Why Python? An Honest Accounting ⚖️

The case for Python

1. It is the shortest distance between an idea and a working thing

Counting a file's words takes one line in Python and a small ceremony in most compiled languages. That difference compounds. When an experiment costs you five minutes instead of an hour, you run twenty times more experiments, and running more experiments is most of what learning and research actually are.

2. It reads like English, which matters more than it sounds

crew = ["Guybrush", "Elaine", "Otis"]

if "Otis" in crew:
    print("Otis is aboard.")

for member in crew:
    print(f"{member} reports for duty.")

Otis is aboard.
Guybrush reports for duty.
Elaine reports for duty.
Otis reports for duty.

You could read that aloud to a non-programmer and they would broadly follow it. Very few languages can say that. It matters because, as the Zen says, code is read far more than it is written.

3. The batteries are genuinely included

Python ships with a standard library that handles dates, files, zip archives, JSON, CSV, SQLite databases, HTTP, email, threading, random numbers, maths, testing, logging and unit conversion, all without installing anything. Then there is PyPI, the public package index, which passed half a million published packages. Whatever you want to do, someone has probably done the boring 80% of it already.

4. It owns science and machine learning outright

This is not marketing, it is where the field actually happens. NumPy and pandas are the standard tools for numerical work. PyTorch is what most AI research is written in. The Event Horizon Telescope team used Python to assemble the first image of a black hole, and LIGO publishes its gravitational wave analysis as Python notebooks. When you learn Python you are learning the language the tools of modern science are written in.

5. It is the glue language

Python is unusually good at telling other programs what to do: shell commands, web APIs, spreadsheets, databases, browsers, your operating system. A huge share of real-world Python is fifty lines that make four other systems talk to each other. Unglamorous, enormously useful, and the fastest route to your first genuinely handy tool.

Who actually runs it

The case against Python

A course that only sells you the upside is an advert. Here is where Python genuinely loses.

It is slow, and the reason is structural

Interpreted, dynamically typed code does far more work per operation than compiled code. A tight numeric loop in pure Python can be tens to hundreds of times slower than the same loop in C or Rust. In practice this rarely matters, because the heavy lifting is done inside libraries that are themselves written in C, Rust or Fortran. NumPy is fast because NumPy is not really Python. But if your hot loop is pure Python, you will feel it.

It lets you make mistakes that a compiler would catch

Python will happily run a program containing a typo in a branch you have not tested yet, then fail at 3am. Static languages catch that class of error before the program starts. Type hints and tools like mypy claw a lot of this back (Lesson 38), but it is opt-in, and opt-in safety is weaker than enforced safety.

Shipping it to other people is awkward

A Rust program compiles to a single file you can email to someone. A Python program is code plus an interpreter plus a set of dependencies, and getting all three onto a stranger's machine has spawned an entire industry of workarounds. It is much better than it was, and Lesson 50 covers the modern answers, but it is still Python's softest spot.

Threads do not do what you expect

For most of its life CPython has had a Global Interpreter Lock, which means ordinary threads do not give you more CPU. There are good workarounds (processes, async, native libraries), Python 3.13 introduced an experimental build without the lock, and Lesson 39 explains the whole situation honestly. But "just add threads" is not the answer here that it is elsewhere.

So when should you not pick Python?

The honest summary

Python is the best first language available, and remains a genuinely excellent tenth language. It optimises for the scarcest resource in any project, which is human attention. You will outgrow it in specific directions, and that is a good outcome: knowing exactly why you are reaching for another tool is what makes you an engineer rather than a fan.

Mark lesson complete +100 XP