Hacker News Reader

Is it really that interesting to point out for every Chinese oss model release?

Yes. The original DeepSeek-R1 answered those questions just fine. The newer models seem to be much more brainwashed.

Is it not relevant to reiterate the bias (or lobotomization) for people new to this space?

No, it’s not really relevant. Should I point out that all the models from providers in the west are very “left-leaning” every time one is released? Is that helpful to the technical discussion, in any way?

If you are using an LLM for historical knowledge, questions, or research, then the chinese censorship is relevant. Or for questions about geopolitics.

If you had a specific example where the LLMs showed “left leaning” bias, then yes, it would be interesting to me.

Like i said i normally don’t point this out, but because you asked, here you go:

https://www.gsb.stanford.edu/insights/popular-ai-models-show...

> On average, models from Google and DeepSeek were seen as statistically indistinguishable from neutral, while models from Elon Musk’s xAI, which touts its commitment to unbiased output, were perceived as exhibiting the second-highest degree of left-leaning slant among both Democratic and Republican respondents.

This is somewhat interesting but this is hardly in line with censorship of historical events.

Reply on news.ycombinator.com