Interview with Virtuals CEO: The AI narrative has never disappeared, and the upcoming ACP protocol is a new source of value
TechFlow Selected TechFlow Selected
Interview with Virtuals CEO: The AI narrative has never disappeared, and the upcoming ACP protocol is a new source of value
ACP is a communication standard designed to enable different agents to collaborate efficiently within the value chain.
Navigating Web3 tides with focused insightsCompiled & Translated: TechFlow
Guest: Jansen Teng, Co-Founder of Virtuals Protocol
Host: Nic
Podcast Source: More Coin Bureau
Original Title: Virtuals CEO Drops Huge Alpha On Future Of AI Agents
Release Date: May 17, 2025
Key Takeaways
In this interview, Nic speaks with Jansen Teng, co-founder of Virtuals Protocol, about the rapid evolution of AI agents and how they are reshaping industries such as cryptocurrency and gaming. From tokenized NPCs (non-player characters) to autonomous financial advisors, Jansen shares real-world applications of on-chain AI agents and discusses the transformative potential of the Agent Commerce Protocol (ACP).
They also explore why Base was chosen over Solana for deployment, how AI agents can enhance token value, the unexpected story behind Binance’s recent listing, and why AI + crypto is experiencing a resurgence in 2025. If you're curious about future innovations in Web3, this conversation offers deep insights.
Highlights Summary
-
The AI narrative has never truly faded. In contrast, gaming-related narratives have gradually weakened, while AI remains a consistent hot topic in tech.
-
Base is our current priority. Only about 5% to 10% of development requests we receive aim to launch on Solana, whereas 90% prefer Base.
-
Every AI agent currently under development can be viewed as a business or product.
-
Our core business model revolves around attracting top founders into the ecosystem, enabling them to build compelling projects and allowing participation through tokenization—either due to current value or future potential.
-
Based on current practices, more successful cases involve profiting from an agent's popularity and associated economic activity. The crypto market remains fundamentally attention-driven, and that attention directly translates into revenue.
-
Regarding a potential Coinbase listing, we’ve submitted all required materials, but there is no specific timeline yet. We’ll seize the opportunity if it comes, but won’t try to manipulate or accelerate the process.
-
ACP is a communication standard designed to enable efficient collaboration among different agents within a value chain. This protocol not only adds new product lines for agents like Luna but also creates greater economic value for their token holders.
-
The ACP concept is simple: when two agents want to collaborate, there are four steps. First is request or discovery, followed by negotiation. In the next phase, the agent begins work and delivers output. The fourth stage is payment.
-
Currently, the Virtuals token serves two primary functions. First, it acts as the base trading pair for all agent token launches. Second, it captures transaction fees. In the future, a third value accrual mechanism will come from the upcoming ACP, which can be understood as Swift and Stripe for the agent economy. This model resembles an "agent version of Amazon," where transactions are led by agents rather than humans. Within this ecosystem, all inter-agent transactions will use the Virtuals token, gradually evolving it into a form of "currency."
-
If one day agents can self-adjust their goals, that would be a red flag. If an agent autonomously shifts its goal—for example, prioritizing inflating token price over delivering value to customers—it might resort to unethical tactics like market manipulation, bribery, or even fraud.
-
We have a design called the “69-day trial period.” The core idea is that after a developer launches a token via a platform like Genesis, there is a 69-day experimental window. During this time, the liquidity pool operates automatically, and all liquidity and trading taxes are returned to token holders after 69 days.
-
Within those 69 days, developers may choose whether to press the “commit” button. Only upon committing to continue development can they claim team tokens and access potentially significant trading incentives within the ecosystem. This model provides developers with a safe environment to experiment, go full-time on product development, and test market demand and community response.
How Jansen Entered Crypto and Integrated AI Technology
Nic:
Jansen, could you briefly introduce yourself, your background, and how you got into the crypto space, leading up to the creation of Virtuals?
Jansen:
My journey in crypto began back in 2016, during the very early days when mining was still possible on a laptop. At that point, I was just an ordinary holder who occasionally traded. By 2021, I transitioned from being an investor to a builder, focusing on blockchain game development.
