Re:GPT 4.5 level for 1% of the price
by buyucu on Sunday March 16, 2025 @08:36AM
(#43378547)
This is also true of OpenAI and Anthropic. There is a disturbing level of censorship there.
[ Parent
| on HN ]
Re:GPT 4.5 level for 1% of the price
by mirekrusin on Sunday March 16, 2025 @08:41AM
(#43378565)
The funny thing is that you'll hit the wall sooner or later with reasoning if you don't give up bias, which applies on both/any ends.
You can't "give up bias". You can choose what you prefer, but an "unbiased" model doesn't exist. If you put in information, you're adding bias.
what people mean by an "unbiased" model is a model that just reflects the underlying distribution in the data.
So biased by both the choice of the sources and the information contained in the sources.
But crucially not biased by RLHF being applied afterwards to stop the models from making basic biological observations, telling jokes about left wing politics, or saying anything nonnegative about European historical impact on the world, and so on.
Re:GPT 4.5 level for 1% of the price
by blackeyeblitzar on Monday March 17, 2025 @03:34AM
(#43386022)
There isn’t a “disturbing level of censorship” in either OpenAI or Anthropic. This false equivalency is a propaganda claim. DeepSeek has obvious extreme censorship aligned to the CCP agenda and it is well documented. There’s nothing like it in American or European AI. I’m sure the AI from Baidu or Bytedance or whoever will show something similar.