Take any two positive numbers, say 4 and 16. Add them and halve it — you get their arithmetic mean: (4+16)/2=10. Multiply them and take the square root — you get their geometric mean: 4×16=8. Notice something? 10≥8. Try it with any other pair of positive numbers you like — the arithmetic mean is never smaller than the geometric mean. That simple, always-true observation is the Inequality of Means, usually written AM ≥ GM.
The precise statement
For two positive real numbers a and b:
AM=2a+b,GM=ab
2a+b≥ab
with equality if and only if a=b. If a=b, the inequality is strict.
Why it is always true
Start from a fact that can never fail: the square of any real number is non-negative.
(a−b)2≥0
Expand the left side:
a−2ab+b≥0
a+b≥2ab
Divide both sides by 2:
2a+b≥ab
That's the whole proof — no assumptions beyond a,b>0 (so that a,b are real numbers). Since (a−b)2=0 exactly when a=b, equality holds exactly when a=b.
Note
The inequality needs a,b≥0. For negative numbers, ab may not even be real, so the "GM" isn't defined there.
Worked example
Find the AM and GM of 9 and 25, and verify the inequality.
Step 1:AM=29+25=17
Step 2:GM=9×25=225=15
Step 3: Check: 17≥15✓ — and since 9=25, the inequality is strict, exactly as the rule predicts.
A useful consequence: inserting a mean between two numbers
If a and b are two positive numbers and G is inserted between them so that a,G,b form a Geometric Progression, then G=ab — precisely the geometric mean. Comparing this G against the arithmetic mean A=2a+b (the number that would sit between a and b in an Arithmetic Progression) is exactly an application of this inequality: A≥G always, so the AM-inserted term never sits below the GM-inserted term.
Watch out
A common slip is writing ab when a or b is negative, or applying the two-number formula directly to more than two numbers. For n positive numbers a1,a2,…,an, the generalised inequality is