This is a really cool idea! I recently came across another project on GitHub: https://github.com/tensorzero/tensorzero that explores a similar direction. You might find it interesting, and perhaps it could offer some inspiration or useful insights for your work as well.
cool! Even though it’s getting easier to use AI to generate commands these days—I often ask AI to help me write commands myself—I still think it’s important to understand and systematically learn the command line. I’ll definitely recommend this book to my friends. One suggestion: maybe you could include some practical examples from real work scenarios in the book to make it even more approachable and engaging.
I don't want to just add to it because the entire idea is to have a short book, but I will be thinking about replacing/improving some sections. The examples already demonstrate the practical scenarios, but I think there could be a few sentences about how a particular concept can be used in real world scenario.
To be honest, I’ve done quite a bit of LeetCode, and while it’s great for practicing algorithms and problem-solving, I’ve found that I rarely use these “fancy” algorithms in my actual work. Most of the time, what really matters is breaking down requirements into clear logic, writing maintainable and collaborative code, and communicating well with the team. These skills are hard to assess through algorithm-focused interviews alone. I really hope interviews can place more emphasis on practical engineering skills and real-world scenarios, which would be more valuable for both companies and candidates.
I’m currently preparing for interviews myself, so having access to high-quality, free resources like this is incredibly helpful. The AI interviewer feature, in particular, looks like it will be very useful for me. Thanks again to the author for making these resources available!
I think it really depends on the job. In most jobs I've had the algorithms rarely get more complicated than just avoiding accidental O(N^2). The sort of thing where juniors would do a linear search.
But I did have one job working on an AI graph compiler which used fancy algorithms all over the place. In practice though I found the space between "use the standard library" and "it's NP-complete; use heuristics" where the answer is "you can use this neat dynamic programming trick" is basically nonexistent.
