see themselves as the smartest ones in the room.
 
No more 3D engines as we know it.
The same way noone uses local gaming devices to play anything anymore.

Anyway of all the AI use-cases, world synthesis is a lot more interesting (and proof positive) than any of its rendering artifacts. Especially if you're chasing realism.
 
Perhaps Tim Sweeny will use AI to completely make games at a rate of every two weeks so he can give them away free instead of using other publishers' games to give away.
 
More dumb reactions from the stock market and investors expecting that videogames will create themselves now with minimal costs involved? Shocking outcome.
 
Oh, you will be doing that, as a software developer myself, I'm using it to write code every single day, everybody in the software industry is.

If you mean games that are using this kind of product, well, there will be a time that you won't be able to notice what was produced using AI tools.

Are you and everybody using it in the software industry paying for it ? Out of curiosity.
 
Very cool. However, I won't be paying any money for AI generated games
More likely you'll just be paying an ever increasing sub for computation time on the companies' servers into which gaming (as we'll know it) will be rolled up.
 
Last edited:
Are you and everybody using it in the software industry paying for it ? Out of curiosity.

My company pays for Gemini Pro license for every developer, we also hace n8n licenses. There are added costs, like, for example, if you want to use the API from Google AI Studio.

Removing documentation, unit testing and writing most of the code from our plate makes the teams incredibly productive. Now, I've been in the industry for over 20 years now, I can actually check the output the AI tools produce is correct, nobody will push code to testing or prod without checking the solution makes sense (at least with the teams I work with).
 
It's very cool tech and to be able to explore is really mind blowing.

l can imagine people using this to author scenes, characters (or whatever objects) quickly and export then import to a traditional game engine.

Wouldn't be surprised to see games being a possibility in future iterations but can't imagine it will be cheap to run on the cloud even for a few minutes worth of gameplay.
 
Last edited:
My company pays for Gemini Pro license for every developer, we also hace n8n licenses. There are added costs, like, for example, if you want to use the API from Google AI Studio.

Removing documentation, unit testing and writing most of the code from our plate makes the teams incredibly productive. Now, I've been in the industry for over 20 years now, I can actually check the output the AI tools produce is correct, nobody will push code to testing or prod without checking the solution makes sense (at least with the teams I work with).

I see. One more question if you'll indulge me. Will your company keep paying for it if the price increases ? If it doubles per seat ? If it triples ? Quadruples ? When's the cost no longer worth it ? You may be extracting value now, but what about the future ?
These big AI companies are burning billions per month and these models are basically subsidized by investors. When the obvious price increase comes as it surely must, because they have to stop burning money at some point and become profitable, will it still be worth it ?
 
Can't wait to get stuck or fall through maps every with every half-assed content update. I could see a use for "look-but-don't-touch" backdrops though.
 
I see. One more question if you'll indulge me. Will your company keep paying for it if the price increases ? If it doubles per seat ? If it triples ? Quadruples ? When's the cost no longer worth it ? You may be extracting value now, but what about the future ?
These big AI companies are burning billions per month and these models are basically subsidized by investors. When the obvious price increase comes as it surely must, because they have to stop burning money at some point and become profitable, will it still be worth it ?

With the proper project tracking you can calculate how much you're saving and how much more productive teams are, now we're doing smaller teams for big, complex features/applications cause it takes less time to build and test them.

What took teams of 7-9 people, now we do with 4-5, so, when new clients are onboarded we are currently just breaking up teams and moving people around trying to avoid layoffs.

To answer your question: I don't know what happens if it became prohibitively expensive, but I, personally, would pay for my seat, or at least use a local open source model. I'm not going back to spend days doing documentation, unit testing or finding a needle in a haystack in the codebase. Its just too useful.
 
Top Bottom