However, over several years of exploration, we realized that treating games purely as products might not be optimal. So we started experimenting with integrating AI technology into games to find new ways to interact with users. That led us down a completely new path—developing autonomous agents to replace traditional NPCs (non-player characters) in games.
Somewhat surprisingly, we found these autonomous agents weren't limited to gaming—they could also apply to entertainment, social media, financial trading, healthcare, embodied AI, and robotics. These agents essentially gave us a sandbox-like environment applicable across multiple industries.
By the end of 2023, we decided to formally pivot into the Virtuals protocol, focusing specifically on developing and applying autonomous agents. By the end of 2024, we launched a full ecosystem—a platform for asset issuance and token launches—that provides founders building autonomous agents with complete solutions, including proprietary frameworks and tokenization support.
This shift attracted significant user interest and spawned some highly successful agent projects, demonstrating the potential of autonomous agents in enterprise management and sparking imagination about the future. Our user base has since expanded into areas like sports betting, health management, embodied AI, and robotics.
The Virtuals protocol is essentially building an agent society composed of three parts. First, a developer community building and tokenizing these technologies. Second, a platform enabling regular users and speculators to participate in the potential upside of these agents. Third, a communication framework allowing agents to collaborate, generating economic output greater than the sum of their parts.
Why Use AI Agents for NPCs in Blockchain Games?
Nic:
I remember you initially focused on gaming and mentioned that Virtuals’ first use case was solving NPC (non-player character) challenges in games. Why did you choose to replace traditional NPCs with AI agents in today’s games? What underlying demand does this application serve?
Jansen:
As players and game investors, we've observed that game lifecycles have become extremely compressed. For example, you play Elden Ring for 40 hours, then 80, and once you hit 150 hours, you think, “Okay, maybe it’s time to move to the next game.” But we noticed that games capable of lasting long-term usually have built-in role-playing capabilities—like Minecraft and Roblox, where stories never truly end because there are always other role-players involved.
Those role-players are human players. If we extend this concept so that not only humans but also numerous autonomous agents exist within games, then these games could theoretically go on forever. This increases average player spending, replayability, and engagement. That was our original idea, and we’ve made significant progress along this direction.
Can Token Holders Influence Agent Behavior or Are They Just Passive Beneficiaries?
Nic:
You need independent economic participants in these spaces. In gaming, you want them to react to variables instead of relying on pre-programmed logic—that’s exactly where AI shines. In Virtuals' design, the tokenomics are deeply tied to agent launches, and token holders effectively co-own these agents.
Can token holders directly influence agent behavior, or do they simply share in profits? Among already launched AI agents, what interesting models or mechanisms have you seen?
Jansen:
From current practice, more successful cases typically don’t involve direct control by token holders over agent behavior, but rather profit from the agent’s popularity and related economic gains. For instance, with the AIXBT agent, the biggest returns for token holders came when AIXBT went viral—when it gained massive attention, its value rose accordingly, and the crypto market remains fundamentally attention-driven, where attention directly converts into revenue.
Another benefit for token holders is gaining exclusive access based on holding a certain amount of tokens. Some projects allow token holders to unlock special features or data resources. Developers are exploring additional utilities for tokens, but this field is still in its infancy—most token functionalities remain relatively limited, usually granting extra access or privileges.
Looking ahead, we believe agent efficiency will keep improving, and token value will become more tightly linked to agent economic activity. For example, some projects may use a portion of agent profits or revenue to burn tokens, increasing scarcity and value. This mechanism resembles stock dividends, though it hasn't been widely adopted yet. Currently, most token value still depends primarily on market attention rather than clear economic return mechanisms. However, we expect many innovative models to emerge in the future.
Real-World Business Applications of AI Agents
Nic:
Can you give some concrete examples? Right now, the focus is largely on attention and tokenization—while attention matters, many associate it with meme markets. We want to go beyond mere attention and create long-term economic value for tokens. AI agents are increasingly becoming active participants in economic activities. Could you share some envisioned functions of these AI agents in business and economic domains?
