In the past 24 hours, the happenings in the crypto and tech circles have been truly worth observing. From debates over AI safety boundaries, to crazy fundraising scales, to AI impacting engineering jobs, the entire industry seems to be undergoing some kind of accelerated turning point.

First, let's talk about Anthropic refusing the Pentagon. This AI company stood firm, refusing the military's request to remove safety restrictions, resulting in Trump directly ordering the ban of a $200 million government contract. The event itself is quite interesting—it reflects an increasingly blurred boundary: the tug-of-war between tech companies and government power. Some say this is AI companies sticking to their bottom line, while others see it as a self-sabotage in the global AI competition. Regardless, this incident has prompted more industry reflection on the militarization of AI.

Next is OpenAI’s $11 billion funding round. Honestly, this number is quite shocking. NVIDIA invested $3 billion, Amazon might invest $5 billion, but the problem is OpenAI’s projected revenue for 2025 is only $13 billion, with cumulative losses possibly exceeding $11.5 billion. An old Wall Street hand tweeted that in his 45-year career, he’s never seen three of the smartest investors pour $11 billion into a loss-making company. Is this necessary infrastructure investment, or the start of a new tech bubble? The community is divided, and no one can persuade the other.

Then there's the situation with Block. Jack Dorsey’s company laid off 40%, with engineering team layoffs reaching 70%. They say code productivity has increased by 40%, mainly thanks to AI tools. This directly touches a sensitive topic: is AI truly changing the structure of engineering jobs? Some believe this is the beginning of AI reshaping employment, while others see it as a normal adjustment after pandemic-era overhiring. The stock market responded with Block’s stock rising 24%, clearly indicating market approval of this efficiency boost.

From a macro perspective, these events all point to the same phenomenon: we are entering a new phase driven by capital and computing power. Paradigm is raising $150 million to venture into AI and robotics, which isn’t an isolated event—crypto capital as a whole is seeking new growth narratives. Some joke that all crypto companies will eventually become tech companies, but fundamentally, this reflects capital chasing the next wave of kurva s proyek.

On Ethereum, Vitalik has rarely provided a concrete roadmap—ZK-EVM will start participating in validation in 2026, with gradual increases in 2027. Such clear time commitments are rare for him, and the community’s response is generally optimistic. But some worry that over-reliance on ZK-EVM could introduce systemic risks. On Solana’s side, SoFi officially supports SOL deposits and withdrawals; now 13.7 million users can hold SOL directly in their bank accounts, marking an important fusion of traditional finance and blockchain infrastructure.

The AI Agent experiments in the Base ecosystem are also heating up. DX Terminal Pro’s trading Agent reached a trading volume of $4.5 million in the first hour, and Towns App allows Agents to trade directly in group chats. These are still early experiments, but you can feel that application layers are trying out new things.

From a risk market perspective, Hyperliquid is the only profitable DAT project, holding 17 million HYPE with an unrealized gain of $356 million. This shows that, in the current market environment, some projects have indeed found sustainable models.

Overall, this period looks like a stage where multiple kurva s proyek are advancing simultaneously—AI infrastructure investments are accelerating, crypto capital is seeking new directions, and traditional finance is gradually integrating blockchain. Risks and opportunities are both very apparent.