1 #+TITLE: Timothy Gowers - What is the 500th digit of $(\sqrt{2}+\sqrt{3})^{2012}$?
4 #+DESCRIPTION: IATC transcription of a short informal talk
5 #+KEYWORDS: mathematical argumentation
7 #+OPTIONS: H:3 num:t toc:t \n:nil @:t ::t |:t ^:nil -:t f:t *:t <:t
8 #+OPTIONS: TeX:t LaTeX:nil skip:nil d:nil todo:t pri:nil tags:not-in-toc
9 #+INFOJS_OPT: view:nil toc:nil ltoc:t mouse:underline buttons:0 path:http://orgmode.org/org-info.js
10 #+EXPORT_SELECT_TAGS: export
11 #+EXPORT_EXCLUDE_TAGS: noexport
14 #+HTML_HEAD: <link rel="stylesheet" type="text/css" href="http://metameso.org/~joe/solarized-css/build/solarized-light.css" />
15 #+STARTUP: showeverything
17 * What the 500th digit of $(\sqrt{2}+\sqrt{3})^{2012}$?
19 What the 500th digit of $(\sqrt{2}+\sqrt{3})^{2012}$?
22 IATC> perf[assert](meta[goal](compute 500th digit of (sqrt(2)+sqrt(3))^2012))
25 Even this, eventually, a computer will be able to solve.
28 IATC> perf[assert](rel[exists](meta[strategy](strategy X)))
29 IATC> perf[assert](rel[has_property](strategy X,HAL)
32 For now, notice that total stuckness can make you do desperate things.
35 IATC> perf[suggest](meta[strategy](simplify and compute))
38 Furthermore, knowing the origin of the problem suggests good things to try.
39 The fact that it is set as a problem is a huge clue.
42 IATC> perf[suggest](meta[strategy](this should be straightforward once we find the trick))
45 Can we do this for $x+y$? For $e$? Rationals with small denominator?
48 IATC> perf[assert](meta[analogy](compute 500th digit of (sqrt(2)+sqrt(3))^2012, compute 500th digit of (x+y)^2012))
49 IATC> perf[assert](meta[analogy](compute 500th digit of (sqrt(2)+sqrt(3))^2012, compute 500th digit of (e)^2012))
50 IATC> perf[assert](meta[analogy](compute 500th digit of (sqrt(2)+sqrt(3))^2012, compute 500th digit of (r1+r2)^2012 where r1 and r2 are small rationals))
51 IATC> perf[assert](rel[has_property](compute 500th digit of (r1+r2)^2012 where r1 and r2 are small rationals,we can really compute this: e.g. 500th digit of (1/10)^2012 is 0)))
54 How about small perturbations of these? Maybe it is close to a rational?
57 perf[suggest](meta[strategy](the trick might be: it is close to something we can compute))
60 \(m\)th digit of $(\sqrt{2}+\sqrt{3})^n$?
63 perf[suggest](meta[auxiliary](generalise(general form of the problem: mth digit of (sqrt(2)+sqrt(3))^n)))
66 $(\sqrt{2}+\sqrt{3})^2$
69 perf[suggest](meta[auxiliary](specialise((sqrt(2)+sqrt(3))^2)))
70 rel[instantiates]((sqrt(2)+sqrt(3))^2,general form of the problem)
71 perf[assert](rel[has_property]((sqrt(2)+sqrt(3))^2,we can really compute this))
74 $2+2\sqrt{2}\sqrt{3}+3$
77 perf[assert](rel[expands](2+2sqrt{2}sqrt{3}+3,(sqrt(2)+sqrt(3))^2))
80 $(\sqrt{2}+\sqrt{3})^2 + (\sqrt{2}-\sqrt{3})^2 = 10$
83 perf[suggest](meta[strategy](eliminate cross terms)
84 perf[assert](rel[expands](2+2sqrt(2)sqrt(3)+2 + 2 - 2sqrt(2)sqrt(3)+3,(sqrt(2)+sqrt(3))^2 + (sqrt(2)-sqrt(3))^2))
85 perf[assert](rel[sums](10,2+2sqrt(2)sqrt(3)+2 + 2 - 2sqrt(2)sqrt(3)+3))
88 $(\sqrt{2}+\sqrt{3})^{2012} + (\sqrt{2}-\sqrt{3})^{2012}$ is an integer!
91 perf[suggest](meta[strategy](binomial theorem))
92 rel[generalises](binomial theorem, eliminate cross terms)
95 And $(\sqrt{3}-\sqrt{2})^{2012}$ is a very small number.
96 Maybe the final answer is ``9''?
99 perf[assert](rel[contains as summand]((sqrt(2)+sqrt(3))^2012 + (sqrt(2)-sqrt(3))^2012,(sqrt(3)-sqrt(2))^2012)
100 perf[assert](rel[has_property]((sqrt(3)-sqrt(2))^2012,small))
101 perf[assert](rel[implements](SOME SUBGRAPH,the trick might be))
102 perf[suggest](meta[strategy](numbers that are very close to integers have 9 in many places of their decimal expansion)
105 We need to check whether it's small enough.
106 $(\sqrt{3}-\sqrt{2})^{2012} < (1/2)^{2012}=((1/2)^4)^{503}=(1/16)^{503}<.1^{503}$, so we're in luck.
109 perf[assert](sqrt(3)-sqrt(2) < 1/2)
110 perf[assert](0<a<b<1 => a^N<b^N)
111 perf[assert]((1/2)^2012=((1/2)^4))^503
112 perf[assert]((1/2)^4=1/16)
113 perf[assert](1/16<.1)
114 perf[assert](.1^503 has 502 0s in its decimal expansion)
115 perf[assert](an integer minus something with at least 502 0s in its decimal expansion has at least 503 9s in its decimal expansion)
118 The answer is indeed 9
120 perf[assert](rel[implies](SOME SUBGRAPH,general conclusion:(sqrt(2)+sqrt(3))^2012=(sqrt(2)+sqrt(3))^2012+(sqrt(2)-sqrt(3))^2012-(sqrt(2)-sqrt(3))^2012 has at least 503 9s in its decimal expansion))
121 perf[assert](in particular:(sqrt(2)+sqrt(3))^2012 has 500 9s in its decimal expansion)
122 perf[assert](rel[specialisation](in particular,general conclusion))