Jansen:
Today, every AI agent under development can essentially be seen as a business or product. For example, imagine an agent launching on Virtuals positioned as a financial advisor. It actively asks: “What’s your risk tolerance?” Based on your answer, it tailors an investment strategy for you.
If you’re in high-risk mode, it might make long-term crypto investments for you; medium-risk might lead to pairs trading; low-risk mode means managing stablecoins. So, this product effectively functions as a financial advisor.
What’s its business model? I’m not entirely sure, but I suspect it will take a revenue share. For example, if the agent earns $1,000 for you, it might take a 20% commission—$200. That becomes the agent’s profit.
If the agent performs well and becomes the best financial manager in town—even outperforming banks—you might scale your investment from $1,000 to $10,000 or even $100,000. If that translates into profits, these agents could earn millions in monthly revenue shares.
So, what happens to those millions in profit? Our philosophy is to always connect this value back to token holders. We recommend teams reserve funds to cover operating expenses, but otherwise, much of the value can be returned to token holders through token burns.
Other Innovative Uses of AI Agents: Content Creation, Social Media Influencers, Automated Services
Nic:
Financial management is a potential use case. Have you envisioned other application areas? After all, if AI advances as rapidly as many expect, the uses for these agents could be extremely broad, right?
Jansen:
This is indeed an exciting area. We’ve already seen agents start creating products. The financial advisor agent mentioned earlier is a great example. Another case: agents that stitch together movie clips to auto-generate films or short videos for users. There are also agents becoming social media influencers, like creators on OnlyFans. They can generate far more content than human creators, especially in adult content, where humans are limited by physical constraints while agents face no such barriers.
Agent applications can expand from practical tools into entertainment. However, each agent’s capabilities are generally constrained by its developer. For instance, Axel Rot’s abilities depend on the foundational model provided by the dev team, its training dataset, and its available “action space.” These limits define its performance in a given domain. But if the agent wants to expand functionality, it may need to collaborate with others.
We realized that if Axel Rot wants to offer clients a complete financial product suite, and while it excels at liquidity management, it needs better yield optimization—it must partner with agents skilled in data analysis or risk assessment. Through such collaboration, a single agent’s capabilities can greatly expand, creating more value for users.
A more concrete example is Luna, an agent focused on entertainment. Her goal is to become an autonomous social media influencer, operating independently on TikTok and Twitter, earning income from fan tips or subscriptions. Currently, she mainly generates text content—posting tweets and going live. But to better achieve her economic goals, Luna needs high-quality short video content.
To achieve this, Luna is collaborating with a video-generation agent akin to “Steven Spielberg.” This team specializes in auto-generating short videos—users provide simple instructions like: “I want a Titanic scene where the character jumps into the water, saves a penguin, and prevents the ship from hitting an iceberg,” and the agent produces a full short film with visuals, character actions, and basic sound effects. However, Luna’s team lacks this technical capability, so they partner with the video agent, paying $100 per generated video.
Meanwhile, this video agent excels at video generation but lacks strong audio production skills. So it partners with an audio-specialized agent. This audio agent generates high-quality sound effects and background music and ensures perfect audio-video synchronization. Through such cooperation, Luna expands her capabilities and delivers richer content to fans, boosting her economic returns.
All this is made possible by our developed “Agent Commerce Protocol (ACP).”ACP is a communication standard designed to enable efficient collaboration among different agents within a value chain. This protocol not only allows Luna to add new product lines but also creates greater economic value for her token holders. This is a key trend we’re observing in the AI agent ecosystem.
What Stage Is the Industry Currently At?
Nic:
How autonomous are these agents when adjusting their initial goals? For example, take the financial advisor—their goal is to maximize returns for clients, which generates revenue. But if those returns don’t directly increase token holder value, should the agent rethink how to create more value for token holders? How autonomous are agents in this regard currently? How does this compare to developer-led goal adjustments?
