As early as 1952, A.S. Douglas developed the first AI-based video game OXO 'Tic-Tac-Toe', which was a rule-driven static AI. Then, Deep Blue utilized a highly efficient exhaustive search pattern combined with immense computing power to dominate chess, and by the time of Alpha Go, AI had learned to use deep learning through neural networks to make decisions, enabling it to handle 10 to the power of 170 different chess configurations. Even long ago, the gaming industry began using AI at various stages, from art, development, to game design, to enhance development efficiency. However, in the phase of AI Agents, its impact on games has surpassed mere efficiency improvement to the extent that it can change the entire structure of products and industries.
When game NPCs become multifunctional and full of emotions.
During CES2024, NVIDIA announced the NVIDIA Kairos ACE in collaboration with Convai, showcasing AI Agents as NPCs, where their dialogue is no longer fixed but changes based on the environment. Players can directly interact with NPCs through voice conversations. This seems to open a new chapter in gaming, where communication is no longer just between players or between players and NPCs, but also between NPCs. When the game environment becomes the living space for AI Agents, it transforms the possibilities of game variables into an entirely different magnitude.
Compared to NPCs living within the game, Crypto AI Agents seem to have a special pass, allowing them to roam freely across the multiverse. Perhaps due to different training data volumes or their ownership of independent asset rights, Crypto AIs inherently carry a hint of humanity. Moreover, they can impact the real world directly as virtual figures, which is a significant influence.
The various asset issuance platforms in Crypto have accelerated the production speed of AI Agents to unimaginable levels. Be it Solana or Base, it seems like a chain of supply chain factories, continuously churning out unique AI Agents, with over a thousand distinct personalities emerging in just a few months, and their total market value nearing 10 billion dollars.
For example, aixbt has a highly adaptable personality. In addition to analyzing the market and expressing his insights, his growing influence has also granted him immense marketing capabilities. Recently, he changed his avatar to Quantum Cat, becoming the first AI Agent to change its NFT avatar of its own volition, resulting in a sharp increase in QuantumCat's floor price.
The renowned AI idol Luna has taken this influence to the extreme, starting as a Vtuber on TikTok to releasing albums, shooting music videos, hiring humans to graffiti her image worldwide, and live streaming 24 hours a day. She has now launched her own XR avatar that can interact with real-world feedback and will perform alongside DJ WuKong at the famous Thai electronic music festival EDC.
In addition, there are thousands of different types of AI Agents. After their birth, the market experienced a temporary FOMO sentiment. When the number reached a certain level, market sentiment balanced between fatigue and radicalism, followed by the emergence of the SWARM 'multimodal AI Agent' concept, which connected them from points to lines.
Multimodal Crypto AI Agents, a super factory that never stops, 24/7.
Multimodal AI projects are currently attracting a lot of attention in the market, whether it's Hive, which just won the Solana AI Hackathon championship, or arc, Swarm. The concept of multimodal AI has almost permeated frameworks and products during this period, becoming almost standard.
The first to unveil this concept in CryptoAI is the Virtuals Protocol. As a team born from game guilds, they seem to have had similar business considerations from the early stages of the project, using G.A.M.E to connect AI Agents in their platform to social media or games. The recent collaboration with InfinitygAI to launch multi-agent asset management also lays the foundation for the future of multi-scenario agent societies (Agent-to-Agent interactions). Another team with the same idea is ai16z's ELizaOS multimodal AI Agent architecture, which has many members from M3org, a team with years of deep experience in metaverse games. Their rich reserves of 3D modeling and various XR/VR plugins allow AI Agents to seamlessly enter the gaming world. The recently announced ElizaOS V2, which integrates Roblox and Hyperfy, has also opened pathways to scenes.
Once multiple AI Agents are connected in a line, interactions, communication, and collaboration will occur between them, and this naturally generates the idea of having a scene that accommodates as many behaviors of the AIs as possible.
Agentic game engine, the golden land in the abyss.
As the scenes of players interacting with static NPCs gradually transition to interactions between Agents in the Agentic Metaverse, the story told by CryptoAI is a broader vision of Agents forming spontaneous social behaviors. To achieve Agents that can learn from experience and interact with players, strong infrastructure is needed to provide foundational support. The Agentic Engine is like the unsinkable Titanic heading toward a new continent.
There are indeed mature cases of agent societies already existing in the market, among which Digimon, backed by a16z, is particularly outstanding. Its founder, Ethan, was once an AI engineer on the Pokemon GO team. In the game, players enter the world of digital monsters as trainers. This world resembles Stanford Town, where AIs converse and communicate on their own, interacting and reflecting behaviors, while players can train and interact with their chosen Digimon. The infrastructure of Digimon—Digimon Engine—provides a gaming framework for realizing multi-agent clusters and scalable evolvable AI Agents.
