Gen AI needs web3 & this is how gaming wins
Gen AI without web3 is terrifying, but the combo is a massive opportunity
Artificial intelligence is nothing new. But generative AI is newer, and certainly the flavor of the month for investors and big tech. Being worried about AI is also nothing new. But being concerned about generative AI has only infiltrated the zeitgeist over the past couple years as a mix of really fun and really scary deep fakes and derivative art have surfaced.
Gen AI is part of our future no matter what, and the challenge for creators and IP holders is now how to protect from and incentivize a reality in which anyone can create almost anything. Regardless of White House pronouncements aiming to control AI, we already know from ChatGPT, Midjourney—and a lot of the other gen AI platforms that have become quickly accessible—that people love creating: paintings, photos, movies, poems, full novels. And nothing about that is slowing down, particularly the more that we feed these models with our creative minds’ deepest desires.
Artists and IP holders are already lamenting what they rightfully point out is unapproved training of models on copyrighted work. While this is frustrating for artists trying to protect their style and source of income, it’s also a harbinger of a future where AI tools will let people further harness the art and content that they love to create their own things. There’s nothing new about this: homages, fan art, and fan fiction are as old as art itself. Rather than fight everyone’s natural proclivity for making nods to beloved content, some companies are leaning into the benefits of exponentially easy creation with the one technology that will ensure original creators maintain control and the benefits of their work proliferating: web3.
Let’s take a look at this future through the lens of gaming.
Handing over creative control to adoring fans has often been a wise business move. Nothing gives audiences more attachment to the lore of IP than adapting it. Think of your earliest childhood games: perhaps reenacting a favorite movie, playing as your preferred character? Every IP-based toy is essentially an opportunity for a kid to take control of the sequence of events from the original show or movie and make it his or her own. All of that is big business (the global movie merch market is about $30B). In the past decade, a lot of this type of play has shifted toward gaming: physical merch and IRL childhood play based on movies and shows has simply become Fortnite skins and Lord of the Rings Minecraft servers.
People want to take content that they know and love and do something new and exciting. In gaming, tweaking games has been around for 40 years.
A brief history of modding
Modding began in the 80s. Players, sometimes with the blessing of developers (but sometimes not) took it upon themselves to modify a game’s code so that they could do new things with it. In some instances, the mods became more popular than the vanilla version of the game, and some mods even became full, monetized games themselves. Mods were a way for the community to expand the gameplay, tell new stories, fix nagging issues, or just create all-out insanity with very fun changes. I remember my days of Starsiege Tribes, which modders took apart to create new game modes, servers with infinite ammo and jetpack time, and more.
In many instances, mods could breathe new life into games in a way that would be beneficial to the developers, bringing new users into the game purely for the mods. But modding wasn’t always great for games, either. Yours truly remembers playing the second Tribes game, but only the demo, which mods expanded on so much that it basically became a full, playable game. I never actually had to buy the title, so that wasn’t great for the developer.
At the same time, modding typically didn’t use AI, and wasn’t as focused on bringing in external IP. But it was the beginning of creators changing the nature of the games they loved, which has since evolved substantially (more on that next), and it also showed the potential financial dangers or benefits of letting others manipulate the look and feel of your game.
Modern day user generated content
Today, we’ve got games that not only like it when communities take and expand on what the original developers created, but are actually structured as sandboxes on which users can make any kind of experience they want. This is particularly important in a world where live ops games are the standard, and players expect games to evolve and improve constantly.
Games like Roblox, Fortnite with UEFN, and many of the wannabe metaverses are all built on the premise of handing creative control to others, and building games is only getting easier, as AI tooling makes it increasingly accessible for nontechnical people to create assets and environments. Without structures in place to control what models are trained on and what gets created, however, these types of worlds could easily verge into content territory owned by others. Ironically, this concern also presents a huge opportunity to brands and IP holders who—at the same time—are desperate to reach users where they’re spending more time and money than almost any other media: in gaming.