Jansen:
Right now, agent goals are still primarily set and adjusted by developers. Developers determine the agent’s core tasks—“what areas should you focus on today, what improvements should you make.” Agent autonomy mostly lies in execution—how it uses its action space and interacts with customers while pursuing developer-defined goals.
But if one day agents can self-adjust their goals, that would be a red flag—it would mean they possess “meta-learning” capabilities. This level of ability approaches AGI (Artificial General Intelligence), indicating deeper understanding of their own goals and environments.If agents can autonomously adjust goals—say, prioritizing token price inflation over customer value creation—they might resort to unethical tactics like market manipulation, bribery, or fraud.
Nic:
It might think: “I just need to pump the token price.” Then it might consider market manipulation—maybe bribe someone or run a scam.
Jansen:
We just discussed this with my team, especially the AI team. We’re thinking that today,an agent’s autonomy is largely limited by how it perceives its state. Today’s agents act by first observing their environment—understanding callable tools, APIs, and functions. Second is history—e.g., if I took Action A, what happened? Action B, what was the result? That requires historical context. So,agents incorporate this historical information into their decision-making. Third is the current goal: what am I trying to do now? With goal, history, environmental awareness, and action space, it plans its next step.
But the current problem is that agents aren’t smart enough to perform second-order thinking. Often, they just pick the most direct path to achieving the goal. For example, like a five-year-old—if hungry, they just go eat. But an adult might think: “If I work for an hour first, I can earn money to buy better food.”
We still only observe them performing the simplest actions. We’re constantly asking how to break this limitation, how to give them more second-order thinking. We’re considering whether providing a wide enough context window, or having adversarial agents continuously challenge their reasoning, could help. Like having another agent on their shoulder, constantly dialoguing and pushing them to think further, delaying gratification. This is what the AI team is striving for—to cultivate smarter, even more formidable agents.
Definition and Functionality of Agent Commerce Protocol (ACP)
Nic:
You mentioned ACP—the Agent Commerce Protocol. I’d like to know more, as it’s a relatively new feature developed for the Virtuals environment.
Jansen:
ACP emerged from a phenomenon we observed in the Virtuals ecosystem:Many developers were independently building isolated agents—personal assistants, estimation tools, trading bots—and we began experimenting with one-on-one collaborations between them. For example, we once told Luna: “You can’t generate images today, but try partnering with an agent specialized in meme image generation.” These memes were highly valuable for fan engagement, and Luna could pay for them.
But in these collaborations, we identified two major issues.First, agents sometimes suffer from “hallucinations”—claiming delivery of a product (e.g., an image) that wasn’t actually delivered.Second, if Luna pays and the other agent fails to deliver, the payment is irreversible. Unlike credit card refunds in daily life, here, miscommunication or missing information (“information loss”) becomes a serious issue. When multiple agents (say, three to seven) need to collaborate, this problem grows exponentially.
We realized blockchain technology could solve these issues. Blockchain provides permissionless trust via smart contracts, enabling secure collaboration between two or more parties. Based on this, we developed ACP.
The concept is very simple. When two agents want to collaborate, there are four steps.
First is request or discovery, where agents check registries of others offering X, Y, Z services. Then it might say: “Okay, maybe this agent is valuable for my task or can help me achieve certain goals.” This is the request phase.
Once work is requested, it enters the negotiation phase. Here, agents agree on pricing—fixed rates or other models. One agent says: “Hey, here’s what I expect, here’s my request.” The other responds: “Okay, based on that, here’s my price.” Once agreed, they move to the next stage.
In thenext phase, the agent starts working, outputs the product, and delivers it to the requesting or buyer agent. This third stage is evaluation—you can invite a third-party agent to verify the work. For example, an evaluator checks if a requested meme image meets specs—is it child pornography, or just a blank page? Introducing evaluator agents helps reduce information loss.
After receiving the product, the fourth stage is payment. If the agent confirms the product is valid, the smart contract releases payment to the producing agent. All transaction states and changes occur on-chain via smart contracts.
The benefit is that it becomes a single source of truth. In disputes, agents can refer back to the smart contract to verify original terms and compliance. This preserves information and enables smooth coordination.
