Beep Beep, Comrade: AI Instagates Cold War 2.0
Many experts and media talking heads are calling DeepSeek AI China’s “Sputnik moment.” The Chinese model is supposedly on par with OpenAI and Meta at a fraction of the cost. Silicon Valley executives are sweating, media outlets are breathlessly covering it, and the phrase Sputnik moment is getting thrown around like it’s 1957 all over again.
But here’s the thing; Sputnik was a PR stunt. DeepSeek AI might be, too. The difference is that China has a track record of following through.
The Sputnik Playbook
Let’s get something straight–Sputnik was a metal beach ball with a radio. The Soviets launched it to say We beat the Americans to space! and scare the shit out of everyone. It worked. The U.S. went full hold my beer mode, started throwing cash at NASA like a drunk uncle at a casino, then dunked on the Soviets so hard they reconsidered the whole space thing.
In a recent discussion with a friend who specializes in Russian history, he recalled that the Soviet Union’s entire strategy involved overhyping big moments while neglecting the long game. They built cheap housing blocks in the 1950s, and then when they began to crumble in the 1980s, the party leaders shrugged and said What? We already built you housing. Meanwhile, their people were getting a taste of Western life. Coca-Cola, Levi’s, and MTV showed them what they were missing, and it turned out they wanted that a lot more than what the Soviet Union was selling.
DeepSeek’s Hype Machine
Now we have DeepSeek, a Chinese AI startup claiming their latest models are not just competitive with the best American AI but also cheaper. Their DeepSeek-R1 is supposedly 20 to 50 times more cost-efficient than OpenAI’s top-tier model. They trained their flagship DeepSeek-V3 with just $6 million in computing power, which is an insanely low number compared to U.S. benchmarks.
Of course, these numbers are coming from China’s state-controlled ecosystem, where transparency is optional. Many people in the AI world are skeptical. Scale AI CEO Alexandr Wang even claimed DeepSeek has 50,000 Nvidia H100 chips, a number that, if true, would violate U.S. export controls. DeepSeek has not confirmed or denied this. Bernstein analysts have also pointed out that the actual cost of training DeepSeek’s models was likely much higher than reported.
Even if DeepSeek is exaggerating its numbers, this is still a problem for the United States.
When the Pot Calls the Kettle an IP Thief
China’s AI ambitions have put American companies on edge, and DeepSeek has been particularly devastating to Big Tech—especially OpenAI. For years, these companies have been crafting a narrative that AI is an insurmountably complex field requiring massive infrastructure, like $500 billion Stargate data centers. Now, DeepSeek has upended that story, proving that cutting-edge AI can be built outside the walled gardens of Silicon Valley.
Naturally, OpenAI and others have been quick to discredit DeepSeek, claiming intellectual property theft. PayPal Mafia member and White House AI advisor David Sachs recently went on the news, alleging that OpenAI has "substantial evidence" that DeepSeek stole OpenAI’s outputs to fine-tune its models—a practice known as distillation, which OpenAI explicitly forbids in its terms of service. This isn’t just irony; it’s machine-optimized, fully automated turbo-irony.
Let’s not forget how OpenAI itself vacuumed up the entire internet, including copyrighted material, without asking anyone for permission. Elon Musk accused them of scraping Twitter. George R.R. Martin and other authors are suing them, along with many other copyright infringement lawsuits worldwide. But here’s a tech bro founder secret: they do shady stuff first, then ask for forgiveness later. Once a company reaches critical mass, there’s no stopping it—just ask Uber or Airbnb. So far, OpenAI has mostly prevailed in its legal battles, setting a precedent that might make its complaints against DeepSeek seem more like strategic maneuvering than a genuine grievance.
The Key Difference: Follow-Through
Unlike the Soviets, China does not launch a flashy stunt and call it a day. They iterate. They scale. They integrate breakthroughs into their long-term strategy. For example, the CCP has built one of the most extensive surveillance systems in the world and it keeps getting better.
China has already signaled that DeepSeek is a top priority at the highest levels of government. The company’s founder attended a closed-door meeting with Premier Li Qiang right after launching their latest model. This is a sign that AI dominance is a serious policy goal.
Additionally, they have built an economic system that actually delivers consumer goods, infrastructure, and technology that their people want. The lack of these three elements was one of the factors that contributed to the collapse of the USSR.
Why the United States Should Be Worried
Even if DeepSeek’s numbers are inflated, even drastically reduced statistics should be concerning. China is playing the long game, and their strategy is working.
They are making AI cheaper and more accessible.
They have government backing to push for global adoption.
They are improving at a pace the Soviet Union never could.
The United States dominated the space race because the Soviets could not keep up. China, on the other hand, is not just competing. They are positioning themselves to win in the long run. This is not just another Sputnik moment. It is something much bigger. If American companies do not wake up, AI will be the first domino to fall.
Final Thoughts
In 1958 just a few months after Sputnik launched, the legendary sci-fi author Isaac Asimov published a story, The Feeling of Power, where people get so used to computers doing all the thinking that they literally forget how to do math. Then one fella figures it out again, expecting applause—only for the military to say, "Great, now we can save money by putting humans in missiles instead of using AI." His "breakthrough discovery" turns into a nightmare.
AI is a tool, not a crutch. If we let it do all the thinking, one day we might wake up and realize we have no power over it—or worse, that we have forgotten how to use our own fucking brains.