It has been about 2.5 weeks since the AI agent market hit rock bottom (with a market value of around $4 billion), and the market has now fully entered a bull market led by @virtuals_io. That's right... only Virtuals is in the lead. This wave of market activity is similar to October to November of last year, when Virtuals launched its agent tokenization platform, establishing itself as a pioneer and the first legitimate player, providing a top-tier distribution network for any AI-related project willing to adopt a "fair launch" token model.
The difference this time lies in the latest feature "Genesis Launch" - a fairer way to launch and reward early supporters. This brings out the first trend:
Fair Launch/Gaming Launch Platform
Genesis Launch has completely changed the state of the primary market—from speculative investors blindly rushing into random/small circle junk projects on Pumpfun, to speculators almost certainly making 5-10 times returns when new projects launch.
The achievement of this goal is attributed to the introduction of the "points" mechanism, which coordinates the objectives of different stakeholders, as well as the fixed market value and fixed supply launch model (each project launches with 112,000 VIRTUAL, approximately 200,000 FDV). Each participant can earn up to 0.5 VIRTUAL based on the points contributed, top agent tokens (Virtuals have a dedicated list), or simply earn points by promoting Virtuals.
The recently introduced cooling-off period has further suppressed selling behavior and enhanced the appeal of the Genesis Launchpad, as sellers must now think twice.
As of now, the Genesis Launchpad has proven to be very successful, with @BasisOS being the most successful case, bringing participants a 200x return. Since then, many projects with returns between 5x and 40x have emerged on Virtuals.
In light of the success of Genesis Launch, funds and attention have once again focused on the Virtuals ecosystem, raising the valuation floor of almost all proxy tokens on the platform.
However, despite the resurgence of interest, the lack of quality projects remains one of the biggest challenges facing the Virtuals ecosystem, which leads to the second trend —
Speculation and trading structures take precedence over substance.
As a trader/speculator, you can make money by investing in projects with mediocre teams and average products—this is entirely due to the mechanism design of Genesis Launch. The probability of a project being pumped from a market cap of $200,000 and then dumped is quite high (especially when you know there are no insiders or pre-sale rounds).
As a project party, if you have a unique idea, you can launch a token first, even without a runnable product. There is no need to specify target users, validate market demand, or worry about revenue and user growth. You just need to create the maximum hype and then launch (having a demo version can add points, but it's okay if there isn't one, haha).
Any project that meets the basic requirements (having decent documentation, a good product idea, and a seemingly reliable team) can successfully launch a token at the Virtuals Genesis Launch.
The lesson for investors here is to view these startup projects as short-term speculative opportunities rather than medium to long-term value investments. Because, nine times out of ten, the so-called "underlying AI projects" being hyped are trash. (If interested, I analyzed a typical case in detail on Substack.)
While the proliferation of garbage projects has occurred, it has also brought a wealth of opportunities and gaps for high-quality projects (whether in the AI field or other areas). This leads to the third trend—
Opportunity Window for Quality DeFi Projects
Two months ago, I communicated with the @logarithm_fi/@BasisOS team. I learned about their products (similar to Ethena's Delta neutral strategy, but not relying on stablecoins), and I recognized this team early on when they were deeply engaged in the LPDFi track (products based on Univ3 liquidity). Out of appreciation, I offered them some suggestions for token economic models and launch plans. At that time, I really didn't expect it to rise this much—although DeFi products have fundamental support, the so-called "AI" part is indeed too early. However, in the end, none of this matters, as the project's performance still crushes the market.
Basis's successful case proves that there are blank opportunities for DeFi projects to issue tokens in Virtuals. Although there is a lack of mechanisms in traditional DeFi to incentivize TVL through token emission, leveraging the built-in traffic and attention of Virtuals (especially when you have solid DeFi products) is enough to ensure that your treasury/strategy receives adequate TVL.
If you have a mature DeFi product and want to optimize user experience or enhance core functions through AI, feel free to communicate anytime. I would be happy to exchange ideas.
