Trillions have been wiped off the value of Nvidia stock overnight after the release of a Chinese artificial intelligence model, that after being tested by US scientists was found to not be reliant on expensive NVidia processors to outperform.
Also falling on the announcement of DeepSeek a free, open-source large-language software offering was Google, Microsoft, and Tesla.
Also hurt in the sell off was Apple, Google parent Alphabet, Meta and Amazon who are now struggling to come to grips with the implications of the announcement.
Close to two $2 trillion dollars’ worth of value was wiped from the stocks with more falls tipped today.
The announcement has also cast doubt over the thesis behind the industry’s entire supply chain and whether the USA actually has an advantage over its chief geopolitical rival a move that new President Donald Trump was pushing last week.
DeepSeek released the ‘R1’ version of its AI a week ago, claiming its performance is ‘on par with OpenAI-o1’, which was unveiled late last year.
The novel thing about both models is how they approach the inference side of AI, which is the process of tapping expensively trained large language models, in response to queries, to yield the best outcomes.
It has taken the rest of the world a week to assess open-source DeepSeek-R1 and it seems to have come to the conclusion that its claims are legitimate. The main reason it’s so disruptive is that the US state has gone to great lengths to deny China the hardware used by the likes of OpenAI – especially Nvidia GPUs. So the latest news from DeepSeek suggests that effort has failed in its stated aim of preventing Chinese AI development.
DeepSeek claims to have developed the advanced model in just two months at a cost of under $6 million — and without access to Nvidia’s best computer chips.
The model’s launch — referred to by tech investor Marc Andreessen as “AI’s Sputnik moment” — triggered a global meltdown that slammed AI firms and chipmakers.
“The DeepSeek news was a Scud missile aimed at a US market that is much more fragile than most will admit,” said Island Capital Investments CEO Anthony Esposito, who noted that AI hype has driven markets to all-time highs.
DeepSeek is a Chinese AI company, which a week ago launched its latest AI model, which it calls R1.
The company said the model was particularly good at problem solving, performing on par with OpenAI’s o1 reasoning model—but at a fraction of the cost per use.
A DeepSeek app is currently top in iPhone download rankings for the U.S.
The Wall Street Journal claimed that conventional thinking prior to the announcement was that AI companies needed expensive, leading-edge computer chips—such as those made by Nvidia—to train the best systems.
That has justified huge spending by the biggest U.S. tech companies, such as Alphabet and Meta Platforms, which are sometimes known as hyperscale’s.
Just last week, companies including SoftBank, Oracle and OpenAI pledged to spend $500 billion to build new AI infrastructure in a venture they call Stargate.
Acording to Ryan Cox, Global Head of Artificial Intelligence at Synechon, DeepSeek shook up the tech-world overnight, the global tech consultancy currently works with three of the large banks in Australia,
The Company claims that “Synechron’s testing of DeepSeek’s 32B parameter model revealed both promising capabilities and important considerations for bias mitigation and compliance standards”.
“What sets DeepSeek apart is its efficient architecture, innovative training techniques, and optimised resource management. This opens the door for businesses of all sizes to adopt advanced AI without the exorbitant costs. Beyond cost savings, this democratisation of AI introduces new competition, spurring better results and innovation for end users worldwide. It also underscores the importance of leveraging global AI talent.
“Synechron’s testing of DeepSeek’s 32B parameter model revealed both promising capabilities and important considerations for bias mitigation and compliance standards before rolling it out more broadly.
“The real challenge for enterprises isn’t just cost optimisation – it’s governance. While open-weight models offer great customisation potential, they demand robust validation frameworks. Some models are developed with some level of censorship, particularly concerning sensitive political topics, cultural issues, or content that might be considered inappropriate or against Chinese government policies. Open-weight models lack built-in security certifications, placing the burden of compliance on the deploying organisation. CIOs and technology leaders must establish rigorous governance and validation frameworks to qualitatively ensure their GenAI solutions meet performance, security, and ethical benchmarks.
“Looking ahead, successful AI implementation in 2025 will require balancing innovation with practical governance. Organisations and technology leaders should focus on three key areas: establishing clear validation frameworks for open-weight models, developing robust compliance protocols, and creating flexible architectures that can adapt to this rapidly evolving landscape”
But DeepSeek didn’t have leading-edge chips—and its models appear to be roughly on par with top U.S. rivals on certain benchmarks that evaluate AI ability. DeepSeek says it uses less-advanced chips, combined with innovative model-training techniques.
In addition, DeepSeek released its R1 model as open source, meaning others can pick up and adapt the model for their own use.
That will mean that other companies will be able to build on DeepSeek’s approach and potentially create other cheap AI alternatives.
DeepSeek also published costs for using R1 that were an order of magnitude below those charged by U.S.-based companies for their most sophisticated models.
In a paper published last week, the Chinese researchers behind DeepSeek said their new model would sometimes suddenly stop and realize it should re-evaluate its initial approach to a problem and allocate more thinking time to do so.
They described the behaviour as the model having an “Aha!” moment.
