Open-source models are making a comeback, and this time it’s foreigners copying the homework.
People used to say Chinese AI only knew how to follow Silicon Valley’s lead. Now the tide has turned.
When foreigners build large models, they’re actually starting to use our open-source models as the yardstick.
It feels like the perennial second-place student in class suddenly became the one writing the exam questions—pretty surreal.
Open source isn’t charity; it’s about fighting for the right to set the narrative.
Closed source plays the black-box game, while open source lays all the cards on the table and lets you copy them. That kind of move really cuts right to the heart.
When developers around the world are helping you iterate, the costs get spread out—so why wouldn’t closed-source giants be nervous?
Technical direction has never been about who shouts the loudest.
To be honest, this time we really hit the mark: instead of competing on parameter counts, compete on ecosystem.
As long as the foundation is solid enough, anyone trying to go around it will eventually circle back and use you as the reference.
Don’t celebrate too early: models are the measuring stick, but applications are the money printer.
If you’re both the referee and the athlete, you won’t make big money by only being the referee.
One last question: how many times has the AI on your phone truly impressed you enough to feel worth it?
#今日财经 #热点财经 #市场观察 #币安广场 #热点速评
People used to say Chinese AI only knew how to follow Silicon Valley’s lead. Now the tide has turned.
When foreigners build large models, they’re actually starting to use our open-source models as the yardstick.
It feels like the perennial second-place student in class suddenly became the one writing the exam questions—pretty surreal.
Open source isn’t charity; it’s about fighting for the right to set the narrative.
Closed source plays the black-box game, while open source lays all the cards on the table and lets you copy them. That kind of move really cuts right to the heart.
When developers around the world are helping you iterate, the costs get spread out—so why wouldn’t closed-source giants be nervous?
Technical direction has never been about who shouts the loudest.
To be honest, this time we really hit the mark: instead of competing on parameter counts, compete on ecosystem.
As long as the foundation is solid enough, anyone trying to go around it will eventually circle back and use you as the reference.
Don’t celebrate too early: models are the measuring stick, but applications are the money printer.
If you’re both the referee and the athlete, you won’t make big money by only being the referee.
One last question: how many times has the AI on your phone truly impressed you enough to feel worth it?
#今日财经 #热点财经 #市场观察 #币安广场 #热点速评
