The release of a new artificial intelligence (AI) model, ironically, evokes a biblical memory. What has transpired in the past few days echoes the story of David versus Goliath, in which the massive and well-armed Goliath is defeated by the comparatively puny David, who comes to the battle with only his staff and sling. 

The modern-day equivalent of David that has set the entire world talking is Chinese company DeepSeek, whose advanced open-source language model DeepSeek V3 provides an alternative to OpenAI’s ChatGPT with better efficiency and a fraction of the cost. The emergence of the AI David has stunned Silicon Valley and shaken Wall Street within days of its release, causing the value of US tech stocks to plummet by nearly $1 trillion. 

There are very few examples of such events occurring within the tech industry nowadays as major breakthroughs are increasingly few and far between, entailing years if not decades of work and astounding amounts of resources. Or that’s what Silicon Valley thought. 

With the launch and rapid rise of ChatGPT in 2022, AI became a trending buzzword and the push for AI dominance saw billions upon billions of dollars spent in funding, resources, and computing power. To bolster their lead, the Western “free world” imposed stringent restrictions on access to core technologies and chips essential to developing these technologies. 

US tech plutocrats were present in the front row at the US presidential inauguration in January, where President Donald Trump heaped praise upon them and announced that the private sector, represented by OpenAI, SoftBank and Oracle, would invest up to $500 billion to build AI infrastructure in the US. The atmosphere was one of a victory parade. 

But the victory became hollow as DeepSeek revealed that it had attained competitive parity with OpenAI’s most advanced model, using substantially fewer resources, with slower hardware due to the restrictions, and in considerably less time. The team also released a detailed research paper laying out the combination of breakthroughs and sheer ingenuity that led to the feat. The stunning parity of DeepSeek-V3 with OpenAI’s most advanced model at a fraction of the time, cost, and resources speaks volumes about the level of talent and ingenuity of the DeepSeek team, comprising China’s new generation of homegrown talent and institutions.

Momentous development  

The event is momentous for multiple reasons. The first is that it dispels the notion that Silicon Valley has “won” the AI race and was firmly in the lead in a way that could not be challenged because even if other countries had the talent, they would not have similar resources. Besides, access to the most advanced American-made chips is only given to close partners and allies of the US. This control, people thought, would ensure total US dominance. 

If, by some miracle, any non-American company managed to achieve a breakthrough, the assumption was that it would take ages. However, DeepSeek has not only achieved the impossible but also, in a move straight out of Sun Tzu’s The Art of War, said its model is both free and open-source. This is a game changer on a tectonic level whose ramifications will ripple across time. It will alter the trajectory of AI development and application. 

OpenAI, contrary to its name, is anything but open. Its valuation was based upon two things: its proprietary trained large language model, and ownership of the vast computing resources – the hardware and software needed for processing data, running applications, and tackling problems. 

DeepSeek-V3 is open-source, which means anyone can download and run it, train it, and tweak it as they please, depending on the available hardware. By making it a public good meant to benefit all, DeepSeek has effectively re-written the AI rulebook and redrawn AI’s technological landscape. 

DeepSeek is also collaborating with Huawei, another Chinese tech giant, and their new AI-focused Ascend series of chips, a milestone in China’s budding AI hardware industry. As former CEO of Intel and tech industry veteran Pat Gelsinger said, “DeepSeek will help to reset the increasingly closed world of foundational AI model work. Thank you, DeepSeek team.” 

Something else is of great importance. Now the Global South, their academics and their universities can also harness AI through unfettered access to an advanced model. This will have a profoundly positive impact, as previously the Global South was at the mercy of Western technology oligarchs. Now, reaping the benefits of AI is within the grasp of the developing and underdeveloped countries. A brave new world awaits.

African Times has published this article in partnership with ChinAfrica Magazine