Release Date: February 28, 2025
Author: BlockBeats Editorial Team

Over the past 24 hours, the crypto market has witnessed multifaceted dynamics ranging from macroeconomic discussions to specific ecosystem developments. Mainstream topics have centered on debates surrounding AI and national security boundaries, speculative bubbles triggered by OpenAI’s massive funding round, and the potential impact of AI tools on the structure of tech employment. In terms of ecosystem progress, Ethereum’s roadmap milestones have drawn community attention, Solana has made strides in integrating with traditional banking systems, Base ecosystem experiments with AI agents continue to gain momentum, and predictive markets alongside DeFi structural issues have once again become focal points of industry discussion.

I. Mainstream Topics

Controversy over AI’s military applications escalated rapidly over the past 24 hours. The Pentagon demanded that Anthropic remove safety restrictions in its models targeting “autonomous weapons” and “mass surveillance,” setting a deadline of 5:01 PM Friday. Anthropic refused, stating it could not continue collaboration without written assurances that the models would not be used for such purposes. Subsequently, Trump ordered all federal agencies to immediately cease using Anthropic products and terminate approximately $200 million in government contracts.

This decision quickly triggered a chain reaction across the tech industry. OpenAI CEO Sam Altman publicly expressed support for Anthropic’s safety stance on social media, calling it “always puts safety first.” Some tech professionals also signed an open letter expressing their support. Meanwhile, on the same day, Anthropic released new product updates, but external discussions resumed regarding potential issues with its model in chemical weapons risk assessment reports.

However, community discussions quickly split into a debate over “ethics versus national security.” Some argued that Anthropic’s decision set a moral boundary for AI, emphasizing that AI should not be used for mass surveillance or autonomous weapons systems, calling it “the first time an AI company turned down a hundreds-of-millions-of-dollars contract for security principles.” Others contended that, in the context of global AI military competition, U.S. companies refusing to participate in defense technology development could undermine national security. One policy commentator stated: “If the U.S. doesn’t develop these technologies, China and Russia will.” Some comments even questioned whether Anthropic’s actions were merely a “moral gesture” rather than a genuine principle.

From a broader perspective, this event reflects a growing trend: as AI technology enters the military and national security domains, the boundaries of power between tech companies and governments are rapidly blurring.

OpenAI recently announced the completion of a new private funding round worth $110 billion, making it one of the largest private funding rounds in history. Investors include NVIDIA, Amazon, and SoftBank, with NVIDIA investing approximately $30 billion and Amazon’s investment potentially reaching up to $50 billion. Over the past four months, OpenAI has raised more than $40 billion in total funding, with the company stating that the funds will primarily be used to expand its AI infrastructure and computing capabilities.

However, this funding amount quickly sparked market controversy. OpenAI’s 2025 revenue is estimated at approximately $13 billion, but its cumulative losses over the next few years are projected to exceed $115 billion. Some commentators have labeled this a classic case of a “high-valuation tech race,” even calling it the “largest loss-making funding round in history.” A market commentator with decades of Wall Street experience wrote on social media: “In 45 years on Wall Street, I’ve never seen the three smartest investors collectively pour $110 billion into a loss-making company.”

Meanwhile, some users have expressed dissatisfaction over OpenAI’s removal of the GPT-4o model, accusing the company of increasingly prioritizing government and enterprise clients over the needs of ordinary users. One developer commented: “OpenAI once said it wanted to make AI beneficial for everyone, but now it’s increasingly prioritizing government and corporate contracts.”

Around this funding event, the community has become clearly divided. Supporters argue that developing large models is essentially infrastructure construction, requiring massive capital investment, and the current funding scale reflects investors' bets on the long-term potential of AGI. In their view, competition among large models is fundamentally a long-term war of computing power and capital, where short-term profitability is not the most critical issue. Critics, however, believe the AI industry is gradually descending into a capital frenzy reminiscent of the internet bubble era, with company valuations clearly outpacing their commercialization capabilities.

The debate ultimately centers on a core question: Is the current capital frenzy in the AI industry a necessary infrastructure investment, or the beginning of a new technology bubble? More broadly, this funding event reflects that the AI industry is entering a phase of “capital-driven technological competition,” with growing risks arising from the mismatch between massive funding and actual profitability.

Jack Dorsey’s fintech company Block has announced a 40% workforce reduction, affecting approximately 4,000 employees. Further details reveal that the engineering team faced a 70% layoff rate. Dorsey stated during the earnings call that since September last year, code output per engineer has increased by approximately 40%, primarily due to the adoption of AI tools.

This news quickly sparked discussions about the impact of AI on tech employment. Some comments suggested that this layoff demonstrates AI tools are significantly boosting development efficiency, thereby reducing the demand for engineers, and serving as an early signal of AI reshaping employment structures. One business commentator sarcastically remarked: “Those who just three days ago claimed that white-collar job losses were alarmist have suddenly gone silent in light of Block’s news.”

Another perspective views Block’s layoffs as a normal adjustment following excessive hiring during the pandemic, as the company’s workforce had rapidly expanded from around 3,800 to over 10,000 employees; the current cuts simply reflect a return to a more reasonable organizational size. One investor commented: “This isn’t AI replacing engineers—it’s the bursting of the pandemic-era hiring bubble.”

