In the past two months, discussions in the crypto market have been extremely intense, with new topics seemingly fermenting almost every day. From large-scale layoffs at Block to Anthropic’s positive response to the U.S. Department of Defense, and to Ethereum’s technical roadmap, these events reflect the industry’s anxiety and repositioning in the AI era.

First, let’s talk about the layoffs at Block. Jack Dorsey announced directly on X that over 4,000 employees were cut, accounting for 40% of the total staff, citing the goal of creating a smaller, more efficient organization through AI tools. As soon as this news broke, the community exploded. Some questioned whether AI was truly mature enough to replace 40% of jobs in just two months. Ansem and other KOLs pointed out that this seems more like using AI as an excuse to mask deeper capital pressures and organizational adjustments. Think back to when Elon laid off 80% of Twitter’s staff—AI wasn’t nearly as popular then. Now, Block uses the same logic, but the AI reason sounds more plausible. Historically, no technology has caused such large-scale layoffs just two months after its emergence. This actually reflects a more fundamental issue: the speed of AI productivity gains has far outpaced companies’ organizational decision-making and adjustment capabilities.

Another topic that sparked widespread discussion was the Ethereum Foundation’s Protocol Strawmap. Justin Drake announced a “draft” roadmap planning a series of hard forks through 2029, targeting high-speed L1, gigabit-level L1, and trillion-level L2. Will this roadmap truly be implemented? dcfgod believes ETH 3.0’s architecture is already quite clear, but Ryan S. Adams pointed out that while AI has enhanced research capabilities, the real difference lies in the team’s vision and execution ability. The underlying issue is whether Ethereum Foundation’s open discussion-oriented decision-making process can keep up with market demands for rapid iteration and efficient execution.

On the topic of AI and crypto integration, community interest is growing rapidly. KOLs like Ansem crypto have recently discussed a more macro issue: astronomical investments in AI infrastructure are draining global liquidity and causing large-scale layoffs. Does this mean the AI capex cycle is causing structural changes in the market? Some believe this is an inevitable pain point of a new growth cycle, while others worry that the current global financial system’s liquidity supply may not keep up with the money-burning development path of AI.

At the ecosystem level, Ethereum is accelerating L2 interoperability. L2BEAT launched the Interop tracker, which can track cross-chain stack progress of various L2s in real time. This tool quickly gained widespread sharing within the ETH community and was seen as an important milestone for L2 standardization. But whether the fragmentation of these tech stacks can truly accelerate integration remains to be seen.

Solana is also iterating rapidly. SIONG, a core team member of Jupiter, tweeted that a certain protocol pushed a destructive update without notice, causing the team to work overnight. Although top projects like Backpack and Armani didn’t comment much, they collectively retweeted the post, reflecting coordination issues within the Solana ecosystem amid high-frequency iterations. In the long run, this may push Solana to establish more mature cross-protocol notification mechanisms, but in the short term, community trust costs have already risen.

Reserve has increased the bug bounty for security vulnerabilities to $250k, seen as a positive sign of DeFi projects investing more in security. DFlow and Phantom collaborated to launch a Solana-specific Claude AI skill pack, marking the official entry of AI tools into Solana’s development process. These moves are lowering development barriers and accelerating application innovation.

In the Perp DEX space, Lighter launched the LIT Fee Credits mechanism, allowing small market makers and high-frequency traders to exchange tokens for fee discounts and latency advantages. This helps diversify liquidity, but the long-term sustainability of the economic model still needs time to be validated. Hyperliquid released the HIP-6 governance proposal, a step toward a more mature DAO governance structure.

Prediction markets are also evolving rapidly. Kalshi started publishing real-time news about key figures like Elon Musk in the form of “JUST IN,” with viewership reaching 140k. Polymarket is tracking major events such as the FAA closing Texas airspace and Anthropic refusing the U.S. Department of Defense’s requests, with odds fluctuating in real time. These platforms are evolving from pure gambling tools into parallel news verification layers and geopolitical information discovery engines. Of course, this also introduces new challenges regarding information authenticity and regulatory compliance.

Returning to the event of Anthropic refusing the U.S. Department of Defense’s request, many see this as a political red line between AI companies and national security interests. Odds on related options on Polymarket fluctuate around 44%, reflecting market real disagreement over this decision. Some see it as a victory for AI ethics, others see it as purely business considerations. Regardless, this event marks that the politicization of AI by 2026 has entered a stage of real risk contest.

Overall, the market movements over these two months are essentially answering a core question: in the context of AI reshaping productivity, global liquidity pressures, and escalating geopolitical tensions, how will the crypto ecosystem find its place? Observers like Ansem crypto have been emphasizing that we are transitioning from a phase of technological narrative to one of real risks. Whether it’s layoffs, roadmaps, or security incidents, they are no longer just technical or commercial issues but early warning signals of systemic risks. Continuous attention is warranted.