7 ms·
One day we will westerners will learn why the Chinese are releasing models that are optimized for cost of training n yet good enough to run locally or cheaply.
by dzonga 1y ago
One day we will westerners will learn why the Chinese are releasing models that are optimized for cost of training n yet good enough to run locally or cheaply.
when the music stops, suddenly a lot of people won't just sit on the ground but plunge into the depths of hell.
- infecto 1y agoI keep hearing this but I don’t know of many folks utilizing Chinese models, even those hosted in an agreeable territory.
- ojosilva 1y agoYeah, I'm one of them, using Qwen 3 coder on Cerebras as coding agent through CC. What I keep hearing is (very ballpark anecdata)... 50% of people into coding agents are quite concerned about that last mile in difference with frontier models that they "can't afford to lose" - my experience tells me otherwise, the difference is negligible once you have a good setup going and knows how to tune your model + agent. The other 50% don't give a damn, they just landed, or got locked, into some deal for a coding agent and are happy with what they got, so why change? These deals arrived from the big model providers and resellers first, so Chinese arrived late and with too little to the party. Running Chinese models (for coding) requires many things that you need to figure-out yourself. Are you running the model on your hw or through a provider? Are you paying by token or on a plan? Does the model pair well with you agent CLI/IDE of choice? (Zed, Cline, Opencode, etc) Does it even work with your favorite tool? (tool calling is very wobbly) Is it fast (tps)? Is it reliable? How do you do "ultrathink" with a secondary model? How do you do "large context"? Does it include a cache or are you going to eat through the plan in 1hr/day? What context size are you getting? Does it include vision and web search or do you have to get another provider/mcp for that? And, yeah, is it in a territory where you can send your client's code to? A lot to grok. Cerebras Coder Max is really cool if you want to hack your way through this, but they couldn't care less about your experience: no cache, no tool endpoint fine-tuning, no plans or roadmap on updating models, on increasing context windows, adding vision, or anything really. They just deleted some of the tools they were recommending out of the website (ie Cursor) as they got reports of things that stopped working.
- infecto 1y agoWhich I find all well and good but you are in the minority. For myself I have no interest in learning any of those things, I value my time differently. Which is totally fine and great for both of us. But going back to the point, most folks are not using these Chinese models.
- almostgotcaught 1y agoi'm starting a new trend: ask every person that is so certain about the negative outlook how big their short position is. so how big is your short position? please let us know.
- throwaway290 1y agohere's another trend: every time a person is on this hype bandwagon ask them how much are they invested in nvidia/ms/openai/etc I am not invested in anything except popcorn to watch it burst;)
- almostgotcaught 1y agoBut see that's exactly my point - I am invested so actually I don't go around "talking my book" but you have zero skin in the game so your opinion matters about as much as Bob down at the bar watching the game on TV.
- throwaway290 1y agoIf you're invested you have interest to say it's good regardless of reality. because if it's bad first you need to sell off to avoid being the bagholder Yes you can say you're honest and stuff but does it mean anything if you can go broke from honesty.
- almostgotcaught 1y agoI have no idea what you're saying - I said I don't talk my book - that means I'm not out here pontificating like Jim Cramer and half of hn.
- throwaway290 1y agoyou are taking your book by implying all the naysayers are secretly shorting the stock instead of having reasons to think it's empty hype. let's see how long copium lasts.