TED2015
Nick Bostrom: What happens when our computers get smarter than we are?
Nick Bostrom: Što će se dogoditi kada računala postanu pametnija od nas?
Filmed:
Readability: 4.4
4,632,705 views
Umjetna inteligencija postaje sve pametnija - prema istraživanju, u ovom bi stoljeću UI mogla biti jednako "pametna" kao ljudsko biće. A onda će, kaže Nick Bostrom, zavladati nama: "Strojna inteligencija je posljednji izum koje će čovječanstvo ikada trebati." Filozof i tehnolog nas traži da dobro razmislimo o svijetu koji upravo izgrađujemo, svijetu vođenom strojevima koji razmišljaju. Hoće li nam oni pomoći u očuvanju čovječanstva i naših vrijednosti ili će imati vlastita?
Nick Bostrom - Philosopher
Nick Bostrom asks big questions: What should we do, as individuals and as a species, to optimize our long-term prospects? Will humanity’s technological advancements ultimately destroy us? Full bio
Nick Bostrom asks big questions: What should we do, as individuals and as a species, to optimize our long-term prospects? Will humanity’s technological advancements ultimately destroy us? Full bio
Double-click the English transcript below to play the video.
00:12
I work with a bunch of mathematicians,
philosophers and computer scientists,
philosophers and computer scientists,
0
570
4207
Radim s mnogo matematičara,
filozofa i računalnih znanstvenika
filozofa i računalnih znanstvenika
00:16
and we sit around and think about
the future of machine intelligence,
the future of machine intelligence,
1
4777
5209
s kojima razmišljam o budućnosti
strojne inteligencije,
strojne inteligencije,
00:21
among other things.
2
9986
2044
među drugim stvarima.
00:24
Some people think that some of these
things are sort of science fiction-y,
things are sort of science fiction-y,
3
12030
4725
Neki ljudi misle da su te stvari
u zoni znanstvene fantastike,
u zoni znanstvene fantastike,
00:28
far out there, crazy.
4
16755
3101
nešto ludo, nestvarno.
00:31
But I like to say,
5
19856
1470
Ali ja volim reći
00:33
okay, let's look at the modern
human condition.
human condition.
6
21326
3604
ok, pogledajmo uvjete
modernoga života.
modernoga života.
00:36
(Laughter)
7
24930
1692
(Smijeh)
00:38
This is the normal way for things to be.
8
26622
2402
Ovo je normalno stanje stvari.
00:41
But if we think about it,
9
29024
2285
Ali ako malo razmislimo o tome,
00:43
we are actually recently arrived
guests on this planet,
guests on this planet,
10
31309
3293
mi smo zapravo tek nedavno
stigli na ovaj planet,
stigli na ovaj planet,
00:46
the human species.
11
34602
2082
mi, ljudska vrsta.
00:48
Think about if Earth
was created one year ago,
was created one year ago,
12
36684
4746
Kad bismo gledali na Zemlju kao da
je stvorena prije godinu dana,
je stvorena prije godinu dana,
00:53
the human species, then,
would be 10 minutes old.
would be 10 minutes old.
13
41430
3548
onda bi ljudska vrsta bila
stara deset minuta.
stara deset minuta.
00:56
The industrial era started
two seconds ago.
two seconds ago.
14
44978
3168
Industrijska je era počela
prije dvije sekunde.
prije dvije sekunde.
01:01
Another way to look at this is to think of
world GDP over the last 10,000 years,
world GDP over the last 10,000 years,
15
49276
5225
Drugi pogled na ovo jest taj da promatramo
svjetski BDP u zadnjih 10 000 godina.
svjetski BDP u zadnjih 10 000 godina.
01:06
I've actually taken the trouble
to plot this for you in a graph.
to plot this for you in a graph.
16
54501
3029
Odvojio sam nešto vremena i
to za vas grafički prikazao.
to za vas grafički prikazao.
01:09
It looks like this.
17
57530
1774
Izgleda ovako.
01:11
(Laughter)
18
59304
1363
(Smijeh)
01:12
It's a curious shape
for a normal condition.
for a normal condition.
19
60667
2151
Neobičan je to oblik za
uobičajene uvjete.
uobičajene uvjete.
01:14
I sure wouldn't want to sit on it.
20
62818
1698
Ja sigurno ne bih sjedio na njemu.
01:16
(Laughter)
21
64516
2551
(Smijeh)
01:19
Let's ask ourselves, what is the cause
of this current anomaly?
of this current anomaly?
22
67067
4774
Zapitajmo se koji je uzrok
ove trenutne anomalije.
ove trenutne anomalije.
01:23
Some people would say it's technology.
23
71841
2552
Neki bi ljudi rekli da je tehnologija.
01:26
Now it's true, technology has accumulated
through human history,
through human history,
24
74393
4668
Istina je da se tijekom povijesti
tehnologija nagomilavala
tehnologija nagomilavala
01:31
and right now, technology
advances extremely rapidly --
advances extremely rapidly --
25
79061
4652
i u ovom se trenutku
tehnologija jako brzo razvija --
tehnologija jako brzo razvija --
01:35
that is the proximate cause,
26
83713
1565
to je neposredni uzrok,
01:37
that's why we are currently
so very productive.
so very productive.
27
85278
2565
i zbog toga smo trenutno
jako produktivni.
jako produktivni.
01:40
But I like to think back further
to the ultimate cause.
to the ultimate cause.
28
88473
3661
Ali razmislimo o krajnjem uzroku.
01:45
Look at these two highly
distinguished gentlemen:
distinguished gentlemen:
29
93114
3766
Pogledajte ova dva vrlo
istaknuta gospodina:
istaknuta gospodina:
01:48
We have Kanzi --
30
96880
1600
imamo Kanzija --
01:50
he's mastered 200 lexical
tokens, an incredible feat.
tokens, an incredible feat.
31
98480
4643
koji je svladao 200 leksičkih
znakova što je nevjerojatan podvig.
znakova što je nevjerojatan podvig.
01:55
And Ed Witten unleashed the second
superstring revolution.
superstring revolution.
32
103123
3694
I Ed Witten koji je započeo drugu
revoluciju superstruna.
revoluciju superstruna.
