Most of the progress in AI is currently happening in the United States, but there’s a nation that’s silently making some quick moves as well.
Former Google CEO Eric Schmidt has said that he’s “shocked” by China’s progress in AI. “This week there were two libraries from China that were released open source,” he said in an interview. “One is a problem solver that’s very powerful, and another one is a large language model that’s equal to, and in some cases exceeds the one from Meta, which they use every day. It’s called Lama 3 (with 400 billion parameters). I was shocked when I read this, cause I had assumed that, in my conversation with the Chinese, that they were two to two to three years late,” he added.
But Schmidt said that China was a lot closer to the US in AI progress than what many believed. “It looks to me like China is within a year (from the US) now. So it’d be fair to say (the main players are) the U S and then China, which is within a year’s time. Everyone else is well behind,” he explained.
Schmidt said that this could have security implications for everyone involved. “Now, I’m not suggesting that China will launch a rogue attack against an American city, but I am alleging that it’s possible that a third party could steal from China, because it’s open source, or from the US, if they’re malevolent, and do that. So the threat escalation matrix goes up with every improvement,” he added.
China has been silently releasing strong AI models from its various labs and universities. In particular, China’s video generation models, such as those produced by Kling.ai, are world-class, and often at par with the best models from the US. China is also innovating on producing high-quality models cheaply, and is looking to build models that cost a fraction of what is US is spending on building them.
Interestingly, India has been largely missing from all these moves. There are hardly any companies building their own models in India, and even those models are being generally focused on Indian languages, and not looking to create general-purpose models. Even India’s tech leaders like Nandan Nilekani have said that India needs to focus on discovering use-cases for AI instead of building models itself. It remains to be seen how that strategy will play out, but the top two economies in the world in the US and China seem to be pulling ahead of the rest of the pack in their technical advancements in AI.