User:Eliora/Sandbox: Difference between revisions
No edit summary |
|||
| Line 58: | Line 58: | ||
and reported a proof that the resulting constant is {{W|transcendental}}. The argument was described as parity-agnostic: rather than depending on the known structure of even perfect numbers or the newly established properties of odd perfect numbers separately, it treated perfectness itself, <math>\sigma \left(n\right)=2n</math>, as the defining arithmetic property of the sequence. Using PSLQ- and LLL-based computational techniques within the ICS framework, Seng worked from the assumption that the reciprocal sum was algebraic of arbitrary finite degree and derived constraints on its computable structure incompatible with the arithmetic information encoded by the constant. The result was presented as an ontological application of ICS, in the sense that the argument depended on what integers qualify as perfect numbers rather than on their parity or a particular parametrization of either class. Consequently, if the proof is correct, transcendence of the full reciprocal sum also establishes that the set of perfect numbers cannot be finite, and hence that at least one of the even or odd classes must contain infinitely many members, without determining which. The proof is currently awaiting independent review. | and reported a proof that the resulting constant is {{W|transcendental}}. The argument was described as parity-agnostic: rather than depending on the known structure of even perfect numbers or the newly established properties of odd perfect numbers separately, it treated perfectness itself, <math>\sigma \left(n\right)=2n</math>, as the defining arithmetic property of the sequence. Using PSLQ- and LLL-based computational techniques within the ICS framework, Seng worked from the assumption that the reciprocal sum was algebraic of arbitrary finite degree and derived constraints on its computable structure incompatible with the arithmetic information encoded by the constant. The result was presented as an ontological application of ICS, in the sense that the argument depended on what integers qualify as perfect numbers rather than on their parity or a particular parametrization of either class. Consequently, if the proof is correct, transcendence of the full reciprocal sum also establishes that the set of perfect numbers cannot be finite, and hence that at least one of the even or odd classes must contain infinitely many members, without determining which. The proof is currently awaiting independent review. | ||
== First elite number == | |||
The first elite number, <math>G=2.368259249829\ldots \times {10}^{136,248,339,111}</math>, is the first discovered counterexample to the Goldbach conjecture and therefore the first known even integer greater than 2 that cannot be represented as the sum of two odd prime numbers. Its discovery disproved the conjecture. | |||
<math>G</math> is a colossally abundant number, and hence by definition – a superabundant number. Based on the estimated distributions of superabundant and colossally abundant numbers, <math>G</math> is predicted to be approximately the 824 billionth superabundant number and the 26 millionth colossally abundant number. Its exact ordinal positions in either sequence are unknown, as <math>G</math> was identified directly rather than through an exhaustive enumeration of all preceding superabundant and colossally abundant numbers. It also holds the record for the greatest known distance between a superabundant number and its nearest prime (typically, such numbers are only 1 away from nearest prime, like 5040 and 5039). The nearest primes lie at | |||
:<math>G-31131211131221</math> and <math>G+31131211131221</math>, | |||
It remains unknown whether <math>G</math> is the smallest elite number. Heuristic models instead predict that the majority of elite numbers, if there is infinitely many of them, have comparatively sparse factorizations, typically consisting of a small power of 2 multiplied by several very large distinct prime factors, | |||
:<math>E=2^{a}p_{1}p_{2}\cdots p_{k}</math>, | |||
where <math>a</math> is relatively small and the <math>p_{i}</math> occur predominantly to the first power. However, such candidates are, given predicted sizes of elite numbers – are impossible to identify using existing methods. Hence, the method responsible for its discovery is primarily capable of identifying elite numbers not amongst that set – but rather only among superabundant or similarly highly divisible integers exhibiting anomalously large distances from nearby primes, and consequently samples only a restricted subset of possible elite numbers. The search therefore concentrated principally on the numbers whose highly structured factorizations permitted much more efficient screening. | |||
=== Search and discovery === | |||
