Artificial Intelligence thread

[dripblackcoffee]

Is Claude that good??? Asking because I don't have a subscription and haven't tried it.

fable is but it depends on what you use it for

 

[siegecrossbow]

Please, Log in or Register to view URLs content!

I find ironic that the same neocons who say that HSR is a terrible investment and would never pay for itself think data centers are a great investment.

 

[tphuang]

Please, Log in or Register to view URLs content!

I find ironic that the same neocons who say that HSR is a terrible investment and would never pay for itself think data centers are a great investment.

well, they care about anything that's related to America's technological leadership. This thing is going to blow up so hard once the market finally realizes the glut of data center coming online with no customers.

Is Claude that good??? Asking because I don't have a subscription and haven't tried it.

I mean I don't use it or ChatGPT, but it's clearly the leader based on what others what done with it. It's also kind of implied that if several Chinese labs do some degree of distillation of Claude, then they think Opus is the best.

 

[anzha]

I mean I don't use it or ChatGPT, but it's clearly the leader based on what others what done with it. It's also kind of implied that if several Chinese labs do some degree of distillation of Claude, then they think Opus is the best.

Normally here for learning, but I use claude on a regular basis for a project. It is really good. I've built something that would have taken 20 engineers six months or more prior to the "AI" era. Scoping, design, algorithms, tests, documentation, and even compliance. I've been careful and had external fleshy reviewers just to be careful. Claude's been very, very good.

PS. I pay for it.

 

[Wrought]

Is Claude that good??? Asking because I don't have a subscription and haven't tried it.

It's legit really good, but not worth the money unless your company is paying the subscription (like mine lol). It can still make the occasional dumb mistake, but it reliably gets you ~80% of the way there even on complex tasks. Also benefits hugely from MCP connectors to internal documentation, messaging, codebase, etc, which lets the agents go wild.

And speaking of not worth it:

Chinese models, which had less than half the number of tokens used as U.S. models at the start of the year, surpassed U.S. models multiple times after February. Their use exploded in June, reaching 25 trillion tokens in the final week of the month, more than double the number at the end of May and 78% more than U.S. models. DeepSeek held the top spot in market share by company in June, at 19%. Such other Chinese companies as Z.ai and Xiaomi also saw significant growth, while Google's and OpenAI's shares shrank. "With cost effective token usage dominating the summer headlines, we'd expect these cheaper-but-still-powerful [Chinese] open source models to continue to attract more tokens over time," OpenRouter said in a report.

Please, Log in or Register to view URLs content!

 

[iewgnem]

Is Claude that good??? Asking because I don't have a subscription and haven't tried it.

Its good if your baseline for comparison is no AI or if you're given unlimited tokens for free from your company, but if you pay for it your self or you run the company, and you do coding for a living rather than dabble (i.e. exceed subscription rate limits) , Claude is garbage compared to Chinese AI, you literally could only do a tiny fraction of what you can do with GLM or Kimi for the same budget.

IMO people get attached to Claude because Anthropic put a lot more effort into giving it "personality" than Chinese models, but the world would not feel a single iota of difference if Claude disapeared tomorrow

 

[iewgnem]

Normally here for learning, but I use claude on a regular basis for a project. It is really good. I've built something that would have taken 20 engineers six months or more prior to the "AI" era. Scoping, design, algorithms, tests, documentation, and even compliance. I've been careful and had external fleshy reviewers just to be careful. Claude's been very, very good.

PS. I pay for it.

All AI can do that, but you can do far more with GLM or Kimi on the same budget than Claude, and thats not to mention the large number of people who dont know the difference between Sonnet and Opus.

I do enterprise level coding on GLM + Kimi these days and its infinitely better than waiting for Claude to reset rate limits.

At end of the day hard reality of money doesnt really care about your preferences, companies who use Claude will die to companies who use Chinese models, its a caculus even major US corporations now understand.

Put $150 into GLM max tier and use it inplace of Claude for one month and you'll understand