AI summary by TwinMind
A Magic Square Breakthrough - Numberphile
Numberphile
17m 13s
Current State of 3x3 Magic Squares of Square Numbers
- Hang Square: 2,2,3,8,1,17,4,3,3,9 - semi-magic square with all rows/columns working, one diagonal fails
- When squared: all rows, columns, one diagonal sum to same number
- Best known solution to date
- Wakeley Square: 1,2,3,4,10,6,1,4,3,6 - semi-magic square
- Works with modular arithmetic (mod 3)
- Smaller numbers than Hang Square
- Parker Square (2016): had repeated entries (24,1s and 29,2s), one diagonal failed
- Classic example of "almost right but not quite"
Historical Magic Squares of Square Numbers
- 8x8 magic square from 1890 by Frénicle
- Double property: magic square before squaring AND after squaring all numbers
- All diagonals work perfectly
- Second 8x8 square from 1891
- 128x128 magic square from 1950s
- Works as magic square for powers 1, 2, AND 3 (cubes)
- Multi-magic square with extraordinary properties
Theoretical Framework and Constraints
- Tony's previous work: algebraic equations create 2D surface in 8-dimensional space
- "Parker surface" (name didn't catch on)
- Constraint analysis for NxN magic squares:
- Variables: N² options (actually N²-1 due to dependencies)
- Constraints: 2N+2 (N rows + N columns + 2 diagonals)
- Larger squares easier: constraints grow linearly, options grow quadratically
- Parker surface has finitely many rational elliptic curves (300+ found)
- Finite number suggests possible zero solutions for 3x3 case
Major Breakthrough: Proof for All Powers
- New paper: "On the Existence of Magic Squares of Powers"
- Proves: for any power D, magic squares of D-th powers exist above certain threshold
- Key results for D=2 (squares):
- Even magic squares: guaranteed to work from 112x112 up
- Odd magic squares: guaranteed to work from 135x135 up
- Combined with other results: covers 50x50 to 140x140
- Final result: ALL magic squares 4x4 and larger have solutions
- Proof stops exactly at 3x3 - doesn't rule out or confirm existence
Thresholds for Other Powers
- Cubes (D=3): definitely work from 191x191 up
- Every possible power has threshold above which all larger squares work
- Non-constructive proof: shows existence without finding actual squares
- Could potentially lower thresholds with more detailed analysis
Prize Money Evolution and New Bounty
- Martin Gardner (1996): $100 prize
- Magic Square website (2010): $1,000 prize
- Pattern: 10x increase every 14 years
- New bounty: $10,000 USD for finding working 3x3 magic square of squares
- Submissions to: ParkerSquarePrize@gmail.com
- Committee verification through Discord community
- Deadline: 10th anniversary of Parker Square (2 years from recording)
