Is AI Coming For OpenAI’s Job?

When the year began, America was in high spirits with the looming Donald Trump second era, and after the inauguration, it seemed he had the support of tech in the US. Shortly after, Donald trump announced Project Stargate, an ambitious $500 billion AI project backed by OpenAI, Oracle and Softbank. The four-year project is expected to improve and build AI infrastructure in the US.

Fast forward a few weeks later. DeepSeek, an open-source Chinese-developed AI model, dropped, sending shockwaves through the tech industry and causing stock prices to tumble. The model’s developers promised it was the most advanced AI yet, claiming it had been built in just a few months by a lean team with an alleged budget of only $5.6 million—a fraction of the cost of OpenAI’s ambitious project.

Less than 24 hours later, DeepSeek soared to the top of the most downloaded apps in the US Apple App Store. OpenAI could not keep quiet—this new player had arrived, and it was gunning for its crown. OpenAI’s director countered the narrative, alleging that DeepSeek had been trained on some of OpenAI’s own models and downplaying its seemingly miraculous cost-effectiveness.

But the disruption didn’t stop there. At the tail end of January, Alibaba announced the release of Qwen 2.5, its latest large language model developed by the Qwen team at Alibaba Cloud. The launch came with a bold proclamation. “Qwen 2.5-Max outperforms GPT-4o, DeepSeek-V3, and Llama-3.1-405B almost across the board.” The timing was strategic—unveiling Qwen 2.5 on the first day of the Chinese Lunar New Year made a statement. Qwen’s modular structure allows it to perform tasks like video analysis, image recognition, internet searches, and even has a developer mode for coding applications.

With AI development in China accelerating at an unprecedented pace, major players like DeepSeek and Alibaba Cloud are shaking up Silicon Valley’s dominance. Investors, who once believed in the near monopoly of OpenAI, Meta, and Google, are now questioning the sustainability of their hefty investments. The lean operational model of DeepSeek, combined with Alibaba’s aggressive push into the AI space, has put the industry on high alert.

Meanwhile, OpenAI is responding with new innovations of its own. The company recently rolled out o3-mini, a more affordable and efficient version of its reasoning models, bringing its advanced AI capabilities to a broader audience. While OpenAI’s reasoning models boast sophisticated “chain of thought” techniques for problem-solving, they are significantly more expensive to run than traditional models. o3-mini is an attempt to address this cost challenge, though it still remains more expensive than DeepSeek’s reasoning model.

The competition for AI supremacy is heating up at a breakneck pace. While OpenAI, DeepSeek, and Alibaba dominate the headlines, other major tech firms are also making moves. Microsoft, an OpenAI backer, has integrated DeepSeek’s R1 model into its offerings, while Amazon Web Services has also embraced the Chinese model. This signals a growing acceptance of non-US AI models in the global market.

Yet, as the AI arms race escalates, one question remains. Will Africa have a seat at the table? The continent has long been seen as a consumer rather than a producer of AI technology, but with the increasing accessibility of AI tools and cloud infrastructure, there is an opportunity for African start-ups and enterprises to carve out a niche. Governments and private sector players must act fast to invest in AI research, talent development, and strategic partnerships to ensure Africa is not left behind in this technological revolution.

2025 is shaping up to be the year of GenAI. With DeepSeek, Qwen 2.5, and OpenAI’s latest developments setting the stage for an intense battle, the landscape of artificial intelligence is being redefined before our eyes. The next few months will determine whether traditional AI giants can maintain their dominance or if the era of lean, cost-effective AI models will usher in a new order. One thing is certain—this is just the beginning.