Mon, 27 Jan 2025

Is China's AI tool DeepSeek as good as it seems?

The artificial intelligence (AI) tool has shocked US markets after bursting onto the scene.
A new Chinese AI-chatbot app called DeepSeek has caused chaos in US markets and raised concerns about America's dominance in artificial intelligence. The app, which launched last week, uses a similar model to OpenAI's ChatGPT and Google's Gemini, providing long and informative responses to user questions. However, unlike its US rivals, DeepSeek often begins its responses by saying the topic is "highly subjective" and may not express an opinion even when directly asked. The app also has access to current information, but only in certain areas, which some speculate is a way to prevent the spread of misinformation. One notable difference between DeepSeek and US-developed chatbots is how it handles sensitive topics banned in China. When asked about subjects such as the Tiananmen Square protests, DeepSeek censors itself, refusing to provide an answer or even discuss the topic further. The app's ability to self-censor has raised questions about its potential impact on free speech and information access. Researchers have also noted that the Chinese government's "hands-off" approach to the app may indicate a shift in strategy, with increased investment in AI development expected in the near future. DeepSeek is built using existing technologies from companies like Nvidia, Meta, and AliBaba, which has sparked debate about the cost-effectiveness of AI development. Some experts believe that DeepSeek's model could signal a shift towards more resource-intensive methods of developing artificial intelligence, potentially challenging traditional monetization strategies in the industry.
  >>