This is going to have an opposite effect that the decels want. It's a huge open source AI accelerant, an accelerant for corporations, and enterprises
Now it's a real race. You either make your own AI infrastructure or you don't have a seat at the table
Many users agree the open source AI surge accelerates progress and infrastructure demands, while others fear it will create elite-only access to top models or trigger export bans.
It's so strange talking about this stuff for such a long time and seeing it play out. A little derealizing
This is going to have an opposite effect that the decels want. It's a huge open source AI accelerant, an accelerant for corporations, and enterprises
Now it's a real race. You either make your own AI infrastructure or you don't have a seat at the table
welcome to tensorpunk
This is going to have an opposite effect that the decels want. It's a huge open source AI accelerant, an accelerant for corporations, and enterprises
Now it's a real race. You either make your own AI infrastructure or you don't have a seat at the table
@yacineMTB Yes man. Even I will dedicate my entire career now to building open source AI. Am quitting my job on Monday. Fuck this
@yacineMTB All they have to do is knock out GPT-5.5 too and my local Kimi-K2.7 reigns supreme!
@BigAl_PHD @yacineMTB You are welcome to your opinion. Local hardware will never be as efficient for cost or performance. Shared hardware creates big gains in efficiency.
@BigAl_PHD @yacineMTB Local models don't have enough compute. The actual answer is in between, with rented hardware in datacenters.
@yacineMTB Open source models running locally has been the obvious path since gpt 4. Even before open source models were good it was clear. It’s actually kind of crazy how easy it is to predict how things shake out
@10_X_eng @yacineMTB I don't think you need a data center
@yacineMTB permanent underclass acceleration intensifies
@yacineMTB i'm already pulling the chinese weights local
@yacineMTB i think this is real
@matchaboar @yacineMTB Even with the resources big labs have currently it’s the ultimate example of shared hardware economic efficiencies and they lose their ass on compute
@yacineMTB should every startup now just raise to setup their own data center and use open source Chinese models cause of this 😂
@yacineMTB Thats right. Pretty big wakeup call to everyone.
@yacineMTB No, the problem is you cant do export controls on open source, so guess what? they will ban it.
@yacineMTB There's an issue with the selected model (claude-fable-5). It may not exist or you may not have access to it. Run /model to pick a different model.
@yacineMTB Are you raising to build a data center now?
@yacineMTB This is the premise from which I have worked from for the past 3 years.
@yacineMTB @yacineMTB very rich coming from a guy who literally didn’t build one. What would you do? Did you build one?
@yacineMTB This validated the middle power ‘CERN’ frontier lab idea from the wonk crowd.