Cixin Liu's 1 book The Three-Body Problem is described widely as hard science-fiction.
This is Wikipedia's (sourced) definition: "Hard science fiction is a category of science fiction characterized by an emphasis on scientific accuracy, technical detail, or both."
With this definition, I approached the science in the book with a degree of trust. A trust that has been violated and left me feeling very, very stupid. I don't think this is Cixin Liu's fault. He didn't classify his own book, I assume.
The problem is that there is a lot of science in this book as I mentioned earlier - some of it true and based on the current understandings of theoretical physics, some of it is a colourful extrapolation that can't be justified based on what we already know. My biggest gripe is with a scene containing an awe-inspiring description of higher-dimension engineering.
Minor spoilers ahead
A proton is 'unpacked' from eleven dimensions into two dimensions and electronic circuits are etched on its near mass-less two-dimensional body turning it into a computer possessing artificial super-intelligence. In my (simple) mind I thought that this was some sort of far-future extrapolation of the current science. This is true of artificial intelligence and probably of the unpacking of the proton. But I can't find a single mention of the idea of turning sub-atomic particles into computers, of the higher-dimensional engineering. Cixin Liu seems to have made that up - which is fine everywhere except hard sci-fi.
Liu's book seems to be hearkening back to Golden Age science fiction where the main objective was not pedagogical or didactic but rather to inspire, to turn minds onto the unfathomable wonders of the scientific project. Which is great. I'm just annoyed and embarrassed because Mukunth laughed for a long time when I asked him about proton supercomputers. Goddamn book-classifiers, you had one job!
How to pronounce the author's name: https://forvo.com/word/%E5%88%98%E6%85%88%E6%AC%A3 . My best approximation: Liou Tsuhshin. In Chinese, surnames precede first names. ↩