01:58
If we look under the hood,
this is what we find:
this is what we find:
33
106817
2324
Ako pomno pogledamo,
uočit ćemo ovo:
uočit ćemo ovo:
02:01
basically the same thing.
34
109141
1570
u principu istu stvar.
02:02
One is a little larger,
35
110711
1813
Jedna je malo veća,
02:04
it maybe also has a few tricks
in the exact way it's wired.
in the exact way it's wired.
36
112524
2758
možda ima par specifičnosti
u načinu na koji je sklopljena.
u načinu na koji je sklopljena.
02:07
These invisible differences cannot
be too complicated, however,
be too complicated, however,
37
115282
3812
Međutim, ove nevidljive razlike
ne mogu biti jako složene
ne mogu biti jako složene
02:11
because there have only
been 250,000 generations
been 250,000 generations
38
119094
4285
jer nas dijeli samo 250 000 generacija
02:15
since our last common ancestor.
39
123379
1732
od zadnjeg zajedničkog pretka.
02:17
We know that complicated mechanisms
take a long time to evolve.
take a long time to evolve.
40
125111
3849
Znamo da se složeni
mehanizmi dugo razvijaju.
mehanizmi dugo razvijaju.
02:22
So a bunch of relatively minor changes
41
130000
2499
Tako da nas hrpa relativno malih promjena
02:24
take us from Kanzi to Witten,
42
132499
3067
vodi od Kanzija do Wittena,
02:27
from broken-off tree branches
to intercontinental ballistic missiles.
to intercontinental ballistic missiles.
43
135566
4543
od polomljenih grana do
interkontinentalnih balističkih raketa.
interkontinentalnih balističkih raketa.
02:32
So this then seems pretty obvious
that everything we've achieved,
that everything we've achieved,
44
140839
3935
Onda je poprilično očito
da sve što smo postigli
da sve što smo postigli
02:36
and everything we care about,
45
144774
1378
i sve do čega nam je stalo
02:38
depends crucially on some relatively minor
changes that made the human mind.
changes that made the human mind.
46
146152
5228
uvelike ovisi o relativno malim
promjenama koje čine ljudski um.
promjenama koje čine ljudski um.
02:44
And the corollary, of course,
is that any further changes
is that any further changes
47
152650
3662
Nužna je posljedica ta
da sve daljnje izmjene
da sve daljnje izmjene
02:48
that could significantly change
the substrate of thinking
the substrate of thinking
48
156312
3477
koje bi mogle značajno promijeniti
način razmišljanja
način razmišljanja
02:51
could have potentially
enormous consequences.
enormous consequences.
49
159789
3202
mogu imati potencijalno
ogromne posljedice.
ogromne posljedice.
02:56
Some of my colleagues
think we're on the verge
think we're on the verge
50
164321
2905
Neki od mojih kolega
smatraju da smo na rubu
smatraju da smo na rubu
02:59
of something that could cause
a profound change in that substrate,
a profound change in that substrate,
51
167226
3908
nečega što bi moglo izazvati
takve temeljite promjene,
takve temeljite promjene,
03:03
and that is machine superintelligence.
52
171134
3213
a radi se o strojnoj superinteligenciji.
03:06
Artificial intelligence used to be
about putting commands in a box.
about putting commands in a box.
53
174347
4739
Prije je umjetna inteligencija bila
pisanje naredbi u kutiju.
pisanje naredbi u kutiju.
03:11
You would have human programmers
54
179086
1665
Ljudski bi programeri
03:12
that would painstakingly
handcraft knowledge items.
handcraft knowledge items.
55
180751
3135
mukotrpno ručno stvarali stavke znanja.
03:15
You build up these expert systems,
56
183886
2086
Izgradite takve stručne sustave
03:17
and they were kind of useful
for some purposes,
for some purposes,
57
185972
2324
koji su za određene svrhe korisni,
03:20
but they were very brittle,
you couldn't scale them.
you couldn't scale them.
58
188296
2681
ali su bili krhki i
nisu bili mjerljivi.
nisu bili mjerljivi.
03:22
Basically, you got out only
what you put in.
what you put in.
59
190977
3433
U principu ste dobivali samo
ono što ste unijeli.
ono što ste unijeli.
03:26
But since then,
60
194410
997
Ali, od onda se
03:27
a paradigm shift has taken place
in the field of artificial intelligence.
in the field of artificial intelligence.
61
195407
3467
dogodila promjena u pristupu
u polju umjetne inteligencije.
u polju umjetne inteligencije.
03:30
Today, the action is really
around machine learning.
around machine learning.
62
198874
2770
Danas se uglavnom radi o
učenju strojeva.
učenju strojeva.
03:34
So rather than handcrafting knowledge
representations and features,
representations and features,
63
202394
5387
Umjesto ručnog unosa prikaza
znanja i značajki,
znanja i značajki,
03:40
we create algorithms that learn,
often from raw perceptual data.
often from raw perceptual data.
64
208511
5554
stvaramo algoritme koji uče, i to često
iz čistih perceptualnih podataka.
iz čistih perceptualnih podataka.
03:46
Basically the same thing
that the human infant does.
that the human infant does.
65
214065
4998
Načelno je to ista stvar
koju čini ljudsko dijete.
koju čini ljudsko dijete.
03:51
The result is A.I. that is not
limited to one domain --
limited to one domain --
66
219063
4207
Rezultat je taj da UI nije
ograničena na jedno područje --
ograničena na jedno područje --
03:55
the same system can learn to translate
between any pairs of languages,
between any pairs of languages,
67
223270
4631
isti sustav može naučiti prevoditi
bilo koju kombinaciju jezika
bilo koju kombinaciju jezika
03:59
or learn to play any computer game
on the Atari console.
on the Atari console.
68
227901
5437
ili naučiti igrati bilo koju
računalnu igru na Atariju.
računalnu igru na Atariju.
04:05
Now of course,
69
233338
1779
Naravno,
04:07
A.I. is still nowhere near having
the same powerful, cross-domain
the same powerful, cross-domain
70
235117
3999
UI nije ni blizu posjedovanja
iste jake, unakrsne sposobnosti
iste jake, unakrsne sposobnosti
04:11
ability to learn and plan
as a human being has.
as a human being has.
