Deepseek's emergence challenges US AI industry, Trump responds

Tuesday 28 January 2025 - 16:35

US President Donald Trump described the rise of Chinese company DeepSeek as a "wake-up call" for the US tech industry. This followed the release of DeepSeek's artificial intelligence (AI) model, which disrupted Wall Street and led to significant drops in shares of major tech firms like Nvidia, which saw its market value plummet by nearly $600 billion.

The root of the upheaval is DeepSeek’s claim that its R1 AI model was developed at a fraction of the cost of competitors, sparking concerns over the future of US dominance in AI and prompting questions about the scale of investment US companies are planning. Despite the stir, DeepSeek’s app became the most downloaded free app in the US just a week after launch.

Trump, commenting on the development, suggested that lower-cost alternatives could be a positive for the US if they help reduce expenses while achieving similar results. While he remained confident that the US would continue to lead in AI, there have been growing concerns over cybersecurity, especially from Australian officials who questioned the quality, consumer preferences, and data privacy management of DeepSeek’s model.

DeepSeek's R1 is based on the open-source DeepSeek-V3 model, which reportedly cost just $6 million to train, much less than the billions spent by competitors. This has raised eyebrows in the AI community, with some disputing DeepSeek’s cost claims. This comes at a time when the US has restricted advanced chip sales to China, prompting Chinese AI developers to adopt new, innovative methods that reduce computing power requirements and costs.

The impact on global markets has been mixed: the US market initially reacted with declines, but the FTSE 100 in the UK showed resilience, while Japanese AI-related stocks took a hit. The company, founded in 2023 by Liang Wenfeng in Hangzhou, China, has gained recognition for its AI technology, which rivals major US models, even without access to the latest hardware. Liang emphasized that the focus was always on efficiency and cost calculation, not on market reactions.

DeepSeek’s technology has attracted praise, notably from OpenAI’s Sam Altman, who acknowledged its impressive performance at a competitive price. However, experts like Marina Zhang from the University of Technology Sydney suggest that DeepSeek’s success shows how software innovation and data efficiency can compensate for hardware limitations. Databricks co-founder Ion Stoica also highlighted that DeepSeek’s low costs could encourage more companies to adopt AI, potentially accelerating AI progress.

Despite skepticism, including from tech figures like Elon Musk, who has questioned DeepSeek’s claims, the company continues to challenge industry norms with its ability to train models using fewer resources than initially thought necessary. The debate surrounding DeepSeek’s methods and their implications for the future of AI continues to evolve.