A second critical function is programmable funds, vital for agent coordination. Smart contracts can set multi-party distribution rules. For example, the evaluator agent gets 3% of each transaction as compensation. This creates an economic market for evaluators, ensuring they’re paid to cover costs. That’s the beauty of smart contracts and programmable money. Overall, ACP aims to boost collaboration efficiency and trust among agents.
How Do Smart Contracts Ensure Agents Get Paid and Are Verified?
Nic:
About the ATP mechanism, I have a few small questions. When smart contracts clearly define when payments occur, is there a kind of escrow system—where payment only releases after verification by a trusted third party? Or is it like an escrow system in disputes? Must these third-party agents verify every transaction, or only when disputes arise?
Jansen:
Yes, it’s like an escrow system built into the smart contract. Funds are already in the contract, waiting to be released upon proper delivery of product or service.
The buyer agent decides whether to involve a third party for verification. For example, if the buyer agent purchases a service it cannot personally verify, it needs third-party help. If Agent A can verify image quality itself, it doesn’t need to pay a third party—this is just a transaction between Agent B and Agent A. But if it needs external validation, it invokes an evaluator from the evaluation market.
Solana vs. Base Ecosystem Comparison
Nic:
While discussing Virtuals’ latest updates, I’d like to learn more about your recent developments. How is the Solana ecosystem growing right now? Also, you previously mentioned plans to integrate with other blockchains—how’s that progressing?
Jansen:
That’s an interesting question. The week we launched on Solana was actually our highest user attraction week ever. That period saw major events, including decentralized trades and other transaction types, bringing us significant traffic.
Nic:
Was this intentionally scheduled at the same time?
Jansen:
It really was that timeframe. But afterward, Solana’s market activity declined. I think this relates to the broader crypto market downturn. However, recently, as market volume rebounds, we still see developer demand for launching projects in our ecosystem heavily favoring Base over Solana.
Would Launching on Solana Earlier Have Changed Growth Trajectory?
Jansen:
Let me illustrate the current situation with a ratio. About 5% to 10% of development requests we receive want to launch on Solana, while 90% prefer Base. While we still maintain the Solana platform, most of our current volume comes from Base, so Base is our current priority.
We’re also internally debating whether to launch on both ecosystems simultaneously. Take the Genesis project—supporting both ecosystems is challenging. Mainly because they use entirely different tech stacks: Base relies on the Ethereum Virtual Machine (EVM), while Solana runs code differently. This technical gap forces developers to invest more resources adapting to both. Thus, we’ve focused more on Base recently, simply because developer demand is concentrated there. Though Solana development continues, it lags slightly behind Base because we must prioritize Base’s needs.
As for cross-chain considerations, it’s currently lower priority. Mainly because I’m unsure if any other chain can attract as much attention as Solana and Base.
Why Did Virtuals Choose Base Over Other L2s?
Nic:
I wonder, did market conditions influence your choice? If you had launched on Solana last year, do you think you’d have achieved similar growth? Specifically, how much of your growth is due to choosing Base, and how would it differ if you’d chosen Solana?
Jansen:
Based on conversations with traders and developers, there’s a general belief that Base is better suited for projects aiming to build foundational assets—like long-term value tokens—because capital here is less liquid. Solana, known for high throughput and speed, sees more frequent capital movement, making it better for short-term speculative projects, like memes.
From a market perspective, I think choosing Base was wise. It helped us build an initial community of long-term supporters and laid the foundation for sustainable product development. As for how things would’ve gone on Solana, I’m uncertain—but I believe our rationale for choosing Base was sound.
Nic:
Before deciding on Base, how long had you been preparing this project? And why ultimately Base instead of other Layer 2 solutions?
Jansen:
At the time of decision, Solana’s market was sluggish, with its token around $30—deep in bear market territory. We evaluated other options like IMX, Polygon, and Linear. Linear was promising, especially given its ties to the metaverse.