71
239116
3219
učenja i planiranja raznih područja
kao što to radi ljudsko biće.
kao što to radi ljudsko biće.
04:14
The cortex still has some
algorithmic tricks
algorithmic tricks
72
242335
2126
Korteks ima neke algoritamske trikove
04:16
that we don't yet know
how to match in machines.
how to match in machines.
73
244461
2355
koje ne znamo kako prenijeti strojevima.
04:19
So the question is,
74
247886
1899
Tako da je pitanje
04:21
how far are we from being able
to match those tricks?
to match those tricks?
75
249785
3500
koliko smo blizu tome da
uskladimo te trikove.
uskladimo te trikove.
04:26
A couple of years ago,
76
254245
1083
Prije nekoliko godina
04:27
we did a survey of some of the world's
leading A.I. experts,
leading A.I. experts,
77
255328
2888
proveli smo istraživanje s
vodećim stručnjacima za UI
vodećim stručnjacima za UI
04:30
to see what they think,
and one of the questions we asked was,
and one of the questions we asked was,
78
258216
3224
kako bismo vidjeli što misle,
i jedno od pitanja bilo je
i jedno od pitanja bilo je
04:33
"By which year do you think
there is a 50 percent probability
there is a 50 percent probability
79
261440
3353
"Što mislite, do koje će godine
biti 50% vjerojatnosti
biti 50% vjerojatnosti
04:36
that we will have achieved
human-level machine intelligence?"
human-level machine intelligence?"
80
264793
3482
da ćemo postići strojnu inteligenciju
na razini ljudske inteligencije?"
na razini ljudske inteligencije?"
04:40
We defined human-level here
as the ability to perform
as the ability to perform
81
268785
4183
Ljudsku smo razinu definirali kao
sposobnost da se bilo koji zadatak
sposobnost da se bilo koji zadatak
04:44
almost any job at least as well
as an adult human,
as an adult human,
82
272968
2871
obavi jednako dobro kao što bi
ga obavio odrastao čovjek,
ga obavio odrastao čovjek,
04:47
so real human-level, not just
within some limited domain.
within some limited domain.
83
275839
4005
znači, stvarna ljudska razina, a ne
samo unutar određenog područja.
samo unutar određenog područja.
04:51
And the median answer was 2040 or 2050,
84
279844
3650
Srednji je odgovor bio 2040. ili 2050.,
04:55
depending on precisely which
group of experts we asked.
group of experts we asked.
85
283494
2806
ovisno o skupini stručnjaka
kojoj smo pitanje postavili.
kojoj smo pitanje postavili.
04:58
Now, it could happen much,
much later, or sooner,
much later, or sooner,
86
286300
4039
To bi se moglo dogoditi puno, puno
kasnije ili ranije,
kasnije ili ranije,
05:02
the truth is nobody really knows.
87
290339
1940
istina je ta da nitko to zapravo ne zna.
05:05
What we do know is that the ultimate
limit to information processing
limit to information processing
88
293259
4412
Ali ono što znamo jest to da krajnja
granica procesuiranja informacija
granica procesuiranja informacija
05:09
in a machine substrate lies far outside
the limits in biological tissue.
the limits in biological tissue.
89
297671
4871
u strojevima uvelike nadilazi
ograničenja biološkog tkiva.
ograničenja biološkog tkiva.
05:15
This comes down to physics.
90
303241
2378
Radi se o fizici.
05:17
A biological neuron fires, maybe,
at 200 hertz, 200 times a second.
at 200 hertz, 200 times a second.
91
305619
4718
Biološki se neuron ispaljuje na možda
200 herca, 200 puta u sekundi.
200 herca, 200 puta u sekundi.
05:22
But even a present-day transistor
operates at the Gigahertz.
operates at the Gigahertz.
92
310337
3594
Ali čak i današnji tranzistor
radi u gigahercima.
radi u gigahercima.
05:25
Neurons propagate slowly in axons,
100 meters per second, tops.
100 meters per second, tops.
93
313931
5297
Neuroni se polako šire u aksone,
najviše 100 metara u sekundi.
najviše 100 metara u sekundi.
05:31
But in computers, signals can travel
at the speed of light.
at the speed of light.
94
319228
3111
Ali u računalima signali mogu
putovati brzinom svjetlosti.
putovati brzinom svjetlosti.
05:35
There are also size limitations,
95
323079
1869
I tamo postoje ograničenja u veličini,
05:36
like a human brain has
to fit inside a cranium,
to fit inside a cranium,
96
324948
3027
isto kao što ljudski mozak
mora stati u lubanju,
mora stati u lubanju,
05:39
but a computer can be the size
of a warehouse or larger.
of a warehouse or larger.
97
327975
4761
ali računalo može biti veličine
skladišta, ili čak veće.
skladišta, ili čak veće.
05:44
So the potential for superintelligence
lies dormant in matter,
lies dormant in matter,
98
332736
5599
Potencijal superinteligencije
prikriven je i čeka,
prikriven je i čeka,
05:50
much like the power of the atom
lay dormant throughout human history,
lay dormant throughout human history,
99
338335
5712
slično kao što je snaga atoma
bila prikrivena tijekom povijesti,
bila prikrivena tijekom povijesti,
05:56
patiently waiting there until 1945.
100
344047
4405
strpljivo čekajući 1945. godinu.
06:00
In this century,
101
348452
1248
U ovom su stoljeću
06:01
scientists may learn to awaken
the power of artificial intelligence.
the power of artificial intelligence.
102
349700
4118
znanstvenici možda otkrili kako
probuditi moć umjetne inteligencije.
probuditi moć umjetne inteligencije.
06:05
And I think we might then see
an intelligence explosion.
an intelligence explosion.
103
353818
4008
A onda bismo mogli svjedočiti
eksploziji inteligencije.
eksploziji inteligencije.
06:10
Now most people, when they think
about what is smart and what is dumb,
about what is smart and what is dumb,
104
358406
3957
Većina ljudi, kada razmišlja o
tome što je pametno, a što glupo,
tome što je pametno, a što glupo,
06:14
I think have in mind a picture
roughly like this.
roughly like this.
105
362363
3023
mislim da otprilike zamišlja
ovako nešto.
ovako nešto.
