21.11.2022 Views

Theory of Knowledge - Course Companion for Students Marija Uzunova Dang Arvin Singh Uzunov Dang

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

will advans in omputing afft

How

To answr this qustion w nd

mathmatis?

hav a grasp on what mathmatis is, and

to

prisly, what mathmatiians do, and

mor

omputational powr fits with that. W

how

that formal proofs an partiularly

mntiond

vn impossil, for a mathmatiian to

long,

so omputr proof assistants hav om

vrify,

hav usd omputrs not just

Mathmatiians

vrify ut to gnrat proofs for som tim.

to

four-olour thorm ompltd in 1976

Th

11.6) was th first suh proof. It was

(Figur

ompliatd that no human ing ould

so

it without having to trust th softwar.

vrify

xampl is th 1998 proof of th Kplr

Anothr

whih ontaind ovr 3 gigayts of

onjtur,

“Proof y xhaustion” is a trm givn to

data.

whr a omputr has hkd all possil

proofs

Proofs of this natur ar onsidrd

ass.

aus thy an’t vrifid y a

prolmati

prson—som popl disrgard thm on

singl

asis, ut othrs apt thm.

this

thorm provrs (ATPs) ar

Automatd

that an prov mathmatial rsults

programs

using a st of axioms. ATPs an also

logially

millions of simulations in mission-ritial

run

suh as in nular powr plants,

appliations,

nsur a systm is working proprly—a task

to

would prov xtrmly tim-onsuming

that

human ings. ATPs ar hanging how w

for

Yt omputrs solv diffrntly

smathmatis.

humans, who look for graful onntions

from

possiilitis. Computr proofs an

twn

and awkward—as long as thy find a

lumsy

ths ar just th prolms assoiatd

Prhaps

a nw thnology; prhaps mathmatiians

with

omputr sintists may vntually al

and

od ATPs to dlivr autiful proofs that

to

nw insights to thir human dvloprs.

rval

prhaps not: whthr softwar is apal

Or

th fats of intuition and imagination that

of

th st xampls of human

haratriz

whthr ATPs will al to math

of

xd thir human ountrparts, th

or

of dvloping thm is dpning our

pursuit

of what it mans, and taks, to

undrstanding

is also possible to let computers loose to explore

It

on their own, and in some cases they

mathematics

come up with interesting conjectures that went

have

by mathematicians. We may be close to

unnoticed

how computers, rather than humans, would do

seeing

mathematics.

ommon far is, of ours, that on softwar is

A

nough at proving rsults, and in diding

good

rsults to prov, human mathmatiians

whih

om irrlvant. How would you did if

will

some supercomputer … reported a

Suppose

… which was so long and complex that no

proof

could understand it beyond the

mathematician

general terms. Could we have sucient faith

most

computers to accept this result, or would we say

in

the empirical evidence for their reliability is not

that

Tymozko oind th trm “nonsurvyal

Thomas

proof” to dsri proofs that ar

for a human mathmatiian to vrify,

infasil

as th 1979 omputr-assistd proof y

suh

and Hakn of th four-olour thorm.

Appl

argud that mathmatial proofs must

Tymozko

Conviningnss: th proof an prsuad a

vrifir of its onlusion.

rational

Survyaility: th proof is assil for

y humans.

vrifiation

Formalizaility: th proof uss only logial

twn onpts.

rlationships

III. Methods and tools

III. Methods and tools

III.1.1 Mathematical proof in the digital age

mak a mathmatial proof.

usful.

(American Mathematical Society 2008)

this far is misguidd or valid?

III.1.2 Non-surveyable proof

enough?

(Tymoczko 1979)

proof, thy hav don thir jo.

mt thr ritria.

mathmatis is an opn qustion. Rgardlss

339

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!