My point is current revenue from China are from established...

My point is current revenue from China are from established businesses that are publicly traded. There is only one company that I know of that is listed as a pure Machine learning company on the exchanges and that is C3.AI (also invested).
Appen must diversify away from ad dependencies but the industry is not mature enough to appease the impatient these juniors are still private. Quantum computing is indeed powerful but also pie in the sky stuff at this point as data produced is biased and rubbish. It will get much better but we are talking maybe 10-15 years before it begins making a material impact.
The junior private companies that build ML applications are data scientists.
Technologists don't work with messy data for good AI - the market isn't big enough yet simple as that but Appen has a stronghold on Big tech in US and China and thats a very good start. You can't train a machine to conduct precision heart surgery on synthetic data - nor can you train a machine to speak logically with synthetic data. That is years away.