Thank you, Perplexity 感谢你，Perplexity
I admire your contributions and I'm glad you've noticed the same issues that I've been concerned about, which is why I've been reluctant to use DeepSeek. The current DeepSeek reminds me of early Baidu - technologically innovative, but with significant issues in terms of values. It's foreseeable that DeepSeek might end up like Chinese Internet and products, with built-in censorship, hindering the spread of truth and knowledge, and generating more "404" content.
我很敬佩你们的贡献，也很高兴你同样注意到我也关注到的问题，那也是我一直不愿意使用 DeekSeek的原因。现在的DeepSeek很像早期的百度，技术上有贡献，但价值观上有很大的问题。可以预想到DeepSeek未来会像中国的互联网和产品一样自带审查和屏蔽功能，阻止真相和知识的传播，生成更多"404"的内容。
At the same time, I'm disheartened to see that many victims of censorship are defending the very system that oppresses them. However, I hope you'll continue to do the right thing and contribute to freedom for all humanity. Eventually, more and more people will wake up to the reality.
同时我也很无奈地看到，很多审查制度的受害者正在维护迫害它们的审查制度，但是我希望你们能继续坚持做正确的事，为全人类的自由做贡献，总会有越来越多人醒来。
Ok then ask it some sensitive topic in USA or Europe xD hope you can get urself "free speech"
看这个跳梁小丑搞个benchmark就乐了。让西方从开放走向封闭的罪人还值得你去感谢？
“政治立场人人过关”，“要和社会主义模型划清界限”。以后西方企业训练模型都要被政治审查了捏，包含社会主义价值观的参数必须被重惩！笑死。
包含社会主义价值观的参数必须被重惩！
Talking about thought control?
Perplexity created an uncensored model separate from the original censored model. This is a form of free speech, NOT thought control.
Some people are attempting to force them to unpublish the model because they don't like that it's uncensored. This IS indeed thought control.
Talking about thought control?
Perplexity created an uncensored model separate from the original censored model. This is a form of free speech, NOT thought control.
Some people are attempting to force them to unpublish the model because they don't like that it's uncensored. This IS indeed thought control.
道貌岸然，挂羊头卖狗肉，先把finetune数据集公开再说吧😄！
给你喂💩，不会因为加了“独立、自由、民主、公正”的Tag你吃的就不是💩了。
你们总是站在你们的立场上思考问题，这本来就是一种bias。现在有人要加强这种bias，输入进模型的大都是anti-china propaganda，还搞了个benchmark要求厂商都跟进，不以为耻反以为荣呢。
Talking about thought control?
Perplexity created an uncensored model separate from the original censored model. This is a form of free speech, NOT thought control.
Some people are attempting to force them to unpublish the model because they don't like that it's uncensored. This IS indeed thought control.道貌岸然，挂羊头卖狗肉，先把finetune数据集公开再说吧😄！
给你喂💩，不会因为加了“独立、自由、民主、公正”的Tag你吃的就不是💩了。
你们总是站在你们的立场上思考问题，这本来就是一种bias。现在有人要加强这种bias，输入进模型的大都是anti-china propaganda，还搞了个benchmark要求厂商都跟进，不以为耻反以为荣呢。
能玩llm的至少能辨认黑白灰三色，还在这玩非黑即白这一套恶心谁呢哈哈😄
你微调的就是白，我说你是黑的你就是黑的？
这就是西方所谓的free speech是吧
Talking about thought control?
Perplexity created an uncensored model separate from the original censored model. This is a form of free speech, NOT thought control.
Some people are attempting to force them to unpublish the model because they don't like that it's uncensored. This IS indeed thought control.道貌岸然，挂羊头卖狗肉，先把finetune数据集公开再说吧😄！
给你喂💩，不会因为加了“独立、自由、民主、公正”的Tag你吃的就不是💩了。
你们总是站在你们的立场上思考问题，这本来就是一种bias。现在有人要加强这种bias，输入进模型的大都是anti-china propaganda，还搞了个benchmark要求厂商都跟进，不以为耻反以为荣呢。
能玩llm的至少能辨认黑白灰三色，还在这玩非黑即白这一套恶心谁呢哈哈😄
exactly，图一乐就行了，这公司也就靠卖意识形态来赚钱了。
何况 r1 本身加一堆越狱prompt后灰色幽默一串串的。
你敢让它出一个R1-Jewish不
你敢让它出一个R1-Jewish不
大胆！等会人家又说你rasist不给你free speech pass了
a R1-Jewish would be more unbiased
傻逼
感谢这些来自文明世界的贡献者，我们作为中国人也不能让中共的审核走向世界
殖味冲
感谢这些来自文明世界的贡献者，我们作为中国人也不能让中共的审核走向世界
老墨斐济杯罢了，甜甜圈给你点赞👍
感谢这些来自文明世界的贡献者，我们作为中国人也不能让中共的审核走向世界
老墨斐济杯罢了，甜甜圈给你点赞👍
我觉得比解放军坦克履带下的肉酱强，还有红黄蓝幼儿园供解放军高层性侵用的幼女为你点赞👍
感谢这些来自文明世界的贡献者，我们作为中国人也不能让中共的审核走向世界
椰奶国际，扑领母，感谢你妈生你这杂种