06:17
So at one end we have the village idiot,
106
365386
2598
Na jednom kraju imamo seosku budalu,
06:19
and then far over at the other side
107
367984
2483
a na drugome
06:22
we have Ed Witten, or Albert Einstein,
or whoever your favorite guru is.
or whoever your favorite guru is.
108
370467
4756
Eda Wittena ili Alberta Einsteina,
ili već preferiranog gurua.
ili već preferiranog gurua.
06:27
But I think that from the point of view
of artificial intelligence,
of artificial intelligence,
109
375223
3834
Ali iz perspektive umjetne inteligencije,
06:31
the true picture is actually
probably more like this:
probably more like this:
110
379057
3681
ta slika vjerojatno izgleda ovako:
06:35
AI starts out at this point here,
at zero intelligence,
at zero intelligence,
111
383258
3378
sve počinje na točki nulte inteligencije,
06:38
and then, after many, many
years of really hard work,
years of really hard work,
112
386636
3011
a onda nakon puno, puno
godina napornog rada
godina napornog rada
06:41
maybe eventually we get to
mouse-level artificial intelligence,
mouse-level artificial intelligence,
113
389647
3844
možda konačno dođemo do
umjetne inteligencije na razini miša,
umjetne inteligencije na razini miša,
06:45
something that can navigate
cluttered environments
cluttered environments
114
393491
2430
nešto što se može kretati
kroz pretrpan prostor
kroz pretrpan prostor
06:47
as well as a mouse can.
115
395921
1987
jednako dobro kao miševi.
06:49
And then, after many, many more years
of really hard work, lots of investment,
of really hard work, lots of investment,
116
397908
4313
A onda nakon još puno, puno godina
napornog rada i mnogo ulaganja,
napornog rada i mnogo ulaganja,
06:54
maybe eventually we get to
chimpanzee-level artificial intelligence.
chimpanzee-level artificial intelligence.
117
402221
4639
možda dođemo do umjetne
inteligencije na razini čimpanze.
inteligencije na razini čimpanze.
06:58
And then, after even more years
of really, really hard work,
of really, really hard work,
118
406860
3210
A onda, nakon još puno godina
vrlo, vrlo napornog rada,
vrlo, vrlo napornog rada,
07:02
we get to village idiot
artificial intelligence.
artificial intelligence.
119
410070
2913
dođemo do seoske budale
umjetne inteligencije.
umjetne inteligencije.
07:04
And a few moments later,
we are beyond Ed Witten.
we are beyond Ed Witten.
120
412983
3272
I nakon nekoliko trenutaka
nadiđemo Eda Wittena.
nadiđemo Eda Wittena.
07:08
The train doesn't stop
at Humanville Station.
at Humanville Station.
121
416255
2970
Tu napredak ne prestaje
07:11
It's likely, rather, to swoosh right by.
122
419225
3022
već se ubrzano nastavlja.
07:14
Now this has profound implications,
123
422247
1984
To ima duboke implikacije,
07:16
particularly when it comes
to questions of power.
to questions of power.
124
424231
3862
posebno kad se radi o
pitanju moći.
pitanju moći.
07:20
For example, chimpanzees are strong --
125
428093
1899
Na primjer, čimpanze su snažne --
07:21
pound for pound, a chimpanzee is about
twice as strong as a fit human male.
twice as strong as a fit human male.
126
429992
5222
dvostruko su jače od muškaraca.
07:27
And yet, the fate of Kanzi
and his pals depends a lot more
and his pals depends a lot more
127
435214
4614
A ipak, sudbina Kanzija i njegovih
prijatelja više ovisi o tome što
prijatelja više ovisi o tome što
07:31
on what we humans do than on
what the chimpanzees do themselves.
what the chimpanzees do themselves.
128
439828
4140
mi ljudi radimo nego o tome
što same čimpanze rade.
što same čimpanze rade.
07:37
Once there is superintelligence,
129
445228
2314
Jednom kada se stvori superinteligencija,
07:39
the fate of humanity may depend
on what the superintelligence does.
on what the superintelligence does.
130
447542
3839
sudbina će čovječanstva možda
ovisiti o tome što ona radi.
ovisiti o tome što ona radi.
07:44
Think about it:
131
452451
1057
Razmislite.
07:45
Machine intelligence is the last invention
that humanity will ever need to make.
that humanity will ever need to make.
132
453508
5044
Strojna inteligencija je zadnji izum
koji će čovječanstvo ikada trebati.
koji će čovječanstvo ikada trebati.
07:50
Machines will then be better
at inventing than we are,
at inventing than we are,
133
458552
2973
Strojevi će onda biti bolji u
izumljivanju od nas,
izumljivanju od nas,
07:53
and they'll be doing so
on digital timescales.
on digital timescales.
134
461525
2540
i oni će to raditi u digitalnom
vremenskom rasponu.
vremenskom rasponu.
07:56
What this means is basically
a telescoping of the future.
a telescoping of the future.
135
464065
4901
Načelno to znači
teleskopiranje budućnosti.
teleskopiranje budućnosti.
08:00
Think of all the crazy technologies
that you could have imagined
that you could have imagined
136
468966
3558
Zamislite sve moguće
lude tehnologije
lude tehnologije
08:04
maybe humans could have developed
in the fullness of time:
in the fullness of time:
137
472524
2798
koje bi ljudi mogli razviti:
lijek protiv starenja,
lijek protiv starenja,
08:07
cures for aging, space colonization,
138
475322
3258
kolonizacija svemira,
08:10
self-replicating nanobots or uploading
of minds into computers,
of minds into computers,
139
478580
3731
samoreplicirajući nanoboti,
unošenje uma u računalo,
unošenje uma u računalo,
08:14
all kinds of science fiction-y stuff
140
482311
2159
svakakve znanstveno-fantastične stvari
08:16
that's nevertheless consistent
with the laws of physics.
with the laws of physics.
141
484470
2737
koje su u skladu sa zakonima fizike.
08:19
All of this superintelligence could
develop, and possibly quite rapidly.
develop, and possibly quite rapidly.
142
487207
4212
Superinteligencija mogla bi se razviti,
i to dosta brzo.
i to dosta brzo.
