DeepSeek, China's AI model: News & Discussion

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.
 
I feel like Deepseek gotten dumber. Sure you can now have threads of almost unlimited length whereas the chat length was capped before. But I asked Deepseek the same questions as last year and now it replies it doesn't know what I'm talking about or there is no info on this topic.

What I still don't understand is, the knowledge base of Deepseek seems to have surpassed Google search engine. It seems Deepseek knows anything about everything. How is this possible when Deepseek is only a small team of (software) engineers and an offshoot of some hedge fund? Where does the knowledge base of Deepseek actually come from?
I don't mean Deepseek gotten "dumb" dumb but rather "ignorant" dumb.

Last year I went thru pages and pages to discuss a particular topic. This year, I asked Deepseek questions on the same topic. It replies there is no info on the topic I discussed in depth last year with Deepseek. Like what the heck?!?!
 
To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.
 
The most impressive fact about Zhipu's GLM-5.2 is that it was trained entirely on Huawei Ascend chips
To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.
 
Chinese AI researchers are the backbone of Chinese and American AI success.
 

Users who are viewing this thread

Back
Top