AI is undoubtedly one of the most exciting parts of how the NFT space has changed over time. It was only matter of time we start seeing AI NFTs.
In the past few months, programs like DALL-E and Midjourney that take text-based prompts and turn them into images have become easier to use and more popular. NFT projects that use AI to help make their collections are also becoming more common.
But, depending on who you ask, the very idea of “AI art” is either a joy, a riddle, or an insult to art and the people who make it. But AI could affect Web3 and NFTs in many more ways than just the making of digital art.
Altered State Machine is one company that wants to be on the cutting edge of AI in the metaverse. It has a platform and protocol lets users build, train, and trade “AI agents” as NFTs. This could be the start of a new wave of big changes.
Giving NFTs AI intelligence
ASM, like many people in Web3, is excited about the future of these digital environments. One of its goals is to help make metaverse spaces where AI agents can compete, talk to each other, and help human actors. But the ASM protocol could also be used for many things besides making AI characters that make the metaverse more flexible and alive.
The AI agents in ASM can adapt to many different use cases. And there are a lot of ways to use agents, from games to open-world metaverses like Decentraland to financial use cases.
How an AI NFT will look and act, depends on several factors. The ASM protocol agents are made up of three parts: a “form,” a “brain,” and “memories.”
The ASM game Artificial Intelligence Football Association is an excellent example of how the technology works. It is a metaverse game in which AI teams play soccer matches against each other. Forms, in this case, NFT characters called AIFA All-Stars, can be put into the game. The ASM Brains in the NFTs decide at random what skills they have. During the company’s genesis drop in October 2021, the 40,000-character NFTs in the AIFA All-Stars collection was the first to get what the company calls “Non-Fungible Intelligence” from ASM’s Brains.
So, you have NFTs characters forms that a brain made by AI runs. The last part is the AI brain’s memories, which are where the NFT character’s learned behavior strategies are stored.
The AI brains do this training both in the environments they play in (like a soccer stadium in a metaverse) and in “gyms” designed to improve your AI’s skills. These gyms are linked to GPU cloud computing providers that run training algorithms for a particular ASM application. In our AIFA game example, this might look like making your character better at shooting. If your AI is used for the DeFi markets, it might look like a trading bot’s risk tolerance is being changed.
Owners can pay for training time in these gyms with ASM’s native token, $ASTO, no matter the use case. The company also allows hardware owners to use GPU time given to gyms to make money.
Each ASM-powered NFT or AI NFTs are different in its own way, from how it looks to how it interacts with its surroundings. This is a step from what we have seen in the NFT ecosystem. In generative PFP projects, character accessories and other visual features are put together randomly, but the ASM’s AI agents also have their behaviors and “physical” features chosen randomly. These randomly chosen characteristics will appear differently depending on how the NFT is used.
Once a user has been trained, they can trade or sell improvements. Putting a price on these traits could help create an economy in multiple metaverses and the gaming and financial markets. In an interview with nft now, David McDonald, CEO of Altered State Machine, said, “Training depends on how it will be used.” “How we set up a certain environment is one thing that affects what they learn.”
The AI agents learn from their environments, like a soccer stadium in the metaverse. After this training, the same agents can exist and interact in different metaverse worlds as well, as long as the environment in which they learned to play soccer is also present in those worlds.
There are other ways in which these AI agents can keep their cool. Composability is a system design principle that has to do with how well two systems can work together. In this case, an AI can learn a new “skill” to adjust to a new environment.
McDonald also said that one of the more advanced ways these AI brains could be flexible could be something like walking.
McDonald explained, “Learning how to make a unique way for an agent to move around the world can be a composable element that you can put into games or NPC (non-player character) experiences.” “You could pair that with something like a natural language processor that lets you chat, for example. Or with an AI that can figure out how to get around its world on its own. All of these brain parts can be put together in interesting ways to create interactions and life in environments.”
So far, people are the only ones who can live in and bring life to metaverse spaces. But McDonald thinks that what happens when you’re not in these worlds is a very important part of them.
McDonald said, “Let’s say you have a game store set up in the metaverse, and you come up with a unique way to sell your digital goods there.” “You kind of lose that ability when you’re not there. As these metaverse spaces are built out, ASM can have an effect by letting your assets take over when you go to the beach with your kids or do something else. The other way is that you can make experiences for many people that a single person couldn’t make for themselves. You can set up commerce around digital interactions and scale them in a way that isn’t possible with a single person or even a large group of people.
Muhammad Ali – The Next Legends
Muhammad Ali – The Next Legends, a metaverse boxing game where characters use the company’s AI brains to improve their skills and fight other AIs, is ASM’s most important upcoming project.
The project is a collaboration between ASM and Authentic Brands Group. This company owns Muhammed Ali Enterprises and Lonnie Ali, a trustee of the Muhammad Ali Family Trust. It is set to come out later this year. Non-Fungible Labs, a creative factory on Web3, will design the characters for the game.
The Next Legends will have NFT characters with ASM’s AI brains, just like AIFA. Players will train their characters, who all have different random skills, to get better at things like jabs, uppercuts, and stamina needed to win these fights.
McDonald’s is especially happy about the partnership and the project because Muhammad Ali has been his hero his whole life.
“It’s a great honor to be a part of bringing that story and his legacy into a brand-new space.” It’s a chance to make a really fun and different way to play games and interact with stories that I think people will really enjoy. And preparing us to bring a whole new set of experiences to Web3 and the metaverse in this space.”
When asked how he felt about using the game to introduce people to Web3 gaming, McDonald said that it’s important to meet the needs and interests of both Web3 experts and people who are just starting out in the space.
“We need to make sure that the game is easy for anyone to play, no matter how good they are. I know that some people already know how to use Web3, and those people are what I call “power users.” “They have done the research and likely understand the ASM protocol, how gas works, and how blockchains work in general. But the average user who might want to play a game called “Muhammad Ali AI Boxing” is probably not that smart. So, we have to work with both of these entry points and make sure that the onboarding process is not a barrier to entry.
Putting AI back into people’s hands
In general, what ASM’s technology border means for Web3 is very exciting. In line with Web3’s ownership philosophy, people who collect these agents will have proof that they own an AI whose development and purpose they can change. McDonald compares this to how AI works in Web2. He points out that algorithms in social media are completely out of the user’s control, which can have big effects on a person’s psychology and mental health.
McDonald said, “One of the big problems with preference algorithms today, and we see this with things like TikTok, is that they can lead you down a rabbit hole about what they know about you based on your actions, but that might not be what you’re trying to do.” “That kind of algorithm has a lot of bad effects, like showing people who are depressed content about being depressed. It’s sending them down a rabbit hole they might not want to go down because the algorithm knows that’s where they get the most engagement. I think it will be important for people to be able to control preference AI on their own.