Artificial Intelligence thread
- Thread starter 9dashline
- Start date
Last edited:
I mean in terms of product, they compete better against GLM and Minimax.
Doubao is on a different level than them.
When I used Qwen 3 web version a few times, the results were always garbage. Which I never had with Kimi 2.5. K2.5 is always excellent. It's puzzling why Qwen-3 large models suck so much.
very interesting, now we have both ByteDance/Doubao and Zai/GLM team having trouble meeting the surging inference demand ahead of Chinese New Year period. The former is limiting audio part of video generation. The latter is looking everywhere for domestic compute options.
This seems like the first time in a while the Chinese AI suppliers are having trouble with inference demand.
Minimax is creating a premium high speed plan to meet the customers willing to pay extra for its API. Minimax coding remains the one that requires the least compute for fast response.
This seems like the first time in a while the Chinese AI suppliers are having trouble with inference demand.
Minimax is creating a premium high speed plan to meet the customers willing to pay extra for its API. Minimax coding remains the one that requires the least compute for fast response.
quantized 8 version of Minimax 2.5 is just 226GB and I think it under 500GB for w4a8 version of GLM-5. So the big question is just who is getting automated away? I don't think they are hurting Claude revenue just yet. But they are clearly being used by some people out there. My personal usage is pretty limited. I use web for GLM-5 and it generates some tokens. But I only use it a few times a week and it already saves me hours of time.
The AIs are more efficient in a few minutes than days of Infosys workers.
Sometime in the past six months Alibaba was outpaced by Minimax, Zhipu and Moonshot. Qwen is tier 2 at this point.
Answer: Those suppressed demand that couldn't afford Claude and tasks couldn't be completed by previous models.
I remember reading a Weibo post from an AI data provider a few months ago. He said that while a few dozen yuan might seem insignificant to a company for completing a small task, software companies actually want to complete it for just a few yuan or even a few cents.
Especially now that agents allow developers to complete complex tasks using a large number of tokens, cost-effectiveness has become extremely important.
one thing that people are not realizing is that Chinese ai firms are currently effectively exporting digital services to other countries via these token generation.
My point is that now that agents allow developers to complete complex tasks, then that means you need fewer human software devs. So who is losing jobs?
I know for my firm, I am not hiring anyone right now because myself + AI is more productive than myself with several mid tier devs because I can explain what needs to be done to AI better than another dev & AI can get the code done better than mid tier devs.
it frankly doesn’t matter that opus is better than GLM because GLM can answer all the tasks I need it to do on 1 pass. So when all the LLMs can do the task, I will pick the lowest cost one.