The term elite number was coined by Mehmet Aytaç, Matthew Powell and Antonio Kʼawiil at University College Dublin. It originated in a recurring joke in their early work that anthropomorphized highly divisible composite numbers as an “aristocracy” and primes as “peasants”. Kʼawiil, a Maya mathematician from National Autonomous University of Mexico, extended the joke through a parodic form of Maya numerology, playing on popular fascination with the Maya calendar and supposed Maya mathematical mysticism: highly composite numbers were described as worthy of ceremonial use in the Long Count, while indivisible primes were “unworthy of worship”. Under the analogy, the Goldbach conjecture forced every even integer to nevertheless “associate” with prime numbers. A hypothetical counterexample was consequently termed an elite number — an even integer that did not have to “associate with the peasants”. | |||
On 24 October 2026, Aytaç, Powell and Kʼawiil at University College Dublin, together with researchers at the National Autonomous University of Mexico and seven universities in California, began co-chairing an international computational programme devoted to the search for elite numbers. The programme marked a departure from earlier approaches to the Goldbach conjecture by inquiring into the existence of elite numbers, rather than attempting to prove the conjecture in the positive formulation, as its principal research question. | |||
The reformulation changed the direction of the search. Rather than attempting to verify the Goldbach conjecture for progressively larger ranges of even integers, the programme sought to determine which classes of integers could contain elite numbers and to eliminate classes for which their existence could be excluded. The researchers described this approach as an inversion of the conventional problem: instead of asking whether every even integer possessed a Goldbach representation, they asked directly whether any even integer possessed none. | |||
The computational programme eventually incorporated 6,769 supercomputers and artificial-intelligence data centres, with substantial portions of the computation devoted to the factorization, divisor structure and Goldbach representation counts of selected families of integers. Early heuristic models developed in late 2026 predicted that an elite number was most likely to possess a comparatively sparse prime factorization, consisting of a relatively small power of 2 and a small number of very large odd prime factors, with a wide separation between their logarithmic sizes. The models additionally suggested an upper bound on the exponent of 2 occurring in such a factorization. Numbers having these characteristics were, however, among the most computationally difficult candidates to investigate directly, particularly when their odd prime factors were extremely large. | |||
The collaboration consequently adopted the initially counterintuitive strategy of examining classes in which elite numbers were considered least likely to occur. These families generally possessed highly regular factorizations and could therefore be studied substantially more efficiently. The purpose of these searches was not primarily to discover an elite number within them, but to establish increasingly broad exclusion results, progressively restricting the possible arithmetic form of any counterexample. | |||
One of the first major investigations concerned perfect powers of even integers. Already before 2026, it was known that if <math>g(n)</math>denotes the number of Goldbach representations of an even integer n, numerical experiments had indicated that, for a fixed even integer n, | |||
<math>\underset{k\rightarrow \infty }{\lim}\frac{g\left(n^{k+1}\right)}{g\left(n^{k}\right)}=n</math>. | |||
For powers of 2, for example, | |||
<math>\underset{k\rightarrow \infty }{\lim}\frac{g\left(2^{k+1}\right)}{g\left(2^{k}\right)}=2</math>. | |||
The relationship had previously been regarded as a heuristic asymptotic property of Goldbach representation counts. In July 2027, researchers associated with the programme proved the limiting relation and strengthened it sufficiently to establish that perfect powers of even integers could not be elite numbers. Hence, this only left the integers of the form <math>2^{a}m^{b}</math>, with <math>m</math>odd and <math>a≠b</math>, and later on more general factorizations <math>2^{a}\prod_{i} p_{i}^{k_{i}}</math>. The programme nevertheless continued to investigate highly structured families considered unfavorable to elite numbers. This strategy ultimately became important to the discovery of <math>G</math>, despite the results that most elite numbers are not similar to <math>G</math> in structure. | |||
== Disproof of the Riemann hypothesis == | |||
Shortly after the discovery of the first elite number, G, was announced in Berkeley, researchers studying its divisor structure determined that it was also the first known integer greater than 5,040 to violate Robin's inequality, | |||
: <math>\sigma \left(n\right)\lt e^{\gamma }n\operatorname{\log{\log}}{n}</math>. | |||