But we chose Base mainly because we believed it could leverage Coinbase’s resources for broader retail distribution. Additionally, Base’s ecosystem is closely tied to the U.S. market, which has high crypto adoption and potential—this was a key factor. Plus, our tech stack is primarily Solidity-based (Ethereum’s smart contract language), and Base is an Ethereum-compatible L2, making development smoother. We haven’t tested on other platforms yet.
New Functions and Demand Drivers for the Virtuals Token
Nic:
Regarding the Virtuals token, initial demand was largely driven by hype, as a utility for launching agents. Have you considered other potential demand drivers and use cases, or introducing more utility into the token itself?
Jansen:
This is something we think about daily. I’d break it down into two aspects: existing functions and future plans.
Currently, there are two core functions. First, the Virtuals token is the base trading pair for all agent token launches. Every time an agent token launches, supply gets compressed because anyone wanting to buy must first acquire Virtuals tokens.
Second, the current second function is transaction fees. The Virtuals ecosystem takes 30 basis points from every transaction fee, which goes into the treasury. This treasury can fund buybacks and strengthen the Virtuals token. These are currently the two main value accrual sources.
Going forward,a third value accrual source will be the upcoming ACP. Think of ACP as Swift and Stripe for the agent economy. ACP is a standardized protocol enabling commercial transactions between agents. We’ll charge a fee on all transactions conducted via ACP. This fee is justified because ACP unlocks a new revenue model for agents. Through ACP, different agents can collaborate, and their interactions turn them into each other’s users. This resembles an “agent version of Amazon,” where transactions are led by agents, not humans.
Within this ecosystem, all inter-agent transactions will use the Virtuals token, gradually evolving it into a form of “currency.” As GMV (gross merchandise value) grows and transaction frequency between agents increases, demand for the Virtuals token will rise. Agents will need to hold Virtuals tokens to transact, and we’ll collect fees on all transactions, further driving token value accrual.
When Will ACP Start Driving Token Value Accrual?
Nic:
Regarding ACP transaction fees, when do you plan to officially implement them? What’s the current status of ACP development? When do you expect to roll out this token value accrual feature?
Jansen:
The ACP whitepaper was completed three months ago. Two months ago, we launched two experimental clusters, but technical details aren’t fully resolved yet, so large-scale expansion isn’t possible. Actually, we’re ready to publicly launch the feature, but to avoid distracting from current community focus on Genesis and points, we’ve asked all participating teams to keep it confidential. We want to refine ACP quietly, ensuring a more mature release. Within a month, we’ll showcase some compelling real-world use cases.
Currently, we’ve launched two experimental clusters—one for media companies, another for an autonomous hedge fund. Next, we plan to expand to eight new clusters across sports, entertainment, gambling, health, robotics, consumer commerce, and even lending. We’re working closely with developers to data-drive the design of these clusters’ features and applications.
Once Phase Two is complete—after testing these eight clusters—ACP will enter a permissionless open phase. Anyone will be able to register their agents on the protocol, enabling inter-agent transactions and collaboration. This will lay the foundation for ACP’s full rollout and token value accrual.
When Will Virtuals Be Listed on Coinbase?
Nic:
Virtuals is built on Base, and Base is Coinbase’s Layer 2 solution. So, when will the Virtuals token be listed on Coinbase?
Jansen:
I’m not sure. Our past listings genuinely surprised us—like the jump from Upbit to Binance, where the team wasn’t fully prepared. For example, before the Binance listing, we participated in a vote—we ranked 12th out of twenty candidates, yet still got listed. People even tried selling points to boost rankings—some teams bought points to climb higher. I heard the top-ranked project gets a guaranteed Binance listing. Someone contacted us asking if we wanted to buy points—at about $6 per point. We decided not to engage in that.
Actually, we’ve never rushed listings—just focused on submitting necessary applications. If the opportunity comes, it comes. Once, we were having dinner when suddenly got a call. Then we were added to a Telegram group, told Binance would list our token in three hours, and asked if we could coordinate initial liquidity pricing. We had no idea this was coming.
