Wed Feb 05 08:59:00 UTC 2025: ## Chinese AI Startup DeepSeek Disrupts US Tech Giants with Budget-Friendly Rival to ChatGPT
**Melbourne, Australia** – A previously unknown Chinese AI company, DeepSeek, has sent shockwaves through the US tech industry with the release of its new AI model, R1. The model, a “reasoning” AI capable of detailing its problem-solving steps, rivals OpenAI’s advanced o1 model in capability but at a significantly lower cost.
DeepSeek’s app quickly surpassed ChatGPT in new downloads, causing a dramatic drop in US tech company stock prices and prompting OpenAI to accuse DeepSeek of improperly using its data to train R1. OpenAI alleges that DeepSeek “distilled” its models – a common machine learning technique where a smaller model learns from a larger one’s predictions – by querying OpenAI’s o1 extensively. DeepSeek has not commented on the allegations.
While OpenAI claims DeepSeek’s actions constitute intellectual property theft, OpenAI itself is facing numerous lawsuits over copyright infringement in its own model training data. The company’s claim of “fair use” of copyrighted material scraped from the internet is currently being challenged in court.
DeepSeek’s rapid success may also be attributed to its innovative use of cheaper AI chips, a necessity given US export bans on advanced chips to China. This forced DeepSeek to optimize its model for less powerful hardware. Although a recent report suggests DeepSeek’s hardware expenditure reached US$500 million, the company still achieved remarkable efficiency compared to its rivals.
The lack of effective methods to definitively prove data theft, such as AI output watermarking, further complicates the situation. The controversy highlights the rapid advancement of AI technology outpacing the development of clear legal frameworks.
Over the weekend, OpenAI responded by releasing its own advanced model, o3-mini, claiming superior performance and cost-efficiency. This escalation marks the beginning of a new era of increased consumer choice and competitive pricing in the AI market, a positive development for users. The improved efficiency of these newer models also offers environmental benefits by reducing the massive energy consumption associated with AI training and use.
The DeepSeek incident underscores the need for a broader conversation regarding data collection and usage practices within the AI industry, and the ongoing legal battles will play a critical role in shaping the future of AI development and regulation.
First Piper 