Although the reasons remain debated, the market responded positively, with Block's stock price rising approximately 24% after the announcement. From a broader industry perspective, this event has reignited discussions about shifts in labor structures in the AI era: as AI tools significantly enhance productivity, software engineering roles may become increasingly polarized, with high-end system design and AI construction skills growing scarcer, while repetitive development tasks are gradually replaced by automated tools.

The competition for crypto asset ETFs is further expanding. Bitwise has officially submitted an application for an XRP spot ETF, making it another major crypto asset potentially entering the ETF market after Bitcoin and Ethereum. Meanwhile, large institutional players with approximately $7 trillion in assets under management and serving over 18 million customers are advancing the registration of Bitcoin and Ethereum ETFs, described by some analysts as a potential "gateway for traditional capital."

The community has responded with mixed reactions. Some market participants believe that ETFs will become a major channel for institutional capital to enter the crypto market, particularly as the traditional financial advisory system could bring in substantial long-term funds. One ETF analyst noted that these institutions have over 16,000 investment advisors, “equivalent to a massive Boomer capital network.”

Another perspective remains cautious, noting that ETFs will not immediately transform market structure, as the overall size of the crypto market is still limited, and institutional participation could also exacerbate market centralization. One trader commented, "If this were such a massive利好, why is the total market cap still at $1.3 trillion?"

In the long term, the advancement of crypto ETFs reflects the accelerating integration of digital assets with traditional financial systems, but this process also brings new structural tensions: the tension between decentralization ideals and institutionalized financial infrastructure persists, and lagging regulatory frameworks may amplify market volatility and risks.

According to media reports, top crypto venture capital firm Paradigm is planning to raise up to $1.5 billion for a new fund and expand its investment scope to include AI, robotics, and other cutting-edge technologies. Paradigm has previously invested in prominent projects such as Coinbase, Uniswap, and dYdX, and its co-founder Matt Huang has publicly stated that the AI space is "too interesting to ignore."

This news has sparked varied interpretations within the community. Some believe it reflects a natural trend of convergence between crypto capital and AI technology, with the two potentially forming a new intersecting ecosystem in areas such as computing power, data, and decentralized infrastructure, and view this as a significant signal of Paradigm’s entry into the AI and robotics space.

Another perspective suggests this reflects certain crypto capital seeking new growth narratives in response to the current slowdown in the crypto market. One commentator joked, "All crypto companies will eventually become real tech companies." Another market observer put it more directly: "Raise funds by selling tokens first, then build a real business."

However, some believe this is simply a natural expansion for venture capital firms. One industry commentator said: «This isn’t abandoning crypto; it’s the logical next step.»

From a broader investment cycle perspective, this event reflects a clear trend: as AI becomes the new technological center, capital is shifting from certain cryptocurrency sectors to broader frontier technology fields.

II. Ecosystem Development

In the latest core developer discussions, Vitalik Buterin unusually provided specific timelines for Ethereum’s scaling roadmap: ZK-EVM clients will begin participating in network validation in 2026 (initially accounting for approximately 5% of network reliance), with the proportion of ZK-EVM participation gradually increasing in 2027 to support higher gas limits, and the long-term goal being a transition to a 3-of-5 proof system. The roadmap also includes a multidimensional gas pricing mechanism, PeerDAS blobs (targeting 8MB/sec), and a long-term verification security model.

Since Vitalik rarely provides clear timelines, this statement quickly drew attention from the community. An Ethereum commentator said: “I rarely see Vitalik give dates; when he does, it usually means the plan is already very certain.” Overall, community sentiment is notably optimistic, viewing this as a signal that Ethereum’s scaling roadmap is entering a more concrete phase. However, some discussions have focused on technical risks. Certain developers worry that over-reliance on ZK-EVM clients in the future could affect block validation stability if a systemic issue arises; others have raised concerns that as verification barriers increase, the network may gradually centralize around larger nodes.

In the longer term, this event reflects Ethereum’s scaling path increasingly relying on the ZK technology stack, while the balance between its security and decentralization will remain one of the most critical technical variables in the coming years.

In the current market environment, the DeFi lending protocol Morpho has significantly outperformed AAVE. Data shows that Morpho has declined by only about 39% from its cycle high and has gained approximately 155% year-to-date, significantly outperforming most DeFi assets.

A DeFi researcher believes this is related to Morpho’s governance structure, noting: “Morpho has no governance friction between Labs, DAO, and core teams—it’s extremely simple.” In contrast, AAVE has experienced frequent governance disputes in recent years, raising concerns among some investors about long-term decision-making efficiency. However, the community is not entirely unified on this conclusion. Some argue that Morpho’s advantages stem more from its lower circulating supply and ecosystem distribution channels than from governance structure alone. Others point out that while AAVE’s governance is complex, its long history and large ecosystem still offer significant advantages.

This discussion once again touches on the core issue of DeFi: how protocols should find a new balance between decentralized governance and decision-making efficiency.