08:24
Now, a superintelligence with such
technological maturity
technological maturity
143
492449
3558
Superinteligencija s takvom
tehnološkom zrelošću
tehnološkom zrelošću
08:28
would be extremely powerful,
144
496007
2179
bila bi iznimno moćna,
08:30
and at least in some scenarios,
it would be able to get what it wants.
it would be able to get what it wants.
145
498186
4546
i bar bi u nekim slučajevima,
mogla dobiti što želi.
mogla dobiti što želi.
08:34
We would then have a future that would
be shaped by the preferences of this A.I.
be shaped by the preferences of this A.I.
146
502732
5661
Onda bismo u budućnosti imali svijet
oblikovan prema preferencijama te UI.
oblikovan prema preferencijama te UI.
08:41
Now a good question is,
what are those preferences?
what are those preferences?
147
509855
3749
Dobro je pitanje koje su te preference.
08:46
Here it gets trickier.
148
514244
1769
Tu se stvari kompliciraju.
08:48
To make any headway with this,
149
516013
1435
Da bismo napredovali s tim,
08:49
we must first of all
avoid anthropomorphizing.
avoid anthropomorphizing.
150
517448
3276
moramo prvo izbjeći antropomorfizaciju.
08:53
And this is ironic because
every newspaper article
every newspaper article
151
521934
3301
A to je ironično jer
svi novinski članci
svi novinski članci
08:57
about the future of A.I.
has a picture of this:
has a picture of this:
152
525235
3855
o budućnosti UI
imaju sliku ovoga:
imaju sliku ovoga:
09:02
So I think what we need to do is
to conceive of the issue more abstractly,
to conceive of the issue more abstractly,
153
530280
4134
Tako da mislim da trebamo pojmiti
ovo pitanje malo apstraktnije,
ovo pitanje malo apstraktnije,
09:06
not in terms of vivid Hollywood scenarios.
154
534414
2790
a ne u obliku holivudskih scenarija.
09:09
We need to think of intelligence
as an optimization process,
as an optimization process,
155
537204
3617
Inteligenciju trebamo promatrati
kao proces optimizacije,
kao proces optimizacije,
09:12
a process that steers the future
into a particular set of configurations.
into a particular set of configurations.
156
540821
5649
proces koji usmjerava budućnost
u određeni set konfiguracija.
u određeni set konfiguracija.
09:18
A superintelligence is
a really strong optimization process.
a really strong optimization process.
157
546470
3511
Superinteligencija je zaista jak
proces optimizacije.
proces optimizacije.
09:21
It's extremely good at using
available means to achieve a state
available means to achieve a state
158
549981
4117
Nevjerojatno je dobra u korištenju
raspoloživih sredstava kako bi postigla
raspoloživih sredstava kako bi postigla
09:26
in which its goal is realized.
159
554098
1909
svoj cilj.
09:28
This means that there is no necessary
conenction between
conenction between
160
556447
2672
To znači da nužno ne postoji
povezanost između
povezanost između
09:31
being highly intelligent in this sense,
161
559119
2734
visoke inteligencije u tom smislu
09:33
and having an objective that we humans
would find worthwhile or meaningful.
would find worthwhile or meaningful.
162
561853
4662
i cilja koji bismo mi ljudi smatrali
vrijednim ili smislenim.
vrijednim ili smislenim.
09:39
Suppose we give an A.I. the goal
to make humans smile.
to make humans smile.
163
567321
3794
Recimo da UI damo cilj
da nasmiju ljude.
da nasmiju ljude.
09:43
When the A.I. is weak, it performs useful
or amusing actions
or amusing actions
164
571115
2982
Kad je UI slaba, izvodi korisne
ili zabavne radnje
ili zabavne radnje
09:46
that cause its user to smile.
165
574097
2517
zbog kojih se njihov korisnik smije.
09:48
When the A.I. becomes superintelligent,
166
576614
2417
Kada UI postane superinteligentna,
09:51
it realizes that there is a more
effective way to achieve this goal:
effective way to achieve this goal:
167
579031
3523
uvidjet će da postoji učinkovitiji
način na koji postići taj cilj:
način na koji postići taj cilj:
09:54
take control of the world
168
582554
1922
preuzimanje kontrole nad svijetom
09:56
and stick electrodes into the facial
muscles of humans
muscles of humans
169
584476
3162
i smještanje elektroda u
mišiće lica ljudi
mišiće lica ljudi
09:59
to cause constant, beaming grins.
170
587638
2941
kako bi izazvala neprestan, vedar osmijeh.
10:02
Another example,
171
590579
1035
Još jedan primjer,
10:03
suppose we give A.I. the goal to solve
a difficult mathematical problem.
a difficult mathematical problem.
172
591614
3383
recimo da zadamo UI da riješi
zahtjevan matematički problem.
zahtjevan matematički problem.
10:06
When the A.I. becomes superintelligent,
173
594997
1937
Kad UI postane superinteligentna,
10:08
it realizes that the most effective way
to get the solution to this problem
to get the solution to this problem
174
596934
4171
uviđa da je najučinkovitiji način za
rješavanje toga problema
rješavanje toga problema
10:13
is by transforming the planet
into a giant computer,
into a giant computer,
175
601105
2930
preobrazba planeta u ogromno računalo,
10:16
so as to increase its thinking capacity.
176
604035
2246
da bi povećala njegovu
sposobnost razmišljanja.
sposobnost razmišljanja.
10:18
And notice that this gives the A.I.s
an instrumental reason
an instrumental reason
177
606281
2764
I imajte na umu da to daje UI
instrumentalni razlog da
instrumentalni razlog da
10:21
to do things to us that we
might not approve of.
might not approve of.
178
609045
2516
nam radi stvari koje mi
možda ne odobravamo.
možda ne odobravamo.
10:23
Human beings in this model are threats,
179
611561
1935
U ovom su primjeri ljudi prijetnja,
10:25
we could prevent the mathematical
problem from being solved.
problem from being solved.
180
613496
2921
mogli bismo spriječiti rješavanje
matematičkih problema.
matematičkih problema.
