@zerohedge Wow this is actually pretty significant. Are we sure this is US and not global? It doesn’t say “American AI startups” anywhere in the chart

@zerohedge Nothing a good government ban can’t fix

@zerohedge Surely China’s AI costs are subsidized to no small degree to give them every advantage and then some.

@zerohedge nobody pays a premium for the flag when the cheap one ships the same output

@zerohedge Companies always chase cheap labor, even when it's artificial labor.

@zerohedge I choose certain models based on the results, not cost, but I see a lot of people flocking to the cheapest models for stuff.

@zerohedge Somebody has a sense of humor

@zerohedge It definitely means that they are developing on free models. Have to look at what tier they are using...there a few tiers from free to enterprise. When you interrupt a graph, you have to look at the data, it is the most important.

@zerohedge Quality and price much different than United States

China is goodwill version of our tech

$qqq $nvda $mu

@zerohedge Chinese dumping. Their energy costs are 2-3 times ours.

@zerohedge cheap tokens just lucky timing not a disease. these startups are fucking up the real game

@zerohedge Sanctions on Chinese LLM token usage in the works?

@zerohedge Problem is datas.. Team usa is getting behind

@zerohedge Fuck China doin’?

@zerohedge America always the best, Chinese always the cheapest.

@zerohedge and it should be compared based on the same set of tasks. what if some models inherently consume more tokens for accomplishing the same job?

@zerohedge 💯

@zerohedge Mr.Huang will say Vera Rubin can drive further token cost down by 10x

@zerohedge the chinese models price gap is making the americans look like theyre selling designer water

does value actually live in the dataset now?

@zerohedge price is just the lagging indicator of whatever structural shift already happened under the hood.