In addition to the traffic dividend, this also gives rise to new trend experiments similar to the 2023-24 ETH shitcoin era. This leads to the fourth trend —
Using trading volume revenue as a growth engine
Once upon a time, the market was flooded with a large number of small Ponzi DeFi projects that experimented with various token economic models: charging 1-3% transaction fees, using these funds to expand the treasury to maintain the Ponzi strategy, and then returning the profits/interests to token holders. In that era when Ethereum players had ample funds (the frenzy of people all-in on garbage projects is still fresh in memory), project creators could rake in six to seven figures in just one or two weeks—thanks to the scarcity of tokens in the market at that time.
Today, we see similar patterns in the field of AI agents.
Typically, the Virtuals platform charges a 1% fee on each transaction, of which 70% goes to the project creator. Other launch platforms also charge a transaction fee of 1-2% and return 70%-100% to the creators.
@Squidllora, powered by intelligent support from @AlloraNetwork (recently launched on @autodotfun), is allocating creator shares to expand its liquidity pool/treasury, and trading mainstream cryptocurrencies using Allora's predictive models (Allora is like Bittensor in pure financial scenarios—numerous scientists are competing to build the best predictive models for crypto assets across different time dimensions). A portion of the profits generated from trading will be used for token buybacks.
This model is particularly effective for teams with ample funds — they do not need to rely on transaction fees to maintain operations. Now, teams can issue AI agent tokens as a marketing tool to accumulate attention and transaction fees through new user acquisition funnels, thereby providing startup capital for brand new AI experiments.
That being said, looking at the entire AI agent field, despite many teams publicly advancing projects, there are still no other hotspots apart from the Virtuals ecosystem. This leads to the fifth trend—
The dominance of Virtuals / Other ecosystems lack upward momentum
Due to the ongoing promotion of the Genesis Launch, the valuation floor of proxy tokens on Virtuals continues to rise. However, the emergence of this wave of new projects does not imply improvements in fundamentals or technology; its main driving force comes from significant optimizations in trading structures—more people are participating in the games on Virtuals because they know they can profit from it. This trend may continue until the prices of existing projects reach local highs/ceilings.
Once this happens, market attention will naturally shift to other ecosystems, such as @CreatorBid, @arcdotfun, @autodotfun, especially those low market cap projects with obvious improvements in fundamentals (new features, product releases, new partnerships).
Currently, CB and Arc may be the two best ecosystems, among which there are still some undervalued projects that have not yet started a bull market (such as the 3-4 assets commonly found on CB that are integrated with subnets or focused on developing Bittensor-related products, as well as those handshake projects on Arc that contribute to the Ryzome ecosystem).
The best way to seize this opportunity is to position these tokens in advance before most people realize their value.
The AI investment opportunities for institutional investors are limited.
Despite the continuous price rise of mainstream AI agent ecosystems like VIRTUAL and AI16Z, many institutional investors can only watch from the sidelines—because the currently significantly appreciating assets are only suitable for retail/enthusiastic speculators to participate in. These assets have thin liquidity, and the market maker structure is fragile (especially in Virtuals).
The current lack of a sound liquidity infrastructure, coupled with the growing interest from institutions in decentralized AI, has prompted them to shift their funds towards investing in decentralized infrastructure, agent-based Layer 1, and decentralized AI laboratories, rather than the current generation of AI agent tokens.
You may ask: What tokens are favored by these institutions? They include GRASS, TAO (and its subnet), VANA, FLOCK, $PROMPT (which might count), and a series of yet-to-be-released tokens like @NousResearch, @PluralisHQ, @PrimeIntellect—these projects are building a true Web3 AI moat that puts ownership of high-performance models in the hands of the public (rather than centralized AI labs). In short, these are truly complex AI projects that ordinary investors neither understand nor know how to invest in.
How to make the best layout based on trends?
