What is Deepseek AI?

Shooting up to the #1 ranking in Apple’s app store immediately upon its release, Deepseek AI has garnered a mixed-bag of praise, suspicion, and fear in an extremely short period of time. Everyone and their mother is talking about Deepseek, and if you’re not in the world of artificial intelligence, perhaps you don’t comprehend how big of a moment this is.

Some reporters worried Deepseek would crash the stock market, many tech influencers speculated this was our generation’s “ Sputnik moment” on social media, and whether you’re choosing to stick with ChatGPT and stay loyal to OpenAI, or you opt for this cheaper, better Chinese AI model, or even if you don’t know or care much about AI, you will choose which side of history you'll be on by voting for one of these products with your dollar.

Table of Contents

• What is Deepseek?

• Why are People Afraid of Deepseek?

• What Is the Future of AI After Deepseek?

• Conclusion

• FAQ

What is Deepseek?

Deepseek is a Chinese company formed in 2023 by Liang Wenfeng. Previous to Deepseek, Liang was the CEO of a hedge fund called High-Flyer which leveraged AI for quantitive trading.

The Chinese startup’s new model, Deepseek R1, is a free, open source Chinese artificial intelligence that is just as easy to use as ChatGPT, delivering data of similar quality, developed at a fraction of the cost.

Like ChatGPT’s o-1 model, R1 is a reasoning model that recreates human thought by showing users the process with which it generates its outputs and information. This is an upgrade from o-1 which only shows users a summary of its reasoning as opposed to the whole thought process.

Another large language model with similar capabilities to o-1 wouldn’t have caused this much a disruption if not for the fact that R1 cost $6 million to train as opposed to ChatGPT’s $100 million +.

How was Deepseek able to train their AI model so cheaply? Well, because Nvidia A100 chips are banned from export to China, Deepseek’s developer paired what Nvidia chips he had with much cheaper chips to innovate a more cost-efficient process.

Why Are People Afraid of Deepseek?

When traders saw how cheap it was to create this R1 compared to many of the stocks they had already invested in, they immediately sold their stock in American startups and corporations aiming to develop AI in the ChatGPT cost range.

AI development previously required so much semiconductor energy to produce Nvidia chips that entering the AI industry was considered too risky for just any startup.

At $6 million though, Deepseek was developed so cheaply, that its release is a watershed moment that will change the industry’s landscape and allow for more companies to launch cheaper tools at competitive prices.

Many US stocks tanked as a result of China’s Deepseek phenomenon, including Nvidia, Nasdaq Futures, Meta, Google, Microsoft, OpenAI and more. Apple, who hasn’t released an AI model, wasn’t as effected by the Deepseek release.

Some experts speculate that Deepseek will drive investment toward Chinese companies over U.S. companies, however, there’s greater relevancy to the Sputnik comparison than most realize.

Sputnik wasn’t just the moment that legitimized Soviet innovation as the first country to break through the atmosphere, it was the moment that ignited the space race that the United States eventually won.

Similarly, China may have developed an LLM that shut down Sam Altman and Silicon Valley for now, but the AI race they initiated is the very climate within which American companies prosper. Historically, Americans work well under pressure.

What is the Future of AI After Deepseek?

There’s never been a more exciting time to be in the AI sector. If AI startups can work well under pressure, they can develop the tools that do more than just reshape how people use technology, but move the needle of global power.

Whether or not China will be victorious in this AI race has yet to be determined. Many have reservations about using their Deepseek model because they see the Chinese government’s propensity for data collection, propaganda, and censorship as a threat to national security.

Deepseek claims this troubling trends have no place in its AI model's training, however, if you ask Deepseek what happened in the Tiananmen Square massacre, it won’t give you a comprehensive answer.

Conclusion

Whether you want to call Deepseek’s release a watershed moment, Sputnik moment, or Pandora’s Box, those in the know cannot deny that the AI sector will never be the same again. In the spirit of competition, we should expect users to move toward Deepseek’s AI assistant given the cheaper cost and improved model.

As tech stocks adjust to the post-Deepseek world, drastic and beneficial changes to AI technology that put the user first like never before are on the horizon. Chatbots will be better, cheaper, and open source from here on.

Given the geo-political climate we find ourselves in, with a new sheriff in the Whitehouse in the form of United States President Donald Trump, the future is harder to predict than ever.

One should hope though, with Trump's positive track record imposing tariffs to leverage a deal, that there is a middle ground where both the U.S. and China can benefit each other with their AI innovations. This is not the Cold War, the future is not a replay of the past, paradigms are meant to be broken, and the new world with AI doesn't have to be one that retains the worst aspects of human nature.

FAQ

Is Deepseek Detectable?

Yes, like ChatGPT, this generative AI isn't trained to write using human style, it's trained to dispense high quality information. If you input text generated by Deepseek to an AI detector, chances are it's going to get flagged just as easily as ChatGPT.

StealthGPT vs Deepseek

If you're a student or professional hoping to leverage AI for content creation, there is still no substitute for high-quality human content. Bypassing AI Detectors isn't simply a function of standard improvements to AI, it takes a specific algorithm to crack the human writing code.