📖 Top 10 China's AI Stories in 2024: A Year-End Review
Qwen, DeepSeek, 01.AI and others show China's paradigm of AI might make more progress in 2025 vs. the Western Cloud sponsored duopoly Labs OpenAI and Anthropic and even Meta, Google, etc...
One of the biggest macro trends in AI in 2024 for me was how well a couple of solutions out of China are positioning themselves as Open-source frontier model builders. Not just leaders, but actual pioneers. Let’s be frank and honest, DeepSeek-V3 really does seem to point to a new direction of China’s capabilities in Generative AI in 2025.
“DeepSeek (Chinese AI co) making it look easy today with an open weights release of a frontier-grade LLM trained on a joke of a budget (2048 GPUs for 2 months, $6M).” - Source.
How could a Quant fund turned Open-source model pioneer change the narrative so much near the end of 2024? CEO Liang Wenfeng’s main venture was High-Flyer (幻方), but today it’s a new paradigm for China’s AI community. The Western media keeps underplaying the capabilities of Chinese Gen AI startups, but for how much longer can they keep doing that?
DeepSeek has the sort of level of transparency you might have hoped from OpenAI years ago, that by the way never materialized. Meanwhile Microsoft and OpenAI decided on a commercial number of when AGI was “achieved”. The two companies signed an agreement in 2023 that defined AGI as a system that can generate $100 billion in profits, The Information reported. While I’m not a huge fan of how product-centric OpenAI and Anthropic now have to perform, innovation does not belong to any single country or region.
Last week, my mom, 58, won a local dancing competition and was asked to give an award speech. Normally, this kind of task would fall on me – just as many others do. But this time, she told me she’d figured it out using an AI app. “The AI wrote it better than you ever could,” she said, implying that she might not need my help anymore.
In 2024, keeping up with AI’s rapid development in China has been a challenge. There are endless stories to tell – from frontier models to real-world applications. Every company is eager to sprinkle a bit of AI magic into its narrative.
There are a lot of Washington DC eyes on China and its news cycle, but few cover its technology and AI community well. Of these Tony Peng is easily among my favorite and those I recommend to my serious readers. The rate of updates that are far more economical of Qwen and DeepSeek alone are showing technical brilliance in China.
Tony’s background as an AI reporter for Synced really shines brightly on the ecosystem. As Global Head of Communications at Baidu his clarity of expression and powers of synthesis makes reading his work tremendously easy. The Recode China AI blog reflects his personal opinions, it obviously should be noted that it does not represent the views of Baidu. But as a China tech nerd suffice to say I hold Tony’s opinion in high regard.
This deep dive is made available to premium readers.
Keep reading with a 7-day free trial
Subscribe to AI Supremacy to keep reading this post and get 7 days of free access to the full post archives.