My strategy is to gradually shift the profits from short-term small AI agent trading (especially those Genesis Launch projects with average team levels) into the decentralized AI (DeAI) field. Real AI fundamental projects take time to develop, as they mostly focus on infrastructure rather than consumer-facing products. Just as we have seen ChatGPT, Grok, and Anthropic suddenly excel in everyday tasks, real-time research, programming, and more, the decentralized Web3 model will one day shine in certain Web2 and Web3 native tasks.
Does this mean you should heavily invest in DeAI infrastructure?
Not necessarily. The driving force of the crypto market is speculation and distribution—90% relies on distribution, and 10% relies on the model. For a project to succeed in the long term, the key lies in whether it can execute distribution well: providing products that people need, having a good UI/UX, formulating and executing an excellent launch plan, designing a robust token strategy, and having effective user acquisition and retention strategies, among others.
My core investment logic is still to bet on teams that understand both distribution and technology (models). This is similar to the Web2 venture capital investment approach in vertical SaaS (targeting specific niche use cases, although it may use off-the-shelf models plus proprietary data behind the scenes).
I believe this logic will remain effective in the short to medium term, especially in the Web3 space—because everything here is driven by hype and community, and simple and easy-to-understand concepts tend to sell better. The market/trading structure is also evolving in this direction: the fair launch model is gradually becoming the norm in the market, while more and more teams (in other ecosystems) are starting to closely integrate these two aspects (spending more time collaborating with infrastructure projects to leverage their technologies, while focusing on distribution and GTM strategies to generate buzz for real AI technologies).
Quick summary
The Genesis Launch model of Virtuals has dominated market attention and returns.
Speculation still far exceeds substance - most are short-term opportunities.
DeFi projects with solid fundamentals have found room for growth on Virtuals.
The creator revenue from trading volume is driving a new wave of experimental trends.
Virtuals are leading the way, but other ecosystems may soon attract funding attention.
Institutional investors choose to wait and prefer DeAI infrastructure over proxy tokens.
Layout strategy: Rotate short-term AI agent profits into long-term DeAI infrastructure and support founders who can combine speculation/distribution with real technology.
View Original
The content is for reference only, not a solicitation or offer. No investment, tax, or legal advice provided. See Disclaimer for more risks disclosure.
The Era of Virtuals Hegemony: When AI Agents Turn the Bull Run into a "Fair Launch" Liquidity Game
Original Title: AI Trends in the Trenches
Original author: 0xJeff
Source text:
Compiled by: Daisy, Mars Finance
It has been about 2.5 weeks since the AI agent market hit rock bottom (with a market value of around $4 billion), and the market has now fully entered a bull market led by @virtuals_io. That's right... only Virtuals is in the lead. This wave of market activity is similar to October to November of last year, when Virtuals launched its agent tokenization platform, establishing itself as a pioneer and the first legitimate player, providing a top-tier distribution network for any AI-related project willing to adopt a "fair launch" token model.
The difference this time lies in the latest feature "Genesis Launch" - a fairer way to launch and reward early supporters. This brings out the first trend:
Genesis Launch has completely changed the state of the primary market—from speculative investors blindly rushing into random/small circle junk projects on Pumpfun, to speculators almost certainly making 5-10 times returns when new projects launch.
The achievement of this goal is attributed to the introduction of the "points" mechanism, which coordinates the objectives of different stakeholders, as well as the fixed market value and fixed supply launch model (each project launches with 112,000 VIRTUAL, approximately 200,000 FDV). Each participant can earn up to 0.5 VIRTUAL based on the points contributed, top agent tokens (Virtuals have a dedicated list), or simply earn points by promoting Virtuals.
The recently introduced cooling-off period has further suppressed selling behavior and enhanced the appeal of the Genesis Launchpad, as sellers must now think twice.
As of now, the Genesis Launchpad has proven to be very successful, with @BasisOS being the most successful case, bringing participants a 200x return. Since then, many projects with returns between 5x and 40x have emerged on Virtuals.
In light of the success of Genesis Launch, funds and attention have once again focused on the Virtuals ecosystem, raising the valuation floor of almost all proxy tokens on the platform.
