6 ms·

I got their z.ai plan to test alongside my Claude subscription; it feels about on par with something between sonnet 4.0 and sonnet 4.5. It's definitely a few st

I got their z.ai plan to test alongside my Claude subscription; it feels about on par with something between sonnet 4.0 and sonnet 4.5. It's definitely a few steps below current day Claude, but it's very capable.

enraged_camel 9mo ago
When you say "current day Claude" you need to distinguish between the models. Because Opus 4.5 is significantly ahead of Sonnet 4.5.
- kachapopopow 9mo ago
  opus 4.5 is truly like magic, completely different type of intellience - not sure.
  - hhh 9mo ago
    most of my experience with 4.5 is similar to codex 5.1, where I just have to scold it for being dumb and doing things I would have done as a teenager
    - kachapopopow 9mo ago
      dumbness usually comes from lack of information, humans are the same way - the difference between other llms is that if opus has information it has a ridiculously high accuracy on tasks.
  - croes 9mo ago
    Magic when it works.
- hatefulmoron 9mo ago
  Yeah, when I say "current day Claude" I'm referring to Opus 4.5, which is what I always use on the max plan.
jijji 9mo ago
z.ai (Zhipu AI) is a chinese run entity, so presumably China's National Intelligence Law put in place in 2018, which requires data exfiltration back to the government, would apply to the use of this. I wouldn't feel comfortable using any service that has that fundamental requirement.
- queenkjuul 9mo ago
  If the Chinese government has the data at least the US government can't grab it and use it in court. Not living in China I'm not too concerned about the Chinese government
- deaux 9mo ago
  Google, OpenAI, Anthropic and Y Combinator are US run entities, so presumably the CLOUD Act and FISA require data exfiltration back to the government when asked, on top of the all the "Room 641A"s where the NSA directly taps into the ISP interconnects, would apply to the use of them. I wouldn't feel comfortable using any service that has that fundamental requirement.
- hatefulmoron 9mo ago
  I wouldn't use any provider: z.ai, Claude, OpenAI, ... if I was concerned about the government obtaining my prompts. If you're doing something where this is a legitimate concern (as opposed to my open source stuff), you should get a local LLM or put a lot of effort into anonymizing yourself and your prompts.

enraged_camel 9mo ago

When you say "current day Claude" you need to distinguish between the models. Because Opus 4.5 is significantly ahead of Sonnet 4.5.

kachapopopow 9mo ago
opus 4.5 is truly like magic, completely different type of intellience - not sure.
- hhh 9mo ago
  most of my experience with 4.5 is similar to codex 5.1, where I just have to scold it for being dumb and doing things I would have done as a teenager
  - kachapopopow 9mo ago
    dumbness usually comes from lack of information, humans are the same way - the difference between other llms is that if opus has information it has a ridiculously high accuracy on tasks.
- croes 9mo ago
  Magic when it works.
hatefulmoron 9mo ago
Yeah, when I say "current day Claude" I'm referring to Opus 4.5, which is what I always use on the max plan.

jijji 9mo ago

z.ai (Zhipu AI) is a chinese run entity, so presumably China's National Intelligence Law put in place in 2018, which requires data exfiltration back to the government, would apply to the use of this. I wouldn't feel comfortable using any service that has that fundamental requirement.

queenkjuul 9mo ago
If the Chinese government has the data at least the US government can't grab it and use it in court. Not living in China I'm not too concerned about the Chinese government
deaux 9mo ago
Google, OpenAI, Anthropic and Y Combinator are US run entities, so presumably the CLOUD Act and FISA require data exfiltration back to the government when asked, on top of the all the "Room 641A"s where the NSA directly taps into the ISP interconnects, would apply to the use of them. I wouldn't feel comfortable using any service that has that fundamental requirement.
hatefulmoron 9mo ago
I wouldn't use any provider: z.ai, Claude, OpenAI, ... if I was concerned about the government obtaining my prompts. If you're doing something where this is a legitimate concern (as opposed to my open source stuff), you should get a local LLM or put a lot of effort into anonymizing yourself and your prompts.