Deepseek不为人知的83小时: 25日-29日，83小时，Deepseek服务器集群，受到每秒超过2.3亿次DDos恶意请求，攻击总量相当于整个欧洲三天的网络流量总和。

Dario, the CEO of Anthropic, has now blamed all his failures on the lack of chip export controls on China after seeing DeepSeek’s rise. It's reasonable—DeepSeek is bad news for OpenAI, but an absolute nightmare for tier-2 LLM players like Anthropic. OpenAI is closed and expensive, but it’s still ahead. Anthropic, on the other hand, is just proving that a closed, worse, and overpriced model can exist—kind of like a $10 Coke at Disneyland. Even with Dario’s high IQ, 'insufficient export controls' might be the only excuse he can offer when angry investors ask where their money has gone. But do you really think chip export controls are going to save Anthropic? The #1 rule in building a successful tech company is simple: create something that meaningfully adds to human civilization. DeepSeek is winning because they open-sourced their advanced reasoning model, driving tons of downstream applications, and shared their methods to accelerate technological progress. Meanwhile, Anthropic is actively slowing down human technological development—forcing people to use more expensive, lower-quality models through political influence, locking up their tech, and trying to monopolize the market and control users. Those who advance human civilization will win. Those who hinder human civilization will die. It’s just that simple, it's just a matter of time.