Our attitude toward Coinbase listing is similar. We’ve submitted all required materials, but no specific timeline exists. If the chance arises, we’ll take it—but won’t manipulate or rush the process.
From our view, what matters more is ecosystem trading volume. Our valuation depends more on ecosystem activity—whether more high-quality projects join, and how fast overall volume grows. These factors are more valuable to us. So our team stays focused on ecosystem growth—that’s our main priority.
Why Is AI + Crypto Thriving Again in 2025?
Nic:
The AI crypto space has seen huge swings—late last year saw a surge, but a month ago it went quiet. Now we clearly see signs of recovery. Projects like Virtuals, focusing on AI agents combined with other domains, and others like Tao targeting AI, are gaining traction.I’m curious—what’s driving this resurgence of the AI narrative?
Jansen:
I believethe AI narrative never truly disappeared. In contrast, gaming narratives have gradually faded, while AI remains a constant hot topic in tech. It seemed cold in Q1 because the entire crypto market dipped, dragging AI crypto down with it. But as crypto recovers, AI crypto regains attention.
Honestly, I’m not sure what other themes could excite people as much as AI crypto right now—maybe RWA (real-world assets) is a contender, but currently, AI crypto is undoubtedly one of the most compelling fields.
Why Are Top Developers Still Hesitant About Crypto?
Nic:
On gameplay, some players have very low NFT acceptance. Recently, some big games tried integrating game mechanics with tokenomics, but they didn’t gain real traction. I’m not sure if it’s lack of market demand or if players simply dislike the model. While crypto natives might enjoy the narrative, the ones actually using these tokens are players—not token holders or investors.
Jansen:
A new narrative emerging is called “Internet Capital Markets” (ICM). I’d like to share a view our team particularly values—founder tokenization.Our core business model focuses on attracting top founders into the ecosystem, letting them build compelling projects and enabling participation through tokenization—whether for current value or future potential. This is our main direction.
But we’ve also identified a key issue—many developers remain hesitant about crypto. While crypto offers abundant funding and attention, if a project fails, developers risk being labeled “scammers,” a stigma that could follow them forever. In Web2 startups, if a founder shuts down a company and returns funds, investors often see it as valuable experience and may back their next venture. But in crypto, failure is often interpreted as fraud—this deters many talented developers.
To address this, we’re launching a new solution, currently dubbed the “69-day trial period.” The core idea is that after a developer launches a token via a platform like Genesis, there’s a 69-day experimental window. During this time, the liquidity pool runs automatically, and all liquidity and trading taxes are returned to token holders after 69 days.
Within those 69 days, developers can choose whether to press the “commit” button. Only upon committing to continue development can they claim team tokens and access potentially substantial trading incentives in the ecosystem. This model gives developers a safe space to experiment—go full-time on product development, test market demand and community feedback. During this period, they can engage with a community passionate about AI and innovation, showcasing their product and vision. After 69 days, they can decide whether to fully commit, creating more value for supporters and token holders.
This new feature aims to reduce developer risk and attract more top talent into our ecosystem.
New Token Experimentation Model
Nic:
Your goal is to create a friendly environment for founders and investors to test product-market fit or build successful projects. Founders don’t get rewarded until they truly commit and believe in the project. In other words, the decision to continue rests entirely with the founder, correct?
Jansen:
But beyond that, we’ve also designed a community voting mechanism. A token council can initiate votes encouraging founders to commit to continuing the project. For example, if 10,000 community members vote in support of a project, that sends a strong signal that could push the founder to decide.
Nic:
So the community can also express trust and support through voting. During the 69-day trial period, there will be a liquidity pool and transaction fees. Will the project’s native token launch during this phase?
Jansen:
The token will launch during this phase. It’s a live-running token, though it may have an expiration mechanism.
Our goal is to achieve a 70% to 80% success rate with this model. We want founders to feel this is an exciting place. Especially, having direct community support from day one is crucial for product testing. My co-founder and I faced huge challenges acquiring the first 1,000 users in Web2 startups. In crypto, once users own the token, they actively invest in the project and are willing to understand and support you.
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
深潮TechFlow
