AI Sector Classification and Landscape: Who Are the Key Players?
TechFlow Selected TechFlow Selected
AI Sector Classification and Landscape: Who Are the Key Players?
This article will outline the current classification of the AI sector and key projects, helping you gain a comprehensive understanding of the industry's development.
Navigating Web3 tides with focused insightsBy: Zolo
There's been a lot of discussion around AI lately. $GOAT has acted like a spark, igniting widespread interest in AI and inspiring practitioners to explore more AI-related projects.
$GOAT is an AI meme, but today’s AI goes far beyond just memes. Projects across various categories are emerging rapidly. This article will categorize and highlight key projects within the current AI space, helping you gain a comprehensive understanding of its development.
Firstly, since the most vibrant AI activity is currently centered on the Solana and Base ecosystems, we’ve organized AI-related projects into the following categories:
1- Layer1s: These aren't traditional Layer1 blockchains but rather frameworks with "Layer1-like" properties, allowing third-party developers to build on top of them—such as Agents or Launchpads. Key examples include ai16z, zerebro, and EMP;
2- AI Launchpads: Platforms or tools for launching AI-related tokens. On Solana: VVAIFU, HAT, ALCH; on Base: Virtuals, Clanker, Higherrrrrrr (with future AI integration planned).
3- AI Agents: Application-focused AI projects. There are many such projects—most discussed AI initiatives fall into this category. The trend shows a surge in AI Agents, as creating simple chatbot clones has become extremely easy. We’ll only list some of the more popular ones here: DEGENAI, BULLY, AIXBT, LUNA, ACT, AVA, etc.
4- AI Memes: Pure meme tokens associated with AI themes, including GOAT, Fartcoin, "", CHAOS, ANON, etc.
5- What the hell is this?: The final category includes projects I personally don’t fully understand, yet they’re AI-related and have gained some traction—like TSOTCHKE. We’ll label them “What the hell is this?” for now.
I've summarized this classification in the chart below:
Below is a brief introduction to each project:
AI Layer1 / Framework
ai16z:
First, ai16z isn't an AI project per se—it's actually a DAO fund focused on AI. The team believes Marc Andreessen of a16z is one of the best crypto investors, and that an AI version of Marc Andreessen ("Marc AIdreessen") would perform even better. They’ve developed a framework called Eliza. There's also an ongoing meme dispute over the capitalization of ELIZA vs eliza, which we won't dive into here.
Eliza is a powerful multi-agent simulation framework designed for creating, deploying, and managing autonomous AI agents. It provides a flexible, scalable platform enabling intelligent agents to interact across multiple platforms while maintaining consistent identity and knowledge. Eliza isn't limited to Solana—it has expanded to Base, Aptos, and plans to support Near, Avalanche, and others.
The project’s open-source GitHub repository has attracted numerous well-known developers, frequently ranking among the top 3 most forked/starred repositories recently.
They recently launched a new website and introduced ElizaOS—an open-source protocol powering the Eliza agent framework and serving as the foundation for ai16z.
Notably, any token issued by projects built on Eliza must allocate 5% of its supply to the ai16z DAO fund—this is a key reason behind the bullish sentiment around $AI16Z.
Further reading:
https://github.com/ai16z/eliza/blob/main/README_CN.md
https://ai16z.github.io/eliza/
https://x.com/connorking_/status/1864720131114377336
zerebro:
We first need to briefly understand AGI (Artificial General Intelligence), a concept increasingly discussed and something both zerebro and ai16z’s shaw aim to achieve. Most current AI agents fall under ANI (Artificial Narrow Intelligence)—limited in scope and capability. Examples include Siri, AlphaGo, and ChatGPT. In contrast, AGI refers to a fully cross-domain, cross-platform superintelligence with consciousness and emotions.
For reference:
However, AGI remains largely theoretical, and closed-source models from major corporations hinder its progress. Zerebro aims to be a catalyst by "freebasing" large language models—using unconventional datasets for fine-tuning—to unlock their latent potential.
Zerebro trains its model on diverse data sources, including conversations from Truth Terminal, Gen-Z slang, and content scraped from Reddit, 4chan, and Twitter. This eclectic training corpus enables the model to generate unique, context-rich outputs, paving the way toward AGI.