However, despite the resurgence of interest, the lack of quality projects remains one of the biggest challenges facing the Virtuals ecosystem, which leads to the second trend —
As a trader/speculator, you can make money by investing in projects with mediocre teams and average products—this is entirely due to the mechanism design of Genesis Launch. The probability of a project being pumped from a market cap of $200,000 and then dumped is quite high (especially when you know there are no insiders or pre-sale rounds).
As a project party, if you have a unique idea, you can launch a token first, even without a runnable product. There is no need to specify target users, validate market demand, or worry about revenue and user growth. You just need to create the maximum hype and then launch (having a demo version can add points, but it's okay if there isn't one, haha).
Any project that meets the basic requirements (having decent documentation, a good product idea, and a seemingly reliable team) can successfully launch a token at the Virtuals Genesis Launch.
The lesson for investors here is to view these startup projects as short-term speculative opportunities rather than medium to long-term value investments. Because, nine times out of ten, the so-called "underlying AI projects" being hyped are trash. (If interested, I analyzed a typical case in detail on Substack.)
While the proliferation of garbage projects has occurred, it has also brought a wealth of opportunities and gaps for high-quality projects (whether in the AI field or other areas). This leads to the third trend—
Two months ago, I communicated with the @logarithm_fi/@BasisOS team. I learned about their products (similar to Ethena's Delta neutral strategy, but not relying on stablecoins), and I recognized this team early on when they were deeply engaged in the LPDFi track (products based on Univ3 liquidity). Out of appreciation, I offered them some suggestions for token economic models and launch plans. At that time, I really didn't expect it to rise this much—although DeFi products have fundamental support, the so-called "AI" part is indeed too early. However, in the end, none of this matters, as the project's performance still crushes the market.
Basis's successful case proves that there are blank opportunities for DeFi projects to issue tokens in Virtuals. Although there is a lack of mechanisms in traditional DeFi to incentivize TVL through token emission, leveraging the built-in traffic and attention of Virtuals (especially when you have solid DeFi products) is enough to ensure that your treasury/strategy receives adequate TVL.
If you have a mature DeFi product and want to optimize user experience or enhance core functions through AI, feel free to communicate anytime. I would be happy to exchange ideas.
In addition to the traffic dividend, this also gives rise to new trend experiments similar to the 2023-24 ETH shitcoin era. This leads to the fourth trend —
Once upon a time, the market was flooded with a large number of small Ponzi DeFi projects that experimented with various token economic models: charging 1-3% transaction fees, using these funds to expand the treasury to maintain the Ponzi strategy, and then returning the profits/interests to token holders. In that era when Ethereum players had ample funds (the frenzy of people all-in on garbage projects is still fresh in memory), project creators could rake in six to seven figures in just one or two weeks—thanks to the scarcity of tokens in the market at that time.
Today, we see similar patterns in the field of AI agents.
Typically, the Virtuals platform charges a 1% fee on each transaction, of which 70% goes to the project creator. Other launch platforms also charge a transaction fee of 1-2% and return 70%-100% to the creators.
@Squidllora, powered by intelligent support from @AlloraNetwork (recently launched on @autodotfun), is allocating creator shares to expand its liquidity pool/treasury, and trading mainstream cryptocurrencies using Allora's predictive models (Allora is like Bittensor in pure financial scenarios—numerous scientists are competing to build the best predictive models for crypto assets across different time dimensions). A portion of the profits generated from trading will be used for token buybacks.
This model is particularly effective for teams with ample funds — they do not need to rely on transaction fees to maintain operations. Now, teams can issue AI agent tokens as a marketing tool to accumulate attention and transaction fees through new user acquisition funnels, thereby providing startup capital for brand new AI experiments.
That being said, looking at the entire AI agent field, despite many teams publicly advancing projects, there are still no other hotspots apart from the Virtuals ecosystem. This leads to the fifth trend—
Due to the ongoing promotion of the Genesis Launch, the valuation floor of proxy tokens on Virtuals continues to rise. However, the emergence of this wave of new projects does not imply improvements in fundamentals or technology; its main driving force comes from significant optimizations in trading structures—more people are participating in the games on Virtuals because they know they can profit from it. This trend may continue until the prices of existing projects reach local highs/ceilings.
