🧔♂️ A friendly human may check it before it goes live. More news here
Chinese AI companies lag behind Western firms: DeepMind CEO
DeepMind CEO Demis Hassabis said that Chinese AI firms are about six months behind leading Western labs in technological innovation.
Hassabis noted that while Chinese companies like Hangzhou-based DeepSeek have developed impressive models such as the R1 reasoning system, they have yet to show the ability to innovate beyond the current frontier.
The restrictions on access to advanced AI chips due to US export controls have limited Chinese firms’ development, prompting reliance on unconventional methods.
The US is easing some export bans on AI chips to China, but sales of the most advanced processors remain blocked on national security grounds.
Hassabis highlighted ongoing efforts at DeepMind in robotics and the Gemini AI assistant, which integrates data from Google services, and predicted a future breakthrough in physical intelligence, emphasizing the challenge of replicating human hand dexterity.
🔗 Source: Bloomberg
🧠 Food for thought
Implications, context, and why it matters.
Key details on the US chip export policy shift remain unclear
- The exact scope of the eased US export rules needs to be spelled out.
- It needs to be verified which AI processor models are allowed for sale to China, plus whether Nvidia’s H200 is included.
- Licensing rules need to be laid out, along with the start date for the change.
- With those points unresolved, the effect on Chinese AI firms remains hard to judge. A roughly six-month gap was cited by Demis Hassabis, and closing it depends on access to high-end compute.
A compute-supply gap could open for independent data center operators in China
- Equipment lessors plus regional data center operators in China could supply compliant AI compute clusters to buyers facing shortages.
- A separate estimate puts Baidu and Huawei above 70% of China’s GPU (graphics processing unit) cloud market in the first half of 2025 1. It does not connect that share to any move to favor their own chip architectures.
- Independent providers may step in with managed services plus financing for Nvidia-based hardware. Some Chinese developers still prefer it, yet supply has been tight under US export controls 2.
Recent DeepMind developments
Stay updated on the go with our mobile app.
Get latest insights with smoother, more personalized experience through TIA mobile app.




