2⁻⁵² prints as 2.2204460492503⋅10−16. That is the spacing of representable numbers just above 1: a double-precision value carries 52 stored mantissa bits, plus one implicit leading bit, so relative changes below about one part in 4.5 × 10¹⁵ cannot be represented at all.
Add it to 1 and the screen answers 1. The value really does change — in binary64, 1 + 2⁻⁵² is 1.0000000000000002 — but this display prints somewhere near fourteen significant digits, so the sixteenth never reaches the screen. Halve the increment once more and it disappears from the arithmetic too: 1 + 2⁻⁵³ rounds back to 1 with nothing left behind. Every floating-point error budget starts from this number — which is why the imaginary residue left by e^(iπ) at 1.2246467991474e−16 is not a surprise but a rounding of exactly this size.
Answer
2.2204460492503⋅10−16
How to read it
The workbench print of 2⁻⁵². Keying 1 + 2⁻⁵² returns 1 — the increment sits at or below the last digit this screen shows.
Source
IEEE 754 binary64 stores 52 explicit fraction bits, so the gap between 1 and the next representable number is 2⁻⁵²
Work it in the calculator
These links open the right calculator with this problem's numbers already typed in. The line above is what the screen should show — check it yourself.