AI summary by TwinMind
Primes and Primitive Sets (an Erdős Conjecture is cracked) - Numberphile
Numberphile
20m 13s
Erdős Primitive Set Conjecture Background
- Paul Erdős conjecture about primitive sets, now proven as full theorem in 2022
- Gauss quote: "Mathematics is the queen of the sciences and number theory is the queen of mathematics"
- Analogy: If primes are diamonds (precious objects), primitive sets are the full collection of crown jewels
- Primitive sets are properties about collections of numbers vs individual numbers
Definition of Primitive Sets
- Set s is primitive if no number in s divides any other number in s
- For different numbers a and b in set: ratios a/b and b/a are never whole numbers
- Example: If 4 and 2 are in set, not primitive because 4 is multiple of 2
- Definition is simple but creates broad class of sets
Examples of Primitive Sets
- All prime numbers form primitive set (infinite)
- Numbers with exactly k prime factors form primitive sets for any k
- k=2: {4, 6, 9, 10, 15...} (2×2, 2×3, 3×3, 2×5, 3×5...)
- Divisors of p×q are {1, p, q, p×q} - none overlap with other two-factor numbers
- General rule: number with k prime factors has divisors with <k factors, multiples with >k factors
Erdős 1935 Result
- Constant c exists such that for any primitive set s: Σ(1/(n log n)) ≤ c
- Series converges for primitive sets, assigns unique "fingerprint" number to each set
- For all primes: Σ(1/(p log p)) = 1.6366 (computed value)
- For all natural numbers ≥2: Σ(1/(n log n)) diverges to infinity
- Primitive sets have finite sums, most sets diverge
The Erdős Primitive Set Conjecture (Now Theorem)
- For any primitive set: Σ(1/(n log n)) ≤ Σ(1/(p log p)) over primes
- Upper bound is 1.6366 - no primitive set can exceed this "fingerprint" value
- Primes maximize this complexity measure among all primitive sets
- Quantitative proof that primes are special within primitive set family
Proof Details (2022)
- Built on Erdős' 1930s ideas that weren't initially seen as relevant
- Probabilistic reinterpretation of the series as size measure
- Identified inefficient steps in original argument and optimized them
- Worked on problem since undergraduate, done "by candlelight" without advisor knowledge
- Proof uploaded to archive, formal journal submission in process
Extended Conjectures and Open Problems
- Composite primitive sets: What's maximum sum without primes?
- Conjecture: Numbers with exactly 2 prime factors maximize among composite primitive sets
- General conjecture: For numbers with ≥k prime factors, maximum achieved by exactly k factors
- Chain of inequalities: primes > 2-factors > 3-factors > ... in fingerprint values
Maximal Primitive Sets
- Sets with exactly k prime factors are maximal (can't add more numbers while staying primitive)
- Adding number with <k factors: divides existing number
- Adding number with >k factors: is multiple of existing number
- These represent special subclass of all primitive sets
Complications in Extended Conjecture
- After k=6, fingerprint numbers start increasing again (not monotonically decreasing)
- Numbers tend to 1 as k→∞, wedged in thin strip
- Breaks Banks-Martin original conjecture
- Modified conjecture: Consider only odd primitive sets to recover nice decomposition
- k=6 represents minimum in the sequence
