Following generative AI trends can be daunting at times, with multiple scientific breakthroughs announced frequently, sometimes simultaneously.
In the latest episode, DeepSeek has tumbled to #51, weeks after dethroning OpenAI’s ChatGPT as the most downloaded free AI app in Apple’s App Store in the United States.
Chinese AI startup DeepSeek R1 V3-powered model has been a hot topic in the AI landscape, predominantly because of its ultra-cost-effective alternative to proprietary AI models like OpenAI’s o1 reasoning model.
DeepSeek has tumbled to 51, while American AI apps still rule the app charts. How come no one is posting or writing about this? I guess the news isn’t sensational enough for the reporters who wrote about DeepSeek the past month. pic.twitter.com/lAInNpShNWFebruary 27, 2025
Meta’s AI lead scientist Yann LeCun attributed DeepSeek’s overnight success to it being an open-source model:
“They came up with new ideas and built them on top of other people’s work. Because their work is published and open source, everyone can profit from it. That is the power of open research and open source.”
DeepSeek’s launch led to a dramatic shift in the market, causing NVIDIA’s stock price to plummet. This translated to a $500 billion loss in market valuation (arguably the biggest loss in history in a single day).
The AI model raised investor concern after it was revealed that it gave proprietary models from sought-after companies, including Meta’s Llama 3.1, OpenAI’s GPT-4o, and Anthropic’s Claude Sonnet 3.5, a run for their money at a fraction of their development cost.
DeepSeek researchers say the R1 model surpasses OpenAI’s o1 reasoning model capabilities across math, science, and coding at 3% of the cost.
However, a separate report suggested that there was more to the R1 model than the researchers were leading on.
DeepSeek reportedly spent more than the highlighted $6 million and 2,048 AI GPUs to train its R1 model. The report details that the Chinese AI startup spent up to $1.6 billion in hardware, including 50,000 NVIDIA Hopper GPUs.
To that end, it’s increasingly becoming difficult to pinpoint the cause of DeepSeek’s downward trajectory, especially after its broad adoption during its launch.
While OpenAI CEO Sam Altman touts DeepSeek’s cost-effective model as impressive, he indicated that OpenAI is better positioned to develop models, potentially creating a clear path toward AGI.
The world is going to want to use a LOT of ai, and really be quite amazed by the next gen models coming. Look forward to bringing you all AGI and beyond.
OpenAI CEO, Sam Altman
This happened after OpenAI had made its $500 billion commitment to its Stargate project, designed to facilitate the construction of data centers across the United States for advanced AI projects.
Microsoft CEO Satya Nadella praised DeepSeek’s model as “super impressive,” urging stakeholders to monitor developments in China.
DeepSeek faces several issues, including the stringent AI chip exportation rules imposed by the Biden-Harris administration, blocking the shipment of chips to China for AI advances.
Former OpenAI board member Helen Toner indicated that DeepSeek is not leading the pack in the AI landscape, but if US President Donald J. Trump revokes the NVIDIA chip export ban, it might be a huge win for China.
“It doesn’t make any sense to try to walk it back. We’ve paid the price … putting China on high alert that chips are a strategic technology, and incentivizing the whole global supply chain to avoid using U.S. components so as not to be subject to extraterritorial controls.”
Is AI just but a fad?
Last year, a report suggested that AI might be a fad, further highlighting that users aren’t incorporating advanced tools into the workflows for more efficiency and effectiveness.
More recently, during Windows Central’s weekend discussion on AI and its usefulness, it became apparent that more users are seemingly hopping onto the AI bandwagon.
The vast majority of the audience indicated that they use Microsoft’s Copilot and ChatGPT for a wide range of tasks. However, some users cited critical accuracy and consistency issues.
Last year, a detailed study by Gartner predicted that 30% of AI-themed projects will be abandoned by investors by 2025 after proof of concept.
Gartner attributed its findings to top executives struggling to identify a profitable path despite their hefty investment in the landscape:
“After last year’s hype, executives are impatient to see returns on GenAI investments, yet organizations are struggling to prove and realize value. As the scope of initiatives widen, the financial burden of developing and deploying GenAI models is increasingly felt.”
Interestingly, a separate report suggested that an overreliance and dependence on these AI tools may lead to critical health issues, including the deterioration of cognitive faculties. The Microsoft study further indicated that AI ‘atrophies’ critical thinking.
Aside from establishing a clear path to profitability, generative AI still faces a fair share of challenges as it rapidly advances, including high operation costs, high electricity and water consumption, and more.
This happened after multiple reports indicated that top AI labs, including OpenAI, Google, and Anthropic, had hit scaling laws, preventing them from developing advanced AI models.
However, top executives, including OpenAI CEO Sam Altman and former Google CEO Eric Schmidt, disputed the claims, indicating that there was no evidence that scaling laws have begun stunting AI progression. “There’s no wall,” added Altman.
To that end, it will be interesting to see how DeepSeek competes against the emerging AI models with reasoning capabilities or if it disappears into the horizon before hitting its prime time.
What’s your experience with AI-powered tools like OpenAI’s ChatGPT, DeepSeek, or Microsoft Copilot? Share your thoughts with us in the comments.
