A Chinese AI lab called DeepSeek has sparked controversy by claiming its model is similar to ChatGPT and was built at a fraction of the cost, using less computing power than US-based labs like OpenAI.
OpenAI, the developer of ChatGPT, is now investigating allegations that DeepSeek used its intellectual property to build its competitor. The company’s spokesperson confirmed it is reviewing the claims but has yet to provide further details.
The controversy centers on a technique called “distillation,” where developers train smaller AI models using the performance of larger ones. While OpenAI claims this practice is common, some tech experts argue that DeepSeek may have used it in an unauthorized manner.
Tech venture capitalist Bill Gurley notes that many AI insights and breakthroughs are based on previous work, and he questions why OpenAI is suddenly objecting to DeepSeek’s claims. Lutz Finger, a senior visiting lecturer at Cornell University, suggests that distillation is a standard practice in the industry and may be difficult to detect.
The incident highlights the need for clearer norms around fair use versus unauthorized replication in the rapidly expanding AI industry. As one tech expert explains, “we’re entering a phase where the AI community will have to define clearer norms.”
Source: https://edition.cnn.com/2025/01/30/business/openai-deepseek-nightcap/index.html