Once this happens, market attention will naturally shift to other ecosystems, such as @CreatorBid, @arcdotfun, @autodotfun, especially those low market cap projects with obvious improvements in fundamentals (new features, product releases, new partnerships).
Currently, CB and Arc may be the two best ecosystems, among which there are still some undervalued projects that have not yet started a bull market (such as the 3-4 assets commonly found on CB that are integrated with subnets or focused on developing Bittensor-related products, as well as those handshake projects on Arc that contribute to the Ryzome ecosystem).
The best way to seize this opportunity is to position these tokens in advance before most people realize their value.
Despite the continuous price rise of mainstream AI agent ecosystems like VIRTUAL and AI16Z, many institutional investors can only watch from the sidelines—because the currently significantly appreciating assets are only suitable for retail/enthusiastic speculators to participate in. These assets have thin liquidity, and the market maker structure is fragile (especially in Virtuals).
The current lack of a sound liquidity infrastructure, coupled with the growing interest from institutions in decentralized AI, has prompted them to shift their funds towards investing in decentralized infrastructure, agent-based Layer 1, and decentralized AI laboratories, rather than the current generation of AI agent tokens.
You may ask: What tokens are favored by these institutions? They include GRASS, TAO (and its subnet), VANA, FLOCK, $PROMPT (which might count), and a series of yet-to-be-released tokens like @NousResearch, @PluralisHQ, @PrimeIntellect—these projects are building a true Web3 AI moat that puts ownership of high-performance models in the hands of the public (rather than centralized AI labs). In short, these are truly complex AI projects that ordinary investors neither understand nor know how to invest in.
How to make the best layout based on trends?
My strategy is to gradually shift the profits from short-term small AI agent trading (especially those Genesis Launch projects with average team levels) into the decentralized AI (DeAI) field. Real AI fundamental projects take time to develop, as they mostly focus on infrastructure rather than consumer-facing products. Just as we have seen ChatGPT, Grok, and Anthropic suddenly excel in everyday tasks, real-time research, programming, and more, the decentralized Web3 model will one day shine in certain Web2 and Web3 native tasks.
Does this mean you should heavily invest in DeAI infrastructure?
Not necessarily. The driving force of the crypto market is speculation and distribution—90% relies on distribution, and 10% relies on the model. For a project to succeed in the long term, the key lies in whether it can execute distribution well: providing products that people need, having a good UI/UX, formulating and executing an excellent launch plan, designing a robust token strategy, and having effective user acquisition and retention strategies, among others.
My core investment logic is still to bet on teams that understand both distribution and technology (models). This is similar to the Web2 venture capital investment approach in vertical SaaS (targeting specific niche use cases, although it may use off-the-shelf models plus proprietary data behind the scenes).
I believe this logic will remain effective in the short to medium term, especially in the Web3 space—because everything here is driven by hype and community, and simple and easy-to-understand concepts tend to sell better. The market/trading structure is also evolving in this direction: the fair launch model is gradually becoming the norm in the market, while more and more teams (in other ecosystems) are starting to closely integrate these two aspects (spending more time collaborating with infrastructure projects to leverage their technologies, while focusing on distribution and GTM strategies to generate buzz for real AI technologies).
Quick summary
The Genesis Launch model of Virtuals has dominated market attention and returns.
Speculation still far exceeds substance - most are short-term opportunities.
DeFi projects with solid fundamentals have found room for growth on Virtuals.
The creator revenue from trading volume is driving a new wave of experimental trends.
Virtuals are leading the way, but other ecosystems may soon attract funding attention.
Institutional investors choose to wait and prefer DeAI infrastructure over proxy tokens.
Layout strategy: Rotate short-term AI agent profits into long-term DeAI infrastructure and support founders who can combine speculation/distribution with real technology.