Eliza OS Founder Announces Termination of the ai16z Token Project

Original author: David, Shenchao TechFlow

ai16z, do you remember this name?

If you remember, you experienced the AI coin boom at the end of 2024. If not, that’s normal—this category of AI Agent tokens has been quiet for a long time.

Crypto and AI now seem like a pseudo-concept and an outdated toy in an era where AI tools are rapidly advancing and being practically implemented.

Today, Shaw Walters, founder of the Eliza OS project, published a lengthy article stating that the ai16z token is completely dead, the foundation has shut down, there will be no buybacks, and token holders must handle their holdings individually.

He suggested you either sell or "get a group of people to pump it," but don't expect him to do anything else.

The tone of this letter doesn’t resemble a founder gracefully announcing the termination of a project—it sounds more like someone storming out in disappointment after yelling at everyone. He referred to the crypto community as “a bunch of spoiled crybabies” and claimed that a law firm had sued him on behalf of token holders.

The Eliza OS Foundation had no money to fight the lawsuit and lost all of its remaining funds.

Shaw also said he once had $25 million worth of AI16Z tokens in his wallet, held onto every single one until they became worthless; now he lives off his savings in a run-down little bedroom in San Francisco, coding every day.

From his perspective, he may feel that he was developing with idealism in a chaotic market, only to be left with a deep sense of ruin, disappointment, and anger.

I have no intention of judging these personal emotions. But standing at the crossroads of shifting trends, industry pivots, and the rise of AI, the shutdown of Eliza OS is a poignant reminder that crypto meets AI truly got up early but arrived too late.

In October 2024, ai16z launched a crowdfunding campaign on DAOs.fun, aiming to raise 420.69 SOL, equivalent to approximately $75,000, to create an AI-governed investment fund.

420.69—anyone familiar with crypto culture knows this is an inside joke number. From the very choice of fundraising amount, the tone of this whole thing was already set.

But it really took off. In less than three months, ai16z’s market cap surged to $2.6 billion. It also sparked an entire category—GOAT led the way, followed by VIRTUAL, with a flood of agent-themed meme coins emerging. The entire AI Agent sector grew from nothing to nearly $10 billion.

Cryptocurrency is conceptually a step ahead. What was the state of the AI market at the end of 2024?

ChatGPT has just turned two and is often criticized for its various hallucinations; Claude still lacks tools to directly interact with computers, and most people’s understanding of “AI Agent” remains at the conceptual level. Yet the crypto market has already assigned a price to this concept.

But there was one detail that not many people cared about at the time.

Less than a week after ai16z launched, crypto media published reports questioning whether the project’s core AI Agent—the so-called “Marc AIndreessen” claimed to make autonomous investment decisions—was actually being operated by humans rather than functioning as a true autonomous agent.

This same skepticism later arose with the well-known crypto market analysis agent, aiIxbt.

At the time, such skepticism struggled to generate any ripple or FUD. Market capitalization continued to rise, the community kept cheering, and new AI agent tokens kept being launched... Looking back, this may have been the most precise microcosm of the entire crypto-AI narrative.

Truth or falsehood doesn’t matter. In an environment of liquidity spillover, a leading narrative alone is enough to spark a speculative frenzy.

The concept is faithful, the direction is correct, but tokens exist before the product, and price precedes technology—this is precisely the essence of crypto’s head start.

In 2026, AI agents are here.

In the West, there are CodeX and Claude; in the East, there are competitors like Workbuddy that are better suited to the domestic environment. Looking back at the hype around crypto AI, with its whitepaper narratives about automatically analyzing market trends, managing workflows, and monitoring public sentiment, these features have now become a reality.

This may be one of the few real-world use cases in the crypto industry, aside from stablecoins.

But the people cashing in have changed. It’s Anthropic, it’s OpenAI, it’s the teams writing code in San Francisco and Hangzhou. Money flows to aggregation platforms like OpenRouter, to cloud providers’ GPU bills, and to every actual API call that runs.

Another sentence from Shaw’s lengthy article evokes endless reflection:

“There are no such issues on the AI side. The people there are optimistic and empowered, building the future instead of tearing each other down. Those are my people.” The final conclusion from someone who once ignited the narrative around crypto AI agents is that this industry is incompatible with AI?

Objectively speaking, these successful AI applications or large models currently have no reason to issue tokens.

Anthropic charges via subscriptions and API calls, while OpenRouter processes over $100 million in annual transaction volume as a middleman. These companies have never issued tokens or built DAOs, yet their business models work—developers vote with their feet, and the bill is the best consensus mechanism.

Looking at it from the opposite perspective, the situation for crypto AI projects is brutal. When the token price holds up, everything seems to be in place—treasuries are full, communities are active, and the narrative is trending. But once the price collapses, everything vanishes at once.

The developers left because the purchasing power of the tokens used for salaries had shrunk by 90%, and the community turned into a group demanding rights. Those who wanted to work seriously may ultimately find that the project can’t even afford to hire a single engineer.

Ironically, the Eliza framework he holds is not dead. The code is still being updated, and companies are negotiating partnerships on it. An open-source AI framework doesn’t need to issue tokens to survive—whether it earns revenue through services or gets acquired, its pathways are far broader than those of a token.

More ironically, the open-source AI framework Eliza, which originated in the crypto space, had Claude as its most active contributor just last week.

An AI developer or project without a token may be worth more in the 2026 market than one with a token.

As a crypto professional, it's bittersweet to see once-popular projects fall one by one.

But feeling nostalgic doesn't mean being entirely pessimistic.

The crypto AI narrative has certainly fragmented. But what’s broken is the path of issuing tokens for AI concepts—the potential between crypto and AI has not been fully severed.

The few remaining projects are doing something entirely different from AI16z. For example, Bittensor is building a decentralized AI compute marketplace, while Render is creating a distributed GPU rendering network.

All of them use tokens to coordinate real computing resources; there is still some market, but it’s more valuable and persuasive during a bull market.

In addition, AI is increasingly being applied and making an impact in scenarios such as on-chain data analysis, vulnerability scanning, whale wallet tracking, sentiment monitoring, and DEX contract strategies.

In 2026, a professional crypto trader opening Hyperliquid will likely have a Claude or GPT-powered strategy window open beside it. This means the relationship between crypto and AI isn’t over—it’s just reversed.

AI empowering all scenarios has become a political correctness, and cryptocurrency is no exception. As for how much potential remains in the crypto space, that may depend on when BTC finally recovers.

On a personal level, what you should do is embrace change. A developer who lets go of token-related burdens is worth far more in today’s AI world than any token could ever provide; similarly, a crypto enthusiast who genuinely embraces AI can achieve far greater returns than in the old, unrefined days.

Being in crypto often means seeing the direction ahead of others—or perhaps we’ve simply misunderstood where we should stand.