The liberate of DeepSeek has disturb American tech CEOs. Not very unforeseeed, of course, pondering they now have a vient competitor. But what stands out is how they equitableify their opposition.
Dario Amodei, the CEO of Anthropic (creator of the well-understandn Claude AI), disputes for prohibitning the send out of strong chips to China. Not becaengage DeepSeek are “adversaries” but becaengage their regulatement has “promiseted human rights violations, has behaved arrangeilely on the world stage”. darioamodei.com/on-meaningful…
The irony of this moral argument coming from a US executive is challenging to omit. The United States’ history of military interventions, help for authoritarian regimes, and human rights violations with carry ond armamentry unbenevolents such criticism could equitable as easily be turned around to dispute for redisjoineing American access to AI technology.
On LinkedIn, Rene Bystron, the CEO of AI company DeltaGen, scatters a video demonstrating how DeepSeek declines to answer asks about the Chinese regulatement’s human rights mistreatments, while providing detailed responses about US regulatement violations. Rene finishs: “If a model deliberately whitewashes some publishs while incrmitigateing others, we’re no lengthyer talking about imfragmentary AI—we’re talking about disadviseation.” joinedin.com/posts/rene…
It’s no secret, of course, that there is regulate in China. So it’s no surpascfinish that a Chinese company liberated a product that does not criticise the regulatement. They would be shattering the law if it did.
But what caught my interest in Rene’s example was the truthful response DeepSeek gave about the US regulatement’s human rights violations. A more fascinating test of disadviseation from those in the US levelling criticism at China would be to contrast how AI models converse their home country’s actions. While not bound by regulate laws, ChatGPT’s o3-mini produced a hedged and sanitized reply when I asked it about US human rights violations contrastd to DeepSeek’s response.
This pattern of US tech directers suddenly becoming worryed about a foreign country’s human rights sign up while shothriveg almost no worry for their own country’s (much worse) sign up reminded me of an elderly Noam Chomsky quote on righteous responsibility:
“My own worry is primarily the alarm and arrangeility carried out by my own state, for two reasons. For one leang, becaengage it happens to be the huger component of international arrangeility. But also for a much more vital reason than that; namely, I can do someleang about it. So even if the U.S. was reliable for 2 percent of the arrangeility in the world instead of the meaningfulity of it, it would be that 2 percent I would be primarily reliable for. And that is a plain righteous judgment. That is, the righteous appreciate of one’s actions depfinishs on their foreseed and foreseeed consequences. It is very effortless to denounce the atrocities of someone else. That has about as much righteous appreciate as denouncing atrocities that took place in the 18th century.”