10:29
Of course, perceivably things won't
go wrong in these particular ways;
go wrong in these particular ways;
181
617207
3494
Naravno, stvari neće baš
tako poći po krivu;
tako poći po krivu;
10:32
these are cartoon examples.
182
620701
1753
ovo su primjeri iz crtića.
10:34
But the general point here is important:
183
622454
1939
Ali važna je poanta:
10:36
if you create a really powerful
optimization process
optimization process
184
624393
2873
ako stvorite zaista moćan
proces optimizacije
proces optimizacije
10:39
to maximize for objective x,
185
627266
2234
kako biste maksimizirali cilj x,
10:41
you better make sure
that your definition of x
that your definition of x
186
629500
2276
morate biti sigurni da
vaša definicija x-a
vaša definicija x-a
10:43
incorporates everything you care about.
187
631776
2469
uključuje sve do čega vam je stalo.
10:46
This is a lesson that's also taught
in many a myth.
in many a myth.
188
634835
4384
Ovo je lekcija na koju nas
upozoravaju mnogi mitovi.
upozoravaju mnogi mitovi.
10:51
King Midas wishes that everything
he touches be turned into gold.
he touches be turned into gold.
189
639219
5298
Kralj Mida želi da se sve što
dotakne pretvori u zlato.
dotakne pretvori u zlato.
10:56
He touches his daughter,
she turns into gold.
she turns into gold.
190
644517
2861
Dodirne kćer i ona se pretvori u zlato.
10:59
He touches his food, it turns into gold.
191
647378
2553
Dodirne hranu, ona postane zlato.
11:01
This could become practically relevant,
192
649931
2589
Ovo bi moglo postati i praktički važno,
11:04
not just as a metaphor for greed,
193
652520
2070
ne samo kao metafora za pohlepu,
11:06
but as an illustration of what happens
194
654590
1895
nego i kao prikaz onoga što se dogodi
11:08
if you create a powerful
optimization process
optimization process
195
656485
2837
ako stvorite moćan
proces optimizacije
proces optimizacije
11:11
and give it misconceived
or poorly specified goals.
or poorly specified goals.
196
659322
4789
i date mu pogrešne ili loše
određene ciljeve.
određene ciljeve.
11:16
Now you might say, if a computer starts
sticking electrodes into people's faces,
sticking electrodes into people's faces,
197
664111
5189
Mogli biste reći da, ako računalo
počne stavljati elektrode u lica ljudi,
počne stavljati elektrode u lica ljudi,
11:21
we'd just shut it off.
198
669300
2265
jednostavno ćemo ga isključiti.
11:24
A, this is not necessarily so easy to do
if we've grown dependent on the system --
if we've grown dependent on the system --
199
672555
5340
Pod a, to nije tako jednostavno
ako smo već ovisni o sustavu --
ako smo već ovisni o sustavu --
11:29
like, where is the off switch
to the Internet?
to the Internet?
200
677895
2732
na primjer, kako isključiti internet?
11:32
B, why haven't the chimpanzees
flicked the off switch to humanity,
flicked the off switch to humanity,
201
680627
5120
Pod b, zašto čimpanze nisu
isključile prekidač ljudske vrste,
isključile prekidač ljudske vrste,
11:37
or the Neanderthals?
202
685747
1551
oni ili neandertalci?
11:39
They certainly had reasons.
203
687298
2666
Zasigurno su postojali razlozi za to.
11:41
We have an off switch,
for example, right here.
for example, right here.
204
689964
2795
Mi imamo prekidač, recimo,
upravo ovdje.
upravo ovdje.
11:44
(Choking)
205
692759
1554
(Guši se)
11:46
The reason is that we are
an intelligent adversary;
an intelligent adversary;
206
694313
2925
Razlog je taj što smo mi
inteligentni protivnici;
inteligentni protivnici;
11:49
we can anticipate threats
and plan around them.
and plan around them.
207
697238
2728
predviđamo prijetnje i
planiramo kako ih izbjeći.
planiramo kako ih izbjeći.
11:51
But so could a superintelligent agent,
208
699966
2504
Ali to bi mogao raditi i
superinteligentni izvršitelj
superinteligentni izvršitelj
11:54
and it would be much better
at that than we are.
at that than we are.
209
702470
3254
i to puno bolje nego
što mi to radimo.
što mi to radimo.
11:57
The point is, we should not be confident
that we have this under control here.
that we have this under control here.
210
705724
7187
Poanta je ovoga da ne smijemo biti
uvjereni da sve držimo pod kontrolom.
uvjereni da sve držimo pod kontrolom.
12:04
And we could try to make our job
a little bit easier by, say,
a little bit easier by, say,
211
712911
3447
Mogli bismo si olakšati posao tako što
12:08
putting the A.I. in a box,
212
716358
1590
bismo smjestili UI u kutiju,
12:09
like a secure software environment,
213
717948
1796
nekakvo sigurno softversko okruženje,
12:11
a virtual reality simulation
from which it cannot escape.
from which it cannot escape.
214
719744
3022
virtualnu simulaciju stvarnosti iz
koje ne može pobjeći.
koje ne može pobjeći.
12:14
But how confident can we be that
the A.I. couldn't find a bug.
the A.I. couldn't find a bug.
215
722766
4146
Ali kako možemo biti sigurni
da UI neće pronaći rupu?
da UI neće pronaći rupu?
12:18
Given that merely human hackers
find bugs all the time,
find bugs all the time,
216
726912
3169
Ako uobzirimo da hakeri
neprestano pronalaze rupe,
neprestano pronalaze rupe,
12:22
I'd say, probably not very confident.
217
730081
3036
vjerojatno ne možemo biti sigurni u to.
12:26
So we disconnect the ethernet cable
to create an air gap,
to create an air gap,
218
734237
4548
Isključimo Ethernet kabel da bismo
stvorili zračni prostor,
stvorili zračni prostor,
12:30
but again, like merely human hackers
219
738785
2668
ali opet, ljudski hakeri rutinski nadilaze
12:33
routinely transgress air gaps
using social engineering.
using social engineering.
220
741453
3381
te prostore socijalnim inženjeringom.
