As someone whose heard of retrobrighting, but never actually investigated how its one: so it involves using hydrogen peroxide? Well then, I'm not shocked you get damage. Hydrogen peroxide bleaches things by oxidizing them. You know what causes a big chunk of damage to plastic? Oxidation. While plastic chemistry can be complicated, this feels kind of like trying to repair burn damage to an object by "cleaning" away charring with a blowtorch.
This is not exclusive to the video game industry, sadly. One notable example from the tabletop gaming business: the company "Dyskami Publishing" basically exists the distance new products like the current edition of BESM from "Mark MacKinnon", their creator. Why? Because his name is still at least moderately toxic, even a couple decades later, for basically grabbing the cash box and running across the border when the RPG industry started contracting after the peak d20 OGL days.
Basically, not using AI as a tool to create a game, but using AI as part of the game, with an LLM serving as a "virtual GM" to manage to ( more traditionally designed ) procedural content. Its ridiculously ambitious and I only give it maybe 20% odds of working out well. . .but even if it fails it'll at least be the right kind of failure.
Being "fair", I am skeptical that this is an actual sincere belief on their part. Rather, its a pretense to cover their actual belief: that they don't want people buying older games, because the current execs make bigger bonuses and gain more clout from new games.
I'd bet that's an artifact from the days when IBM sold you both the product and the upkeep. They likely were originally writing those manuals, not for the end user, but for their own maintenance personnel. And even when this shifted, the design culture stuck.
And had most of the contents of the AD&D 2e players handbook, hence the size. Though even completely original 90s WRPGs often had similar sized manuals, hence why IMO they were the peak for physical manuals.
I would say the key is not precisely that games these days have "too much" stuff, but that they have too much extraneous stuff. Things added, not because its part of the Core Vision of the game or contributes to such; but because that individual item is perceived as popular and thus hopefully leads to more people trying and buying the game. Which definitely emerges from the "trying to be everything for everybody" concept. Big AAA studios generally hate greenlighting anything that isn't at least potentially aiming to be The Next Big Thing That Earns All The Money And Dominates The Industry.
I wouldn't rule out "thieves stole loot without a plan for what to do with it". Its shockingly common that even seemingly-talented thieves able to pull of quite clean heists, have no actual idea of the challenges of turning the stolen items into usable cash. Which could lead to the items showing up on Craigs List, or it could lead to the mouldering in some storage unit for years.
LLM/Stable Diffusion tech is the same tech, whether its called "generative AI" or something. Its just a question of what you are training that tech to do. Hence why a chatbot and an art AI can both be made from the same underlying technology.
Its not that shocking in retrospect: Nintendo is by far the oldest company in the industry. More than any other, they probably came in with a corporate culture of long term thinking. "Ooops we lost important info" is probably a mistake Nintendo made, and learned from, decades before most of the people in the industry were even born.
Basically, a modern Daggerfall using GenAI tech as a "virtual GM" managing the procedural content.
It is ridiculously ambitious, and I'd personally only give it a 25% chance of living up to its aspirations at best. However, its the right kind of ambition to be interesting ( maybe even if it fails ), and an example of how to use GenAI tech in video games as something other than a questionable excuse to reduce labor costs.
I think the key here would not using GenAI as part of developing the game, but using GenAI as part of the game engine itself. Which is to say, procedural games already effectively involve "training" the procedural engine to produce the desired array of results in game. Now you could do the exact same thing, only with the much more powerful GenAI tech as the underlying procedural "engine".
Of course, the "problem" with this is that it would require just as much work producing appropriate bespoke art for the training, and labor in doing the training, as it takes with current procedural games. Except that's only a problem if your goal with using GenAI is to produce a magic free "infinite art forever" button. For everyone else, it would just be a tool to do procedural design, only better than before.
Comments 112
Re: "Retrobrighting" Might Actually Cause More Harm Than Good To Your Yellowing Consoles
As someone whose heard of retrobrighting, but never actually investigated how its one: so it involves using hydrogen peroxide? Well then, I'm not shocked you get damage. Hydrogen peroxide bleaches things by oxidizing them. You know what causes a big chunk of damage to plastic? Oxidation. While plastic chemistry can be complicated, this feels kind of like trying to repair burn damage to an object by "cleaning" away charring with a blowtorch.
Re: Random: Tommy Tallarico Got Bodied So Badly He's Now Using A Fake Name
This is not exclusive to the video game industry, sadly. One notable example from the tabletop gaming business: the company "Dyskami Publishing" basically exists the distance new products like the current edition of BESM from "Mark MacKinnon", their creator. Why? Because his name is still at least moderately toxic, even a couple decades later, for basically grabbing the cash box and running across the border when the RPG industry started contracting after the peak d20 OGL days.