Because Robin's theorem establishes that the inequality holds for every <math>n>5040</math> if and only if the Riemann hypothesis is true, the result constituted an indirect disproof of the hypothesis before any non-trivial zero outside the critical line had been identified. The Berkeley team subsequently contacted Coraline Seng and Luke Bannon, whose group operated Solace, an artificial-intelligence research system developed specifically for investigations of the Riemann hypothesis. | |||
Seng and Bannon later recalled in interviews that they initially “couldn't quite believe what they were seeing”, particularly because the result appeared to resolve the hypothesis through an unexpected divisor-theoretic consequence of <math>G</math>, rather than through the direct discovery of an exceptional zero. Following independent checks of the Robin inequality calculation, Seng and Bannon retuned Solace to search explicitly for zeros outside the critical line, incorporating Gtogether with other exceptional numbers reported earlier that day. Solace autonomously identified the first odd perfect number, <math>N</math>, as a second relevant divisor-theoretic object and tested combinations involving the two newly discovered numbers. At 16:15:34 UTC, it reported | |||
:<math>\boxed{\zeta \left(\frac{1}{N}+Gi\right)=0,}</math> | |||
providing the first explicit non-trivial zero known to lie outside <math>\Re (s)=\frac{1}{2}</math>. While the violation of Robin's inequality had already done so, the aforementioned complex number supplied the direct zeta-function counterexample that researchers had sought immediately afterward. The rapid progression from the discovery of <math>G</math>, through the unexpected Robin violation, to Solace's identification of an explicit zero became one of the most widely reported episodes of the Day of Mathematics. | |||
== Disproof of Collatz conjecture == | |||
Later that day, Kindred, an artificial-intelligence research system operated since 2028 with the objective of resolving the Collatz conjecture, independently processed the factorization data produced following the discovery of <math>N_{2}</math>. The analysis was conducted principally by Kindred servers operated at the Massachusetts Institute of Technology (MIT). As part of its automated search, the system ingested all 10,572,384,227 prime factors identified in the factorization of the second known odd perfect number and examined their Collatz trajectories and associated parity structures. | |||
At 16:40:00 UTC, Kindred identified the second prime factor of <math>N_{2}</math>, | |||
:<math>C\approx 2.158158\times {10}^{5.1284\ldots \times {10}^{383838472427}}</math>, | |||
as a counterexample to the Collatz conjecture. Rather than eventually entering the <math>4 \rightarrow 2 \rightarrow 1</math> cycle, the orbit of <math>C</math> was shown to be unbounded. Kindred subsequently produced a proof that the trajectory continued to grow without bound, establishing Cas the first known '''non-Collatzian number'''. The result was particularly unexpected because <math>C</math> had not been identified through a direct search over integers: it had entered Kindred's analysis only as one of the prime factors of <math>N_{2}</math>, discovered minutes earlier. | |||
=== Second counterexample ot the Riemann hypothesis === | |||
The Kindred result was automatically incorporated into the continuing computations performed by Solace. The system combined <math>C</math> with <math>F_{M}</math>, the newly discovered Fermat prime that had independently been identified as the special prime of <math>N_{2}</math>. At 16:46:24 UTC, six minutes after Kindred's announcement, Solace reported a second non-trivial zero outside the critical line, | |||
:<math>\boxed{\zeta \left(\frac{1}{F_{M}}+Ci\right)=0.}</math> | |||
The result attracted particular attention because both coordinates of the new zero were derived from exceptional prime factors of the same odd perfect number: its real component was determined by <math>F_{M}</math>, while its imaginary component was the newly identified Collatz counterexample <math>C</math>. It followed the first explicit counterexample, | |||
<math>\zeta \left(\frac{1}{N_{1}}+Gi\right)=0<math> reported by Solace at 16:15:34 UTC, giving an interval of only 30 minutes and 50 seconds between the discovery of the two off-critical-line zeros. | |||
Bannon later stated that he was present in the office when confirmation of the second zero arrived. According to his account, Simon Decourcy, head of the Berkeley Department of Mathematics, approached him shortly after the result had been assessed through the Kindred and Solace systems and announced, ''“The Collatz is down, and a second counterexample has hit the Riemann hypothesis.”'' The remark, widely reproduced in subsequent reporting, became one of the best-known quotations associated with the Day of Mathematics. Its resemblance to breaking-news language was subsequently amplified in online coverage, where the rapid succession of automated alerts from Kindred and Solace was frequently parodied as “a second hypothesis has been hit.” | |||