DeepSeek Beats OpenAI's ChatGPT, Turns No. 1; 5 Reasons Why Chinese App Created Buzz

It was a David vs Goliath battle on the digital landscape represented by Apple App Store and shockingly, the former emerged victorious! Notably, DeepSeek’s AI Assistant today muscled out top rival OpenAI's ChatGPT to take the top spot among the free apps on the US Apple App Store. Inevitable? Read all about it.

So, for most people, Chinese startup DeepSeek was a “Johnny-come-lately” on the artificial intelligence front till recently. But it has just taken the Silicon Valley tech bulls by the horns and brought them to their knees.

As DeepSeek upended the AI sector with its ground-breaking product release that is being lapped up at breath-taking speed, global markets and tech stocks reacted. Friday had already seen a decline in Wall Street's major indexes due to selloff of tech stocks after the DeepSeek AI programme was introduced. Today, Nvidia Corp. and ASML Holding NV stocks plummeted, and questions were posed about the future of US tech behemoths like Google, and Meta apart from OpenAI, of course. Also, Microsoft as its AI fate is intertwined with that of OpenAI.

DeepSeek, a little startup based in Hangzhou, China, was founded in 2023 by a hedge fund and AI veteran Liang Wenfeng, and is entirely financed by Wenfeng’s quantitative hedge fund, High-Flyer. The startup was formed around the time Baidu unveiled the first Chinese AI large-language model. However, little else is known about the startup.

According to app data analytics company Sensor Tower, after its Jan. 10 release, DeepSeek’s AI Assistant—which is based on the DeepSeek-V3 model—has seen tremendous growth. As per DeepSeek developers, the DeepSeek-V3 model competes with some of the most cutting-edge closed-source models in the world and performs much better than many open-source models.

Shockingly, according to a paper reportedly published last month by DeepSeek researchers, DeepSeek-V3 model was trained using Nvidia’s H800 CPUs, with less than $6 million spent on training. DeepSeek avoids the need for expensive hardware that is in any case subject to US export restrictions by using open-source technology and less expensive chips, in contrast to its US competitors.

Notably, Nvidia has built and staked its reputation on its latest AI chips that the world has been buying with gusto. It has also led to Nvidia shares skyrocketing. The rise of DeepSeek, has put a huge question mark on Nvidia's actual valuation. This has been indicated by the fall in its share price.

This AI model that was launched just weeks back is believed by many as being equally competitive if not better than OpenAI. Although several Chinese IT companies have in the past released AI models, they have made little news out in the Western tech circles. DeepSeek is the first to be acclaimed in this manner for matching or even outperforming the top AI models coming out of the U.S.

However, there is a lot of risk looming over it. Considering it is a Chinese company, experts noted that it faces computational disadvantage because US has slapped export limitations on China regarding cutting-edge semiconductors. To grow its business and compete internationally, DeepSeek will need to leverage advanced chips. Nevertheless, it needs to be noted that DeepSeek has come this far without depending on Nvidia's latest AI chips and if it ever gets access to them, its performance is expected to skyrocket.

Notably, the US has restricted sales of cutting-edge Nvidia AI processors to numerous nations, and prohibited the export of sophisticated semiconductor technologies to China in an effort to slow that nation's advancements in AI. That said, the recent DeepSeek rally raises the possibility that China’s AI development may not be as limited as previously believed. And that is visible to all the world as the top tech companies on Wall Street are virtually shedding blood.

Explained in brief: 5 reasons why DeepSeek has created a buzz

1. Industry-Leading Capabilities: DeepSeek R1 boasts impressive capabilities that rival major players by demonstrating its potential to not just compete effectively in the AI landscape, but to beat the best at the game.

2. Free and Unlimited Usage = Cost Efficiency: The core reason for the buzz is its unique offering of free and unlimited usage. This removes financial constraints for users within the AI domain. Unlike many AI models that require subscriptions, DeepSeek R1 offers unlimited free usage. This removes a significant barrier to entry for anyone (individuals to businesses), thereby making advanced AI more accessible to all - a very democratising move. DeepSeek R1 provides access to high-performance AI capabilities without the sky-high costs typically associated with such technologies.

3. Open To All In Every Way: DeepSeek R1 is open-source, meaning its code is publicly available. This fosters transparency, allows for community contributions and improvements, and accelerates AI innovation. It empowers researchers and developers to tailor the model in their own way.

4. Efficiency: DeepSeek R1 requires less computational power compared to some other large language models. This can translate to lower energy consumption. Not just that, it will also reduce its ecological footprint. Environmentalists may find that an endearing quality.

5. Potential for Disruption: DeepSeek R1's unique combination of capabilities, cost-effectiveness, and open-source nature has the potential to disrupt the AI landscape and pave the way for development in ways still undreamt of.