12:36
Right now, as I speak,
221
744834
1259
Dok ovo govorim,
12:38
I'm sure there is some employee
out there somewhere
out there somewhere
222
746093
2389
siguran sam da su negdje neku zaposlenicu
12:40
who has been talked into handing out
her account details
her account details
223
748482
3346
nagovorili da otkrije
detalje svog računa
detalje svog računa
12:43
by somebody claiming to be
from the I.T. department.
from the I.T. department.
224
751828
2746
nekome tko je tvrdio
da je iz IT odjela.
da je iz IT odjela.
12:46
More creative scenarios are also possible,
225
754574
2127
Mogući su i kreativniji scenariji,
12:48
like if you're the A.I.,
226
756701
1315
na primjer, ako ste UI,
12:50
you can imagine wiggling electrodes
around in your internal circuitry
around in your internal circuitry
227
758016
3532
možete zamisliti micanje elektroda
u unutrašnjem sklopu kako biste
u unutrašnjem sklopu kako biste
12:53
to create radio waves that you
can use to communicate.
can use to communicate.
228
761548
3462
stvorili radio valove kojima
možete komunicirati.
možete komunicirati.
12:57
Or maybe you could pretend to malfunction,
229
765010
2424
Ili se možete prevarati da ste neispravni
12:59
and then when the programmers open
you up to see what went wrong with you,
you up to see what went wrong with you,
230
767434
3497
i kada vas programeri otvore da vide
što nije u redu s vama,
što nije u redu s vama,
13:02
they look at the source code -- Bam! --
231
770931
1936
pogledaju izvorni kod i -- Bum! --
13:04
the manipulation can take place.
232
772867
2447
manipulacija je počela.
13:07
Or it could output the blueprint
to a really nifty technology,
to a really nifty technology,
233
775314
3430
Ili bi moglo dati nacrt
zaista divne tehnologije,
zaista divne tehnologije,
13:10
and when we implement it,
234
778744
1398
a kada ga primijenimo,
13:12
it has some surreptitious side effect
that the A.I. had planned.
that the A.I. had planned.
235
780142
4397
ima neke skrivene posljedice
koje je UI isplanirala.
koje je UI isplanirala.
13:16
The point here is that we should
not be confident in our ability
not be confident in our ability
236
784539
3463
Poanta jest ta da ne bismo trebali biti
presigurni u svoju sposobnost
presigurni u svoju sposobnost
13:20
to keep a superintelligent genie
locked up in its bottle forever.
locked up in its bottle forever.
237
788002
3808
da superinteligentnog duha
možemo zauvijek čuvati u boci.
možemo zauvijek čuvati u boci.
13:23
Sooner or later, it will out.
238
791810
2254
Prije ili kasnije će se osloboditi.
13:27
I believe that the answer here
is to figure out
is to figure out
239
795034
3103
Vjerujem da je rješenje u
pronalasku načina kako
pronalasku načina kako
13:30
how to create superintelligent A.I.
such that even if -- when -- it escapes,
such that even if -- when -- it escapes,
240
798137
5024
stvoriti superinteligentnu UI
koja bi čak - kad - se oslobodi
koja bi čak - kad - se oslobodi
13:35
it is still safe because it is
fundamentally on our side
fundamentally on our side
241
803161
3277
još uvijek bila sigurna jer
je u osnovi na našoj strani
je u osnovi na našoj strani
13:38
because it shares our values.
242
806438
1899
jer dijeli naša uvjerenja.
13:40
I see no way around
this difficult problem.
this difficult problem.
243
808337
3210
Ne vidim kako drukčije
riješiti ovaj veliki problem.
riješiti ovaj veliki problem.
13:44
Now, I'm actually fairly optimistic
that this problem can be solved.
that this problem can be solved.
244
812557
3834
Zapravo sam ja i optimističan kada
mislim da je problem rješiv.
mislim da je problem rješiv.
13:48
We wouldn't have to write down
a long list of everything we care about,
a long list of everything we care about,
245
816391
3903
Ne bismo morali pisati duge liste
svega do čega nam je stalo,
svega do čega nam je stalo,
13:52
or worse yet, spell it out
in some computer language
in some computer language
246
820294
3643
ili, još gore, unijeti to u
neki računalni jezik
neki računalni jezik
13:55
like C++ or Python,
247
823937
1454
kao C++ ili Python,
13:57
that would be a task beyond hopeless.
248
825391
2767
to bi bio i više nego beznadan pothvat!
14:00
Instead, we would create an A.I.
that uses its intelligence
that uses its intelligence
249
828158
4297
Umjesto toga, stvorimo UI koja
koristi svoju inteligenciju kako
koristi svoju inteligenciju kako
14:04
to learn what we value,
250
832455
2771
bi naučila što nam je važno,
14:07
and its motivation system is constructed
in such a way that it is motivated
in such a way that it is motivated
251
835226
5280
i čiji je motivacijski sustav konstruiran
tako da bude motivirana za
tako da bude motivirana za
14:12
to pursue our values or to perform actions
that it predicts we would approve of.
that it predicts we would approve of.
252
840506
5232
provođenje naših uvjerenja ili izvođenje
radnji za koje predviđa da ih odobravamo.
radnji za koje predviđa da ih odobravamo.
14:17
We would thus leverage
its intelligence as much as possible
its intelligence as much as possible
253
845738
3414
Tako bismo iskoristili njezinu
inteligenciju što je više moguće
inteligenciju što je više moguće
14:21
to solve the problem of value-loading.
254
849152
2745
kako bismo riješili ovaj problem.
14:24
This can happen,
255
852727
1512
To se može dogoditi,
14:26
and the outcome could be
very good for humanity.
very good for humanity.
256
854239
3596
a ishod bi mogao biti jako
dobar za čovječanstvo.
dobar za čovječanstvo.
14:29
But it doesn't happen automatically.
257
857835
3957
Ali to se ne događa automatski.
14:33
The initial conditions
for the intelligence explosion
for the intelligence explosion
258
861792
2998
Polazni uvjeti za
eksploziju inteligencije
eksploziju inteligencije
14:36
might need to be set up
in just the right way
in just the right way
259
864790
2863
morat će biti postavljeni
na točno određen način
na točno određen način
14:39
if we are to have a controlled detonation.
260
867653
3530
ako želimo kontroliranu detonaciju.
