Wang Jin, Nadav Kunievsky and Bowen Lou of the Chapman, Chicago and Southern California universities (et al.) describe how the U.S.’s attempts to retard China’s AI development ambitions have backfired.
A decade or so ago the U.S. government decided the best way to crimp China’s technological progress was to restrict access to the nuts and bolts of technology development i.e. hardware and related systems, flowing at that time freely into China.
Since then China’s response has been to promote the use of open-source models at the forefront of its technological development, which now most importantly means AI, whilst AI development in ‘The Land of the Free’ has been almost exclusively that of the walled-garden variety.
China’s preference for an open-source AI ecosystem is a direct response to U.S. government policy. Open-source promotes faster development, encourages wider adoption and is implicitly a more robust architecture independent of externally controlled systems.
This is a game still in progress so the paper doesn’t get into the business of where this is heading so I’ll throw in my own two-pennyworth. Isn’t this like the VHS versus Betamax tussle?
VHS triumphed (it seems an urban myth that pornography had a lot to do with it, I checked) because it was good-enough and cheaper than the alternative. That was it. Sound familiar?
You can read the paper in full via this link U.S. Policies Unintentionally Accelerated China’s Open AI Ecosystems.
Happy Sunday.