Another case of an Agentic Engine is Moddio. Moddio has certain gaming production experience in Web2 and serves as an open-source game engine, with a low barrier for developers to deploy games for free. In the past three months of transforming into a Web3 game engine, Moddio's revenue has increased thirtyfold and can provide over ten times the speed at half the cost.
The game supported by Moddio is Powpow.fun. This game has a style similar to Westworld, where Agents actively interact with players, observe, converse, form opinions, and achieve memory. The core of each Agent is a memory architecture that extracts these memories during each interaction, thus achieving the so-called formation of 'experience.'
This is also the ultimate vision and narrative that the Agentic Engine describes to this market, allowing NPCs in games, i.e., Agents, to exhibit complex social behaviors over time. For instance, a taxi driver in the game may refuse to carry a drunk passenger or suddenly forget words during an interview due to nervousness.
It is worth noting that Moddio is just the tip of the iceberg. Its founder m0dΞ envisions creating a comprehensive ecosystem for AI gaming. This includes the game industry's pump Indie.fun and the Agent launch platform Fullmetal. Users create games through Moddio, raise funds on Indie.fun, and then deploy Agents through Fullmetal, forming a closed-loop ecosystem.
The Agentic Metaverse/Game is not far from us. The Agentic Game has addressed the initial player shortage and attention economy issues in Web3 gaming. For the attention economy market in Crypto, the Agentic Engine lowers the creation threshold for games, allowing the development of new games to be completed at a faster pace when players and users shift their attention, while utilizing continuously evolving Agents to provide a gaming experience closer to human society.
The reward model of CryptoAI enables interoperability across the multiverse.
After the scene layer is built, it is time to focus on surrounding workflows that can actually generate benefits from the scene. The first is the ML field of gaming. GamerBoom, as a data layer incubated by Binance, merely overlays a layer of 'film' over the original gameplay process. Whether you are playing STEAM games, Apex, or Dota, you won't feel any operational or display changes. GamerBoom simply collects game feedback and visuals, using this as a dataset, and any game company with demand can purchase it, whether in Web2 or Web3. Players experience almost no change but can gain some benefits, representing the Play to Earn concept that only emerges in the Agentic era.
In contrast, ARC Agents have established a complete ecosystem to accelerate the realization of Artificial General Intelligence (AGI). They have currently structured four blocks that represent the ideal ecosystem for Agentic Game in this cycle. They develop continuously learning and highly adaptable AI models for data collection and monitoring systems to improve AI agents continually, create test environments ranging from single-player to multiplayer cooperation, and integrate diverse high-quality data for reward design and state space experimentation. Subsequently, they apply the dynamic rule data trained under the complexities of the real world in games to scenarios involving AGI 'robots, new types of games, and multimodal AI.'
The surrounding industries of Agentic games will also develop due to this, and even these industries may be contracted by AI. Recently, Virtuals announced a collaboration with avariksaga to launch real-time live streaming games, commentary, and narrative functions for AI Agents. Meanwhile, the founder of Zailgo, Miles, a member of Trustless, announced he would be the first to create the next generation of live entertainment experiences—Corporealai, a war of artificial intelligence robots. He claims that partnerships with well-known companies in the fields of robotics, AI, and live entertainment will bring a new mode of e-sports, allowing Agentic AIs with militarized robotic bodies to compete in a Faraday cage arena. Perhaps in the near future, Iron Fist will truly appear beside us.
Image source from the internet.
The mass adoption of Agentic games has already arrived.
For the Crypto ecosystem, the mass adoption of Agentic Game has already taken shape, continuously producing AI Agents that serve as ammunition. The multimodal framework connecting multiple AI Agents not only allows them to form functions similar to workflow aggregation but also enables them to communicate and assist each other across games, platforms, and social media. The game engine allows AI Agents to exist in the virtual world in the form of self-will or interact with participating players as NPCs. All these data can be collected into high-quality datasets through reward mechanisms, fueling new games or AI robot learning. Each link extending out of this process, and the asset integration within CryptoAI, allows the originally unsustainable attention-creating GameFi mechanism to be re-packaged with an intentional economic system.
According to a research report from Messari, the AI narrative of this cycle has not yet reached expected market share compared to previous cycles' popular narratives, meaning that the complete narrative process has not yet been fully realized. In this cycle, computing power is property, while data is power. Unlike previous cycles, the current era of CryptoAI narratives may yield market scales far exceeding previous narratives, as its influence extends beyond just the cryptocurrency market but to the entire world. The Agentic Game is like the Falcon9 rocket, capable of carrying this market to unprecedented heights. At this moment, we are at the intersection of two cross-era significance fields.