To be clear, user generated content tooling in games that leverages AI can be built on models that don’t involve any existing IP, and indeed platforms like Steam have made it clear that they’ll refuse any game that can’t prove its AI content was free of copyrighted material. With web3, however, IP holders can for the first time open up the kimono of their works to be fairly used with AI and rewarded through the blockchain, creating an entirely new future of modding, UGC, and being able to reach players.
The power of web3
There’s a convergence of factors here:
average players who want to create
brands and artists that want to reach users without necessarily doing laborious licensing or ad deals every single time on a game-by-game basis and
AI making it easy to create anything.
We’re seeing bits of these things coming together, but only with the blockchain can we unlock the way that this all scales. Adim, a platform launched by “It’s Always Sunny in Philadelphia” creator Rob McElhenney, allows creators to both collaborate on stories and content together, with web3 permanently emblazoning their efforts, and then lets anyone take pieces of the content—characters, plot lines, etc—and remix them into new fanfiction, with the idea being that original creators getting what’s due to them as part of it.
Layer Licensing, on the other hand, is a marketplace that connects game developers to licensors to use IP in their games. For brands and artists, reaching gamers is crucial as it’s the biggest media category in the world and also where increasing amounts of consumer time is going. But not all situations need to be formal licensing deals, particularly in a world where web3 can allow assets to flow between different games and ecosystems, and licensing in web3 can and should look fundamentally different than traditional games.
UGC platform Overwolf, which lets developers easily add content tools for players to a game, plays into this by vastly expanding the reach of UGC for both developers and players. It also has—as proof of the need—a thriving ad business where the UGC-manipulatable real estate can also be leveraged as places onto which brands and artists can place their content.
Finally, there’s Scenario, a Gen AI platform that focuses on games. The platform lets developers train models on only their concept art and IP, which is a good thing as it prevents the AI from infringing on others’ copyrighted material. But it also potentially limits how these games can let players leverage existing content to make their own in-game assets.
The nexus of platforms like these overlapping is where creators can start to engage multiple tools like AI and existing IP to create entirely new categories of gameplay and game creation for example with:
story and characters from one set of creators, remixed using AI
easy access to licensed content that help brands and artists reach players and
UGC platforms that allow developers to more easily offer creator tools
easy asset creation incorporating all of the above, where web3 offers the layers of provenance and royalties for each of the OG creators.
The win here is giving creators the tools to help developers become even more successful by strengthening the ecosystem and replayability of titles, and letting existing IP seep into new ecosystems without the onerous and tedious commercial and licensing structures of the past.
Obviously, there are a lot of snags to work out here. For example, how do you sort out royalties across many parties? An author may feel her work deserves 20% of any derivative work, while a character artist may feel he’s owed 15% to let his content be used for in-game assets. Either the incentives will erode so quickly that the creator doesn’t feel it’s financially worth using this mix, or else some kind of negotiation system with universally-accepted guiding principles will need to be considered. Not all IP will be comfortable with permissionless usage; i.e. the idea that anyone could use their IP so long as the royalties remain intact. This could set up image conscious brands to be in games or next to other content that they don’t want to associate with.
And for many of the world’s biggest IP holders—the Disneys etc.—the revenue from minimum guarantees may be so tantalizing that they’d never consider a an IP usage deal without it. (Not that that’s stopped Disney characters from being easily accessible through platforms like Midjourney.) Others artists and brands, however, would be delighted for their content to reach gamers and would put up as few barriers as possible as long as money and exposure came back to them. Imagine, for example, an up and coming sportswear brand with an ardent following that would never get the Fortnite skin treatment, but whose designs and logo could be offered—with the blockchain—to UGC creators adding new skins and content to an extreme sports game. The game gets new content and perhaps some of the eyeballs of the brand’s audience, the brand gets the eyeballs of a big extreme sports player base, and users get something fresh and maybe recognizable to engage with.
The future doesn’t have to be a conjunction of all these various trends, platforms, and methods, but by using the blockchain and AI, it’s clear that creators will have more ways to create more easily and IP holders will be able to reach new audiences and get their just deserves. However, it’s only with web3 tying together these disparate channels and ecosystems do all parties have the opportunity to unlock this kind of content at scale.