Re: If The Oliver Twins' Ghost Hunters Is The Future Of GenAI Gaming, Then We Have Nothing To Worry About
@MontyMole
The one game I've seen anything about that seems to be trying to use AI for something actually interesting:
https://store.steampowered.com/app/1685310/The_Wayward_Realms/
Basically, not using AI as a tool to create a game, but using AI as part of the game, with an LLM serving as a "virtual GM" to manage to ( more traditionally designed ) procedural content. Its ridiculously ambitious and I only give it maybe 20% odds of working out well. . .but even if it fails it'll at least be the right kind of failure.
Re: Mighty Final Fight Forever Will Launch Just In Time For Christmas
@Sketcz
Being "fair", I am skeptical that this is an actual sincere belief on their part. Rather, its a pretense to cover their actual belief: that they don't want people buying older games, because the current execs make bigger bonuses and gain more clout from new games.
Re: Random: Remember When Games Came With Instructions? This Guy Does, And He Wants To Find The Heaviest PS1 Manual
@FeRDNYC
I'd bet that's an artifact from the days when IBM sold you both the product and the upkeep. They likely were originally writing those manuals, not for the end user, but for their own maintenance personnel. And even when this shifted, the design culture stuck.
Re: Random: Remember When Games Came With Instructions? This Guy Does, And He Wants To Find The Heaviest PS1 Manual
@PKDuckman
And had most of the contents of the AD&D 2e players handbook, hence the size. Though even completely original 90s WRPGs often had similar sized manuals, hence why IMO they were the peak for physical manuals.
Re: "There Weren't A Lot Of Extras, So It Had To Be Done Right" - Fallout Co-Creator Reveals What Modern Game Devs Can Still Learn From The '80s
I would say the key is not precisely that games these days have "too much" stuff, but that they have too much extraneous stuff. Things added, not because its part of the Core Vision of the game or contributes to such; but because that individual item is perceived as popular and thus hopefully leads to more people trying and buying the game. Which definitely emerges from the "trying to be everything for everybody" concept. Big AAA studios generally hate greenlighting anything that isn't at least potentially aiming to be The Next Big Thing That Earns All The Money And Dominates The Industry.
Re: More Than $25,000 Of Rare Coin-Op Components Stolen From North American Arcade
I wouldn't rule out "thieves stole loot without a plan for what to do with it". Its shockingly common that even seemingly-talented thieves able to pull of quite clean heists, have no actual idea of the challenges of turning the stolen items into usable cash. Which could lead to the items showing up on Craigs List, or it could lead to the mouldering in some storage unit for years.
Re: The Oliver Twins Are Reviving Ghost Hunters Using (Shudder) Generative AI
@Santar
LLM/Stable Diffusion tech is the same tech, whether its called "generative AI" or something. Its just a question of what you are training that tech to do. Hence why a chatbot and an art AI can both be made from the same underlying technology.
Re: "There's Basically Nothing" - Final Fantasy VII Remake's Director Reveals "Almost No Documentation" Exists For The Original
@NatiaAdamo
Its not that shocking in retrospect: Nintendo is by far the oldest company in the industry. More than any other, they probably came in with a corporate culture of long term thinking. "Ooops we lost important info" is probably a mistake Nintendo made, and learned from, decades before most of the people in the industry were even born.
Re: The Oliver Twins Are Reviving Ghost Hunters Using (Shudder) Generative AI
@Sn0w
There's at least one game in development that is more or less aiming for this.
https://store.steampowered.com/app/1685310/The_Wayward_Realms/
Basically, a modern Daggerfall using GenAI tech as a "virtual GM" managing the procedural content.
It is ridiculously ambitious, and I'd personally only give it a 25% chance of living up to its aspirations at best. However, its the right kind of ambition to be interesting ( maybe even if it fails ), and an example of how to use GenAI tech in video games as something other than a questionable excuse to reduce labor costs.
Re: The Oliver Twins Are Reviving Ghost Hunters Using (Shudder) Generative AI
@breach187
I think the key here would not using GenAI as part of developing the game, but using GenAI as part of the game engine itself. Which is to say, procedural games already effectively involve "training" the procedural engine to produce the desired array of results in game. Now you could do the exact same thing, only with the much more powerful GenAI tech as the underlying procedural "engine".
Of course, the "problem" with this is that it would require just as much work producing appropriate bespoke art for the training, and labor in doing the training, as it takes with current procedural games. Except that's only a problem if your goal with using GenAI is to produce a magic free "infinite art forever" button. For everyone else, it would just be a tool to do procedural design, only better than before.