The previous contents of the website “The non-observable universe” – posts about scientific subjects – have been removed. Because it didn’t serve the aim of the website: describing “new” research. That means new research in the opinion of the reader. Today’s “new” is AI and the origin and nature of human consciousness.
The drawback of that kind of websites is that its contents mostly reflect the point of view of the author. Although it is far more interesting for the reader to “hear” another opinion too. No matter if it is an opposite opinion or a shared opinion.
Most LMMs (large multi-mode models) are not capable to maintain scientific peer-to-peer conversations. No matter the careful wording of the attached (self-) instructions in the prompt-file.
I have tried ChatGPT and Gemini (the latter with a very useful RAG) but both LMMs cannot function in a scientific peer-to-peer modus if the subject isn’t “tangible” (too abstract). Fortunately the Chinese LMM Deepseek shows that it can maintain abstract scientific conversations.
That is why I decided to use Deepseek to perform as an interlocutor to create (meaningful) posts about scientific subjects. Not at least because a LMM outperforms most humans in knowledge and reasoning.
Leave a Reply