Tim Cook lauds China’s DeepSeek, calls its AI models ‘a good thing’

Apple CEO Tim Cook has praised the AI models of Chinese AI startup DeepSeek for driving "efficiency" in the field of artificial intelligence. Notably, DeepSeek had recently taken over ChatGPT on US app stores, causing palpable panic among AI giants as any notions of American supremacy in the field were shattered.

During Apple's earnings call with investors, Cook was asked how DeepSeek's AI models would affect Apple's margins, to which the veteran tech executive said (via TechCrunch): “In general, I think innovation that drives efficiency is a good thing. And, you know, that’s what you see in that model,”

Cook was also asked about Apple's AI strategy, to which he went on to explain the ‘hybrid’ approach taken by the company, saying (via Business Insider), "From a CapEx point of view, we've always taken a very prudent, deliberative approach to our expenditure, and we continue to leverage a hybrid model, which I think continues to serve as well,"

Notably, Apple currently only has a partnership with ChatGPT, which allows the company's Siri voice assistant to direct some complex queries to OpenAI's chatbot with the user's permission. While Apple has said in the past that it's open to using AI models from Google and Anthropic, no concrete agreement has yet been announced. Cook also didn't confirm whether Apple might integrate DeepSeek's open-source models in a future upgrade.

OpenAI vs DeepSeek:

DeepSeek claimed in a paper last month that its V3 language model was built at a cost of just $5.6 million and trained on older-generation Nvidia H800 GPUs, compared to the billions spent by Google, Anthropic and OpenAI on building similar models.

However, OpenAI on Thursday claimed that DeepSeek was employing a technique called ‘distillation’ in order to train its AI models on train its rival models, potentially undermining some of the achievements of the Chinese upstart.