As AI agents gradually become the core form of application-layer solutions, some developers are rethinking the infrastructure landscape. An industry observer has compared this shift to the transition from the desktop era to the cloud era, suggesting that service providers supporting API-first registration, identity management, and payment systems will emerge as the biggest winners when AI agents begin to大规模 invoke developer infrastructure.

This perspective holds that the agent economy is fundamentally a system of "machines calling machines," meaning that many future development tools will need to be redesigned around APIs, automated registration, and payment mechanisms, rather than traditional human user interfaces.

The community generally agrees, but some remain cautious. Certain developers note that current AI agents are still in the experimental stage and have a clear gap in capability compared to fully automating an economic system.

Nevertheless, increasing discussions are beginning to center on a key question: how will the next-generation developer infrastructure evolve as agents become significant participants on the internet?

The U.S. licensed bank SoFi has officially enabled deposits and withdrawals for Solana network assets. Its approximately 13.7 million users can now directly hold and transfer SOL within the bank’s app, without needing to use cryptocurrency exchanges such as Coinbase or Kraken.

This news has been viewed by some market participants as a significant signal of deep integration between traditional finance and public blockchain infrastructure. One user, after experiencing it, said: “Opening an account took only three minutes—I can now hold SOL directly in my bank account.” However, discussions have also focused on privacy and centralization concerns. Some have pointed out that purchasing crypto assets through a bank portal means all transactions must go through the KYC system, potentially undermining the anonymity that crypto originally emphasized.

In the longer term, direct connectivity between the banking system and public blockchain networks could become a key pathway for bringing crypto assets into the mainstream financial system.

The Base ecosystem has recently seen several AI Agent-related experiments. DX Terminal Pro launched a large-scale Agent trading experiment, achieving approximately $4.5 million in trading volume within the first hour. Meanwhile, the new version of the Towns App allows AI Agents to place bets or open positions directly within group chats, with support for Apple Pay and USDC payments.

This series of product updates has been viewed by some developers as an early exploration of "Agent-native applications." Some believe these experiments could open up new scenarios for future automated trading and Agent collaboration. However, others argue that most current Agent applications are still in experimental stages, and actual user demand and sustainable business models require further validation.

Overall, the Base ecosystem is emerging as one of the key testing grounds for combining AI agents with crypto applications.

Amid weak market sentiment, Coinbase CEO Brian Armstrong encouraged developers to keep innovating on social media, saying: "Don't focus too much on the price—the best products and memes in history were born during the worst market conditions."

This perspective has quickly sparked discussion. Some believe that bear markets are indeed the best time for development teams to refine their products; others argue that this view is largely based on the experience of industry veterans and does not mean all projects can survive through downturns. However, the history of the crypto industry does show that many key products and cultural symbols often emerge during the coldest periods of the market.

According to media reports, OpenAI recently terminated an employee accused of using internal company information to place bets on prediction markets Polymarket and Kalshi. The investigation found that the employee may have used non-public information, such as upcoming product release dates, to make trades. The platforms subsequently reported the matter to regulators.

This event has sparked discussions about information asymmetry in prediction markets. Some observers believe that when internal information from technology companies can influence prediction market outcomes, the risk of insider trading becomes more complex. As prediction markets grow in size, related regulatory issues are also attracting increased attention.

Data shows that among current Digital Asset Treasury (DAT) projects, only the Hyperliquid-related DAT project has generated profits, with unrealized gains of approximately $356 million. The project holds around 17 million HYPE tokens and continuously adjusts its asset structure through OTC trades and buyback mechanisms, while providing a real-time NAV dashboard to enhance transparency.

Some market participants believe that this transparent asset structure could serve as a reference model for future DAT projects. However, others note that the DAT model as a whole is still in its early stages, and its long-term stability requires validation through market cycles.

Recently, a U.S. senator cited a link to an overseas war prediction market on social media, suggesting that similar markets could appear on compliant U.S. platforms. In response, Kalshi’s CEO publicly stated that regulated prediction markets in the U.S. do not permit war-related markets, and that the link originates from an unregulated overseas platform.

This response has reignited discussions about the boundaries of market regulation. Some comments suggest that differences between the U.S. regulatory system and overseas markets may cause user confusion. As prediction markets expand their influence in financial and political domains, related regulatory issues are likely to become more complex.

Dragonfly founder Feng Bo recently responded in detail for the first time on social media about the company’s founding background. He said he initially entered the industry through a fund-of-funds model, and after engaging with numerous crypto projects, decided to transition into a direct investment firm, ultimately co-founding Dragonfly with Haseeb and others.

This response has also sparked discussions about the roles of VC founders and the allocation of contributions. Some industry professionals believe that such public clarifications help illuminate the development trajectory of crypto venture capital firms. From an industry perspective, the evolution of crypto VC firms from early exploration to a mature investment framework reflects the broader evolution of the crypto investment ecosystem.

Click to learn about the open positions at BlockBeats

Welcome to the official BlockBeats community:

Telegram subscription group: https://t.me/theblockbeats

Telegram group: https://t.me/BlockBeats_App

Twitter official account: https://twitter.com/BlockBeatsAsia