My brain is absolutely hopeless at top-10-like things. But here are a few gems from some of those categories.
Mathematics: John Conway's "On Numbers and Games". (If you've run across the term "surreal numbers", this is the original source. Absolutely gorgeous, but heavy going if you aren't actually a mathematician.) David McKay's "Information theory, inference, and learning algorithms". (What it says on the tin. Don't expect this to be a practical guide to training neural networks or anything like that; it is a mathematics book.)
Philosophy: Derek Parfit's "Reasons and Persons". (Contains some exceptionally clear thinking on thorny questions about personal identity and the like.) Gary Drescher's "Good and Real". (Drescher is primarily a computer-science-and-AI guy rather than a philosopher. He addresses a bunch of hairy philosophical questions from that perspective and I personally like it a lot. I confidently predict that some people will hate it.)
Science: Gnaedig, Honyek & Riley's "200 puzzling physics problems". (Just what it says. Roughly undergraduate-physics level, some easier, some harder.) Steven Vogel's "Life in moving fluids". (Explores, e.g., the very considerable differences in what living in water means for a bacterium and for a whale on account of their different size. Fairly technical; he has other lighter books.)
Computing: Donald Knuth's "TeX: the program". (A single volume of literate code, constituting the whole of his TeX typesetting system. The style is pretty old-fashioned but he's a genius.) Steven Skiena's "The algorithm design manual". (This would not be my recommendation if you want just one algorithms book; that should be Cormen, Rivest, Leiserson & Stein. But it's an excellent complement to a more standard sort of algorithms book.)
Oh wow, thank you for this! I have been fascinated by Information Theory for quite some while now, so I hope David McKay's book is a good introduction to it.
One of my personal favourite in Philosophy is "Finite and Infinite Games" by James P Carse. It had a profound impact on how I look at situations. I hope one of your suggestions does the same :)
Mathematics: John Conway's "On Numbers and Games". (If you've run across the term "surreal numbers", this is the original source. Absolutely gorgeous, but heavy going if you aren't actually a mathematician.) David McKay's "Information theory, inference, and learning algorithms". (What it says on the tin. Don't expect this to be a practical guide to training neural networks or anything like that; it is a mathematics book.)
Philosophy: Derek Parfit's "Reasons and Persons". (Contains some exceptionally clear thinking on thorny questions about personal identity and the like.) Gary Drescher's "Good and Real". (Drescher is primarily a computer-science-and-AI guy rather than a philosopher. He addresses a bunch of hairy philosophical questions from that perspective and I personally like it a lot. I confidently predict that some people will hate it.)
Science: Gnaedig, Honyek & Riley's "200 puzzling physics problems". (Just what it says. Roughly undergraduate-physics level, some easier, some harder.) Steven Vogel's "Life in moving fluids". (Explores, e.g., the very considerable differences in what living in water means for a bacterium and for a whale on account of their different size. Fairly technical; he has other lighter books.)
Computing: Donald Knuth's "TeX: the program". (A single volume of literate code, constituting the whole of his TeX typesetting system. The style is pretty old-fashioned but he's a genius.) Steven Skiena's "The algorithm design manual". (This would not be my recommendation if you want just one algorithms book; that should be Cormen, Rivest, Leiserson & Stein. But it's an excellent complement to a more standard sort of algorithms book.)