promptdojo_

Read the names in a tiny tool — step 3 of 8

ch 01 · first code-reading lab3/8

= doesn't mean what it did in math class

This one trips up almost every non-coder exactly once, and then never again. It's the highest-leverage three minutes you'll spend in this whole chapter.

In math, = is a question

x = 7 in algebra is a claim that two things are equal. You can rearrange it (7 = x), you can chain it (x = y = 7), and the equals sign points both directions. It's a mirror.

In Python, = is a command

user_score = 7 in Python is do this thing: take the value on the right, stick the label user_score on it. The arrow only points one way:

user_score  ←  7

You can't write 7 = user_score in Python — it's a syntax error. Python isn't asking whether they're equal. It's telling Python put that label on that value, now.

The rebind move (the part that confuses everyone)

Here's where AI-generated code looks contradictory if you carry the math model. Look at the editor on the right:

user_score = 7
print(user_score)

user_score = 12
print(user_score)

To a math brain, line 4 looks like a contradiction — but you said user_score was 7? Python doesn't see a contradiction. It sees four commands, in order:

  1. Stick the label user_score on the value 7.
  2. Print whatever has that label. → 7
  3. Now stick the label user_score on the value 12 instead. (The 7 is no longer labeled by anything.)
  4. Print whatever has that label now. → 12

Run the code and watch the printed value change. The variable is a label you can move, not a fact you're declaring.

Where AI specifically breaks this for you

The single most common pattern AI tools write is the accumulator loop:

total = 0
for item in items:
    total = total + item.price

total gets rebound on every iteration. The math brain reads total = total + item.price as a contradiction — total equals total plus something? impossible. The Python brain reads it as: grab the current value labeled total, add item.price to it, move the label to the new sum.

If you understand this one move, you can read most loops an AI tool will ever write you. That's not an exaggeration — accumulators show up constantly in scoring, totaling, summarizing, list-building, and reduce operations.

One more thing: variables don't lock the type

Python lets the same label sit on a number, then a string, then a list, across the life of the program:

result = 0          # result is now an int
result = "ok"       # now it's a string
result = [1, 2, 3]  # now a list

This is called dynamic typing, and AI uses it more than you'd expect. When a variable's type seems to change between lines, that's not a bug — that's Python letting the label move to a totally different kind of value. Sometimes it IS a bug, and the next chapter will teach you how to spot the difference. For now, just know: the label has no opinion about what it's stuck on.

read, then continue.