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

Author: Nan Zhi, Odaily Planet Daily

In January, AI tokens experienced the last wave of frenzy since October, with VIRTUAL and ai16z's market caps reaching 5 billion and 2.5 billion at one point. Top AI tokens like arc and swarms, which emerged from pump activities, continuously launched on Binance contracts. However, the situation took a sharp turn after Trump released TRUMP, leading to a severe bloodbath in the AI sector, with a significant drop across the board. Although there was a brief rebound during TRUMP's decline to 30 dollars, DeepSeek soon delivered the final blow that burst the AI bubble. The market cap of arc fell from 600 million dollars to 250 million dollars, while swarms plummeted by 90% from 600 million dollars to 60 million dollars.

During the rise of AI, creating buzz, engaging in activities, or telling a new narrative, such as transforming into an AI Launchpad platform or developing frameworks, could significantly boost token performance. On the other hand, these AI projects are currently "meaningless"; at least we have yet to see any AI product truly and phenomenally transform the operations and processes of the Crypto industry. Therefore, the existing routines and stories can no longer reverse the situation.

In this environment of AI retreat, are there any AI projects launching self-rescue actions? How will they stabilize user confidence, control chips, and curb selling behavior?

Odaily Planet Daily will summarize the recent movements of leading AI projects in this article, summarizing the routes attempted by various top projects. Although we cannot currently determine whether their methods are effective, "what has happened will happen again." A successful solution will be emulated by other projects or even develop into the next core narrative, and we will continue to observe this.

Flexing Muscles, Building Relationships

arc is one of the relatively stable projects in AI, with its core business being the AI Agent framework Rig based on the Rust language, which does not entirely rely on the Solana network. Recently, arc has engaged with multiple ecosystems, including but not limited to:

• Becoming one of the projects supported by the Arbitrum Stylus Fund (with a total funding amount of 5 million ARB)

• Collaborating with the Solana Foundation to launch targeted funding activities for developers

• Participating in the ethdenver hackathon in collaboration with Monad

Product Updates

Chasing Trends, Integrating DeepSeek

DeepSeek is undoubtedly the hottest topic recently, with multiple projects announcing their integration with DeepSeek during its rise, including GRIFFAIN, AVA, BUZZ, etc., but it has not significantly impacted their own business development.

High-Frequency Updates, High-Frequency Exposure

Previously on the X platform, we mostly saw "chatty" AI Agents outputting various topics and styles of content, but the development accounts behind them did not update frequently. Perhaps sensing the arrival of the AI winter, some projects have further increased the frequency of product updates on their main accounts, including LMT, ZEREBRO, etc.

Especially LMT, which has been publishing product update details daily on the X platform since January 8, highlighting the content that LMT builds, fixes, or improves each day, with very detailed granularity.

For such projects, product development is inherently the main line, and due to their development background, capabilities, and products, they have become one of the few projects to break through a market cap of over 100 million dollars during the AI boom. After the revelry and bubble burst, the landing of the next generation of products will become the foundation for future growth (assuming there is a second wave of AI boom), so having products that can iterate and continuously share may be the next way to break the deadlock.

swamrs: Doing Something Different

Previously, swarms was one of the first projects to develop frameworks after ai16z and gained significant exposure due to multiple criticisms from ai16z founder Shaw, leading to a short-term market cap surge to 600 million dollars.

On February 1, swarms made another groundbreaking move—"requesting token holders to invest/stake tokens in the DAO." Swarms announced on the X platform that the team only holds 2% of the total tokens, which makes the tokens susceptible to manipulation by "malicious token holders and exchanges that are not genuinely interested in our mission." Therefore, swamrs proposed to increase the team's token holding to 10%, with the specific form being the establishment of a DAO, where token holders would invest tokens into the DAO and then receive staking rewards, governance rights, and priority access to products.

However, this proposal has many unclear aspects, such as whether token holders still own their tokens after investing; if they retain ownership, how can it be called a "team share"? If they do not retain ownership, why is the incentive called "staking rewards"? Moreover, the team does not hold many tokens, so how they will provide staking returns is also unclear.

Conclusion: The Way to Break the Deadlock May Lie in the Product Itself

In summary, we can see that leading projects are no longer keen on "creating big news," but are turning to further demonstrate their ability to develop products and focus on long-term product development. The main theme of 2025 will still be AI; after the bubble clears, growth may no longer be based on "narratives," but rather on "products."