NQ's ideas about UGC are wrong

[blundertwink]

Dual Universe obviously didn't work out the way NQ intended, but everything they've posted indicates their leadership is still very huge into the idea of UGC (user generated content) as a vehicle to create gaming experiences. 

 

In my opinion, games like BG3 and Starfield show how "content" is misunderstand by executives that look at day-to-day development as a cost center.

 

Content is not the visuals, models, music, or "things" in a game. As an executive that pays for those "widgets" day to day, they might think that's the case...but really, it isn't. 

 

Content is about interactivity and immersion, not just having a big world for its own sake (as Starfield seems to prove). It's a holistic thing, not just about assets. 

 

The idea of UGC is that everyone can contribute to assets...but it still depends on a strong layer of content crafted by actual professionals. 

 

Trying to empower players to create every facet of engagement and immersion is like turning the Steam or Google Play store into a game, where most the content isn't that good. That's why services like Steam exist to begin with...publishing as a concept still makes sense for all parties, especially consumers. 

 

I think people like NQ want to capitalize on this idea of user-built-everything, but they don't really understand what content actually is and will spend a lot of time and money just to realize that they've created a product that has no real market fit.  

Edited November 15, 2023 by blundertwink
i accidentally hit enter and didn't even want to make this thread tbh

[CptLoRes]

The entire concept of user generated content is a double-edge sword.

 

 To get user generated content worth playing, the players need freedom to express themself and tools so that they can make actual content and not just window dressing. This is what DU was supposed to be, but never delivered.

But at the same time, giving players such freedom in a MMO would be a recipe for disaster and impossible to balance. Just look at how much drama we had (dirt walls, space towers, building above markets etc.) with the limited tools available.

[blundertwink]

17 hours ago, CptLoRes said:

The entire concept of user generated content is a double-edge sword.

 

 To get user generated content worth playing, the players need freedom to express themself and tools so that they can make actual content and not just window dressing. This is what DU was supposed to be, but never delivered.

 

I think there will definitely be a reckoning on the horizon, especially for the many firms focusing on AI-driven UGC or AI-driven assets in general. 

 

And when you think about games like Minecraft (often cited as an example of the power of UGC), the actual engagement point (at least IMO) isn't really the things people create, it's the act of creation itself...

 

When studio CEOs and leaders talk about how generative AI or UGC will change gaming, it's almost always from the widget-driven perspective where these people view making games like a factory, since that's the process they can understand. Producing more "widgets" with less people will of course lead to better things from their perspective. 

 

Not better, though, because they don't talk in language like that....they talk about how it will "vastly increase retention".  

 

It's kind of like claiming that AI will make writing novels much better, since AI can provide "more words with less time". Only idiots would believe that the quality of a novel relates to the "amount of content" it provides. 

 

When you're making something as complex and artistic as a game, CEOs and leadership often get in the way. This push for trying to make their own users develop the game so that they can capitalize on it "forever" is going to backfire because those same CEOs can only view games as cash registers. 

 

I hope that games like BG3 send a clear message about what the market actually craves -- quality, not just quantity. I highly doubt NQ's leadership has learned this even after investing 9+ years in DU. 

[Wyndle]

On 11/17/2023 at 10:06 AM, blundertwink said:

 

I think there will definitely be a reckoning on the horizon, especially for the many firms focusing on AI-driven UGC or AI-driven assets in general. 

 

And when you think about games like Minecraft (often cited as an example of the power of UGC), the actual engagement point (at least IMO) isn't really the things people create, it's the act of creation itself...

 

When studio CEOs and leaders talk about how generative AI or UGC will change gaming, it's almost always from the widget-driven perspective where these people view making games like a factory, since that's the process they can understand. Producing more "widgets" with less people will of course lead to better things from their perspective. 

 

Not better, though, because they don't talk in language like that....they talk about how it will "vastly increase retention".  

 

It's kind of like claiming that AI will make writing novels much better, since AI can provide "more words with less time". Only idiots would believe that the quality of a novel relates to the "amount of content" it provides. 

 

When you're making something as complex and artistic as a game, CEOs and leadership often get in the way. This push for trying to make their own users develop the game so that they can capitalize on it "forever" is going to backfire because those same CEOs can only view games as cash registers. 

 

I hope that games like BG3 send a clear message about what the market actually craves -- quality, not just quantity. I highly doubt NQ's leadership has learned this even after investing 9+ years in DU. 

 

Aye, AI is a tool with near limitless potential but just how many monkeys fudge-flinging around typewriters would it actually take to stumble across Shakespeare?  The tool is only as good as the person or process using it (doubly so for processes).  If you're going to toss BG3 and Minecraft on the comparison heap then I'll raise you NMS, SC. 

Afterwards we can take Todd and Bethesda behind the shed.

[sHuRuLuNi]

Oof ....

I hade made a video on DU from my "What happened ..." Series about exactly this issue - that was 2 years ago .... At that time I was still a tiny bit hopeful that things might change ... not so much now.
The same thing now happened to Starfield as well (among many, ahm, other things). Using AI and proc-gen will never produce engaging content or worlds, just a whole lot of nothing:

 

 

[Maxim Kammerer]

4 minutes ago, sHuRuLuNi said:

Using AI and proc-gen will never produce engaging content or worlds

 

At least without crossing red lines. Today the limitation is no longer the AI itself but the data that is used for training. It must be as varied as possible but it must not contain - lets say - problematic content. That requires manual preselection. With the amount of data required for high quality results this is very expensive - even if outsourced to low-wage countries. And with very expensive I mean very expensive. In case of large language models like ChatGTP we are talking about billions of US$. And you need additional manual postselection for reinforced learning to sort out unwanted results. This is not just an additional cost factor but it also reduces the creativity of the system. The tighter the jail the more boring and repetitive the result becomes. But the more you let the AI off the leash the higher the risk of unacceptable results.