The world of artificial intelligence has taken a dramatic turn as Chinese company DeepSeek has released AI models that rival those of OpenAI, the pioneer behind ChatGPT. This move has sent shockwaves through Silicon Valley, questioning the notion that US companies hold a secret recipe for building AI and require enormous resources to do so.
Sam Altman, CEO of OpenAI, had previously claimed that scale was key to progress in AI development. However, DeepSeek’s latest achievement threatens to upend this narrative. The Chinese company’s AI models are reportedly capable of matching OpenAI’s best offerings but at a significantly lower cost and computational power requirement.
Despite the success of DeepSeek’s model, concerns have been raised about its limitations. The AI’s answers adhere closely to China’s government party line, and it even self-censors in real-time. Security researchers have also warned that the model lacks robust guardrails against inappropriate use.
The arrival of DeepSeek on the scene brings both welcome competition and new concerns. As more players enter the generative AI market, the stakes are higher than ever. However, cheaper models that require less computational power may open doors to entirely new applications for the technology, potentially making it more affordable and environmentally friendly.
This development marks a significant turning point in the history of AI, highlighting the need for innovation and regulation in this rapidly evolving field.
Source: https://www.newscientist.com/article/mg26535293-300-deepseek-has-burst-the-ai-hype-bubble-now-all-bets-are-off