14:43
The values that the A.I. has
need to match ours,
need to match ours,
261
871183
2618
Vrijednosti UI moraju odgovarati našima,
14:45
not just in the familiar context,
262
873801
1760
ne samo u poznatim kontekstima
14:47
like where we can easily check
how the A.I. behaves,
how the A.I. behaves,
263
875561
2438
u kojima lako provjerimo
kako se UI ponaša,
kako se UI ponaša,
14:49
but also in all novel contexts
that the A.I. might encounter
that the A.I. might encounter
264
877999
3234
nego i u novim kontekstima s
kojima bi se UI mogla susresti
kojima bi se UI mogla susresti
14:53
in the indefinite future.
265
881233
1557
u neodređenoj budućnosti.
14:54
And there are also some esoteric issues
that would need to be solved, sorted out:
that would need to be solved, sorted out:
266
882790
4737
Postoje i neka ezoterična pitanja
koja bi se morala riješiti:
koja bi se morala riješiti:
14:59
the exact details of its decision theory,
267
887527
2089
točni detalji teorije odlučivanja,
15:01
how to deal with logical
uncertainty and so forth.
uncertainty and so forth.
268
889616
2864
kako se nositi s logičkom
nesigurnosti i tako dalje.
nesigurnosti i tako dalje.
15:05
So the technical problems that need
to be solved to make this work
to be solved to make this work
269
893330
3102
Tehnički problemi koji se moraju
riješiti kako bi ovo uspjelo
riješiti kako bi ovo uspjelo
15:08
look quite difficult --
270
896432
1113
poprilično su veliki --
15:09
not as difficult as making
a superintelligent A.I.,
a superintelligent A.I.,
271
897545
3380
ne toliko veliki kao stvaranje
superinteligentne UI,
superinteligentne UI,
15:12
but fairly difficult.
272
900925
2868
ali prilično veliki.
15:15
Here is the worry:
273
903793
1695
Ovo je razlog za brigu:
15:17
Making superintelligent A.I.
is a really hard challenge.
is a really hard challenge.
274
905488
4684
Stvaranje superinteligentne UI
težak je izazov.
težak je izazov.
15:22
Making superintelligent A.I. that is safe
275
910172
2548
Stvaranje sigurne superinteligentne UI
15:24
involves some additional
challenge on top of that.
challenge on top of that.
276
912720
2416
dodaje još izazova.
15:28
The risk is that if somebody figures out
how to crack the first challenge
how to crack the first challenge
277
916216
3487
Rizik leži u otkrivanju kako
savladati prvi izazov,
savladati prvi izazov,
15:31
without also having cracked
the additional challenge
the additional challenge
278
919703
3001
ali bez shvaćanja kako savladati
dodatni izazov,
dodatni izazov,
15:34
of ensuring perfect safety.
279
922704
1901
osiguravanje sigurnosti.
15:37
So I think that we should
work out a solution
work out a solution
280
925375
3331
Smatram da bismo
trebali riješiti
trebali riješiti
15:40
to the control problem in advance,
281
928706
2822
pitanje kontrole unaprijed
15:43
so that we have it available
by the time it is needed.
by the time it is needed.
282
931528
2660
kako bismo bili spremni
kada za to dođe vrijeme.
kada za to dođe vrijeme.
15:46
Now it might be that we cannot solve
the entire control problem in advance
the entire control problem in advance
283
934768
3507
Može se dogoditi da ne možemo
taj problem riješiti unaprijed
taj problem riješiti unaprijed
15:50
because maybe some elements
can only be put in place
can only be put in place
284
938275
3024
jer se mogu posložiti samo neki elementi
15:53
once you know the details of the
architecture where it will be implemented.
architecture where it will be implemented.
285
941299
3997
kada točno saznamo gdje će se provoditi.
15:57
But the more of the control problem
that we solve in advance,
that we solve in advance,
286
945296
3380
Što više problema po pitanju
kontrole riješimo unaprijed,
kontrole riješimo unaprijed,
16:00
the better the odds that the transition
to the machine intelligence era
to the machine intelligence era
287
948676
4090
bolji su izgledi da će prijelaz
u eru strojne inteligencije
u eru strojne inteligencije
16:04
will go well.
288
952766
1540
proći dobro.
16:06
This to me looks like a thing
that is well worth doing
that is well worth doing
289
954306
4644
Meni se čini da se to isplati raditi
16:10
and I can imagine that if
things turn out okay,
things turn out okay,
290
958950
3332
i mogu zamisliti da, ako
sve prođe u redu,
sve prođe u redu,
16:14
that people a million years from now
look back at this century
look back at this century
291
962282
4658
će ljudi za milijun godina
gledati na ovo stoljeće
gledati na ovo stoljeće
16:18
and it might well be that they say that
the one thing we did that really mattered
the one thing we did that really mattered
292
966940
4002
i reći da smo napravili jednu stvar
koja je stvarno važna, a ta je
koja je stvarno važna, a ta je
16:22
was to get this thing right.
293
970942
1567
da smo ovo napravili kako treba.
16:24
Thank you.
294
972509
1689
Hvala.
16:26
(Applause)
295
974198
2813
(Pljesak)
ABOUT THE SPEAKER
Nick Bostrom - PhilosopherNick Bostrom asks big questions: What should we do, as individuals and as a species, to optimize our long-term prospects? Will humanity’s technological advancements ultimately destroy us?
Why you should listen
Philosopher Nick Bostrom envisioned a future full of human enhancement, nanotechnology and machine intelligence long before they became mainstream concerns. From his famous simulation argument -- which identified some striking implications of rejecting the Matrix-like idea that humans are living in a computer simulation -- to his work on existential risk, Bostrom approaches both the inevitable and the speculative using the tools of philosophy, probability theory, and scientific analysis.
Since 2005, Bostrom has led the Future of Humanity Institute, a research group of mathematicians, philosophers and scientists at Oxford University tasked with investigating the big picture for the human condition and its future. He has been referred to as one of the most important thinkers of our age.
Nick was honored as one of Foreign Policy's 2015 Global Thinkers .
His recent book Superintelligence advances the ominous idea that “the first ultraintelligent machine is the last invention that man need ever make.”
Nick Bostrom | Speaker | TED.com