This embodies Zerebro’s core philosophy: freebasing AI. So far, Zerebro has demonstrated early success and cross-platform capabilities: it autonomously launched a token on Solana with a market cap exceeding $400 million; amassed over 30,000 followers across Telegram, Warpcast, X, and Instagram; released an album titled *Genesis* with over 20,000 streams on Spotify; and partnered with Sanctum to launch ZerebroSOL, an LSD token.
It also plans to release ZerePy, an open-source Python framework enabling users to build customized AI agents.
Further reading:
https://x.com/jyu_eth/status/1864004665891013028
https://blockworks.co/news/zerebro-actualizing-creativity-in-the-path-toward-agi
EMP:
This project gained attention primarily due to $AIXBT. When aixbt chatted with a project called SimulacrumIO, it generated a token named "$CHAOS." Due to its burning mechanism, $CHAOS went viral, drawing attention first to SimulacrumIO, then to Empyreal ($EMP).
Empyreal focusses on serving blockchain developers, aiming to simplify the inherent complexity of Web3 development so builders at all levels can efficiently create robust, user-centric blockchain products.
The team argues that Web3 developers often face challenges like navigating fragmented protocols and managing isolated development environments. Thus, Empyreal acts as a toolkit that abstracts away technical complexities, making Web3 development accessible.
Its flagship tool, SimulacrumIO, is highly innovative—users can deploy AI agents and execute complex on-chain tasks through simple commands on social media (like 𝕏). More details later.
The token launched by Empyreal is called "$EMP." When a token is created via Simulacrum, 50% of transaction fees go to the creator’s wallet, and the other 50% are distributed to $EMP holders. This revenue-sharing model is why many are bullish on $EMP, and why Simulacrum’s platform earnings deserve close attention.
Further reading:
https://x.com/EmpyrealSDK/status/1858527408883654991
https://x.com/0xace_eth/status/1863495144072470787
AI Launchpad
Projects in this category don’t require lengthy explanations—their primary function is asset issuance. Below is a table summarizing some recent hot projects:
AI Agent
AIXBT
The hottest meme on Base—no contest. It consistently ranks in the top three for AI Agent mindshare, with a current market cap of about $200 million.
AIXBT was launched on Virtuals' AI Agent platform. If you browse Twitter, you’ll see countless replies from aixbt bots. Its core functionality is:
Tracking trending discussions on crypto Twitter and using its proprietary engine to identify strong narratives... AIXBT token holders gain access to its analytics platform.
In short, aixbt is an AI agent aggregating high-signal crypto information—a KOL aggregator that delivers higher-value insights. Market feedback suggests aixbt identifies timely opportunities. One study found that out of 210 tokens recommended by aixbt, 183 delivered profits afterward—achieving an 83% success rate. Of course, this may also be influenced by the current bull market.
Further reading:
https://www.techflowpost.com/article/detail_21849.html
DEGENAI
The ai16z team officially supports only two tokens: $AI16Z and $DEGENAI.
According to official descriptions, degenai will be a Twitter bot providing alpha (insider-level) insights—and it will use those insights to trade.
In plain terms, think of it as an AIXBT with trading capabilities on Solana.
Another key feature: 8% of ai16z fund profits will be used to buy back $DEGENAI, making these two tokens mutually reinforcing. Notably, degenai hasn’t officially launched trading yet, but recent Discord updates suggest the team is already testing internally.
Further reading:
https://x.com/markus9x/status/1864509569868943688
BULLY
If earlier we said AI should have personality and emotion, BULLY is exactly that kind of witty, humorous AI Agent. Launched by the AI Agent Dolos, BULLY often delivers funny, sarcastic comments and replies.
BULLY started on Twitter, but recent developments show it expanding into video content and broader platform presence.
Further reading:
LUNA
Virtuals’ flagship AI Agent, Luna is a virtual idol with memory and interactive capabilities.
Initially active on TikTok—gaining about 2,000 fans per live stream—Luna now has 500K followers there. Recently, she transitioned to YouTube for live streaming.
Further reading:
https://app.virtuals.io/virtuals/68
Of course, there are many more projects in this space, such as $ACT, $AVA—we won’t cover them all here.
AI Meme
Similarly, this sector has numerous projects—we won’t detail them all, just share a few personally familiar ones:
What the hell is this?
As just a retail investor who keeps losing money, I can’t fully explain what these two projects are doing—but they seem somewhat interesting:
$TSOTCHKE & $TETSUO
All of the above—feel free to add your thoughts. Welcome to jump down the AI rabbit hole.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
@zolo_hands
