3/ Finally, in line with global research trends, Chinese firms are experimenting with techniques that would reduce the computational intensity of training large-scale AI models. They're effectively trying to squeeze more runway out of less chips or less powerful chips.
@_karenhao
-
Chinese Firms Develop Heterogenous Computing for AI Model Training
By
–
2/ Chinese firms are also trying to advance so-called heterogenous computing, training large-scale AI models across different types of chips, to avoid a reliance on any one type of hardware. This is rarely seen among US firms because of the technical challenges to get it to work.
-
Chinese Firms Bypass AI Chip Sanctions Using Multiple Downgraded Processors
By
–
We discovered three strategies: 1/ Chinese firms can no longer buy Nvidia A100s & H100s, the best AI chips, but they can buy downgraded A800s & H800s, which Nvidia designed to meet sanction requirements. The firms are now trying to use 3 or 4 x A/H800s to simulate 1 x A/H100.
-
Chinese AI Firms Navigate US Sanctions Without Advanced Chips
By
–
US sanctions are driving Chinese firms, like Huawei, Alibaba, Baidu, to seek ways to advance AI without cutting-edge chips. We dug through open-source research papers and spoke with employees & analysts to understand what they're doing. w/ @raffaelehuang
-
ICLR 2023 in Rwanda: Reflections on Africa’s First Major AI Conference
By
–
That’s a wrap! I had so many amazing conversations and learned so much from this year’s ICLR, the first major AI research conference in Africa. Here are my initial reflections. See you next time, Rwanda.
-
First Major AI Research Conference in Africa Begins in Kigali
By
–
Hello from Kigali and the very first major AI research conference in Africa, which kicked off today with a stunning performance!!
-
Karen Hao Joins Yale Panel on US-China Tech Relations and AI
By
–
So thrilled to be joining @yangyang_cheng
, @SammSacks
, Nick Frisch, and so many other people I admire at Yale tomorrow and Thursday to discuss all the topics near and dear to my heart: science & technology in U.S.-China relations, data flows, AI development, and journalism. -
Karen Hao Discusses Generative AI with WSJ Readers
By
–
I had so much fun answering @WSJ reader questions about generative AI. We ran the gamut from ChatGPT hallucinations to AI boyfriends, with a brief interlude into quantum computing. Loved the curiosity! Thanks @zoegthomas & @WSJPodcasts for having me!!
-
Fake URLs vs Working URLs: ChatGPT Misinformation Risk
By
–
Yea, that’s definitely worse. In this instance, you click on the url and realize it’s fake. In the other instance, you click on the url, it works, and you may just take ChatGPT’s summary at face value.