Justin Lin, the head of Alibaba’s Qwen team, estimated the probability of Chinese companies surpassing leading players like OpenAI and Anthropic through fundamental breakthroughs within the next 3 to 5 years to be less than 20%.
Even for their own services – i.e., inference – they’re consuming so much capacity that they don’t have enough compute left to devote to research.
“After a year of gung-ho news about China’s gains in artificial intelligence, some elite Chinese AI researchers are coming to a more pessimistic conclusion. The country’s chances of catching up to the U.S. are slim in the short run, they say, because of a bottleneck in chips.
“The truth may be that the gap is actually widening,” Tang Jie, founder of the Chinese AI startup Zhipu, said at a conference last weekend in Beijing.
chippawanka on
Obviously. China is all propaganda… it’s a house of cards and nothing more
Getafix69 on
They’ll just pour money into it and get there faster than anyone predicts again.
Joseph20102011 on
That’s why China wants to invade Taiwan next year so that it will have better chips by grabbing TSMC.
jon_the_mako on
Try Fritos. They are the working man chips. No flash, no big marketing, no fancy shapes, flavors are plain and chili cheese.
Spara-Extreme on
I’m actually super excited for China to invest in chip technology that can break nvidias stranglehold
-Mediocrates- on
Deep Seek just literally released a new version with a one of a kind scaling architecture. Soooooo dunno man… looks like improvements being made.
esmelusina on
There is an upper ceiling in what AI can practically do and accomplish and in what ways it can _actually_ benefit business or humanity.
Are Chinese tech leaders doomposting to get more funding and support?
US AI folks are all VC motivated and have been culturing an insane hype bubble for years. If China buys that hype bubble, they would ofc feel behind, but I’m not sure how true that is.
China seems to leverage AI in a way that’s more responsible and meaningful, even if it isn’t as scaled or appealing in an investment sense.
Seems like Chinese tech sector is using the hype bubble as an excuse to get more share of gov funding.
Grombrindal18 on
Has China considered being nice to Taiwan, respecting their sovereignty, and asking politely for some of their chips?
Electroboy101 on
This is EXACTLY the sort of shit I would also be saying if I was trying to make my enemy think I was useless so they might as well sell me those enabling chips I need to destroy them in 5 years time.
Zatetics on
At one point this was the outlook on ev’s as well. Short term bottlenecks are not a long term problem for China.
They also hold massive unmined repositories of neodymium, dysprosium, praseodymium, yttrium etc. all pretty crucial for AI data warehouses.
And if the US pivots to nuclear power for data warehouses, China has a surplus of gadolinium as well.
The US needs rare earth minerals to continue at the pace they are on, and the guy in charge keeps putting major tariffs on the countries that can supply them. Of course China will catch up and surpass the US.
GuitarGeezer on
Yes, but what is the real prize for winning the race to overinvest in a wildly unprofitable venture? What if even success at the ai ‘race to skynet’ just leaves the US with a giant national and corporate debt and no product commiserate with the investment to have justified it? Nobody seems to be asking this. China might wanna just let the west ‘win’ this pyrrhic victory.
12 Comments
Justin Lin, the head of Alibaba’s Qwen team, estimated the probability of Chinese companies surpassing leading players like OpenAI and Anthropic through fundamental breakthroughs within the next 3 to 5 years to be less than 20%.
Even for their own services – i.e., inference – they’re consuming so much capacity that they don’t have enough compute left to devote to research.
“After a year of gung-ho news about China’s gains in artificial intelligence, some elite Chinese AI researchers are coming to a more pessimistic conclusion. The country’s chances of catching up to the U.S. are slim in the short run, they say, because of a bottleneck in chips.
“The truth may be that the gap is actually widening,” Tang Jie, founder of the Chinese AI startup Zhipu, said at a conference last weekend in Beijing.
Obviously. China is all propaganda… it’s a house of cards and nothing more
They’ll just pour money into it and get there faster than anyone predicts again.
That’s why China wants to invade Taiwan next year so that it will have better chips by grabbing TSMC.
Try Fritos. They are the working man chips. No flash, no big marketing, no fancy shapes, flavors are plain and chili cheese.
I’m actually super excited for China to invest in chip technology that can break nvidias stranglehold
Deep Seek just literally released a new version with a one of a kind scaling architecture. Soooooo dunno man… looks like improvements being made.
There is an upper ceiling in what AI can practically do and accomplish and in what ways it can _actually_ benefit business or humanity.
Are Chinese tech leaders doomposting to get more funding and support?
US AI folks are all VC motivated and have been culturing an insane hype bubble for years. If China buys that hype bubble, they would ofc feel behind, but I’m not sure how true that is.
China seems to leverage AI in a way that’s more responsible and meaningful, even if it isn’t as scaled or appealing in an investment sense.
Seems like Chinese tech sector is using the hype bubble as an excuse to get more share of gov funding.
Has China considered being nice to Taiwan, respecting their sovereignty, and asking politely for some of their chips?
This is EXACTLY the sort of shit I would also be saying if I was trying to make my enemy think I was useless so they might as well sell me those enabling chips I need to destroy them in 5 years time.
At one point this was the outlook on ev’s as well. Short term bottlenecks are not a long term problem for China.
They also hold massive unmined repositories of neodymium, dysprosium, praseodymium, yttrium etc. all pretty crucial for AI data warehouses.
And if the US pivots to nuclear power for data warehouses, China has a surplus of gadolinium as well.
The US needs rare earth minerals to continue at the pace they are on, and the guy in charge keeps putting major tariffs on the countries that can supply them. Of course China will catch up and surpass the US.
Yes, but what is the real prize for winning the race to overinvest in a wildly unprofitable venture? What if even success at the ai ‘race to skynet’ just leaves the US with a giant national and corporate debt and no product commiserate with the investment to have justified it? Nobody seems to be asking this. China might wanna just let the west ‘win’ this pyrrhic victory.