After the AI tide recedes, how will the leading projects choose to save themselves and break through?

Original | Odaily Planet Daily ( @OdailyChina )

Author｜Nan Zhi ( @Assassin_Malvo )

In January, AI tokens staged the last wave of craze since October. The market value of VIRTUAL and ai16z once reached 5 billion and 2.5 billion, and the top AI tokens from pump, such as arc and swarms, were listed on Binance contracts. However, after Trump released TRUMP, the situation took a sharp turn for the worse. The AI sector was severely drained and plummeted across the board. Although it rebounded briefly when TRUMP fell to $30, DeepSeek immediately became the last blow to burst the AI bubble. The market value of arc fell from $600 million to $250 million, and swamrs fell 90% from $600 million to $60 million.

In the rising period of AI, doing something, making a living or telling a new narrative, such as transforming the AI Launchpad platform, developing a framework, etc., can greatly boost the performance of tokens. But on the other hand, these AI projects are meaningless at present. At least we have not seen any AI products truly and phenomenally change the way the Crypto industry operates and processes, so the original routines and stories can no longer reverse the situation at present.

In this AI recession, are there any AI projects taking self-help actions? How will they stabilize user confidence and control the chips and selling behavior?

In this article, Odaily Planet Daily will summarize the recent trends of the top AI projects and the routes that each of the top projects has tried . Although it is not certain whether their methods are effective at present, what has happened will happen again, and a successful solution will be imitated by other projects and even develop into the next narrative core. We will continue to observe this.

Showing off muscles and building relationships

Arc is one of the most stable projects in AI. Its core business is the Rust-based AI Agent framework Rig, which is not completely dependent on the Solana network. Recently, Arc has integrated multiple ecosystems, including but not limited to:

• Become one of the Arbitrum Stylus Fund support projects (total funding is 5 million ARB)

• Launching targeted grants for developers in partnership with the Solana Foundation

• Joining Monad at the ethdenver hackathon

Product Updates

Follow the trend and integrate DeepSeek

DeepSeek is undoubtedly the hottest topic recently. When DeepSeek became popular, many projects announced their access and integration with DeepSeek, including GRIFFAIN, AVA, BUZZ, etc., but it had little actual impact on the development of their own business.

High frequency update, high frequency exposure

Previously, on the X platform, we could see more chatty AI Agents outputting content of various topics and styles, but the development accounts behind them were not updated frequently. Perhaps sensing the coming of the AI winter, some projects have further increased the frequency of product updates on their main accounts, including LMT, ZEREBRO, etc.

In particular, LMT has published product update details on the X platform every day since January 8, highlighting what LMT has built, fixed or improved every day, with very detailed granularity.

For such projects, product development is the main line, and due to their development background, capabilities and products, they have become one of the few projects with a market value of over 100 million US dollars during the AI boom. After the carnival and bubble subside, the implementation of the next generation of products will become the basis for the rise (assuming there is a second wave of AI boom), so products that can be iterated and continuously shared may be the next way to break through.

swamrs: do something different

Previously, swarms was one of the first projects to develop a framework after ai16z, and it received a lot of exposure due to repeated criticism from ai16z founder Shaw. It was second to none in terms of storytelling, and its market value surged to US$600 million in a short period of time.

On February 1, swarms made another groundbreaking move - requesting token holders to invest/stake tokens in the DAO. Swarms posted on the X platform that the team only held 2% of the total tokens, which made the tokens vulnerable to manipulation by malicious token holders and exchanges that have no real interest in our mission. Therefore, swamrs launched a proposal to increase the teams token holdings to 10%. The specific form is to establish a DAO, where token holders invest tokens in the DAO and then obtain staking rewards, governance rights, product priority experience rights and other rights.

However, there are many unclear points in the proposal, such as whether the token holder still owns the token after investing. If the ownership is retained, how can it be called team share? If not, why is the incentive called staking reward? In addition, the team does not have many tokens, so it is not clear how to provide staking income.

Summary: The solution may lie in the product itself

In summary, we can see that no one in the top projects is keen on making big news anymore, but instead turns to further demonstrating their ability to make good products and develop products for the long term . The main theme in 2025 will still be AI. After the bubble is cleared, the rise may no longer be based on narratives but on products.