TEDGlobal 2007
Kwabena Boahen: A computer that works like the brain
Kwabena Boahen o računalu koje funkcionira poput mozga
Filmed:
Readability: 4.2
718,375 views
Istraživač Kwabena Boahen traži načine kako oponašati super računalne moći mozga u silikonu -- jer zbrkani, suvišni procesi u našim glavama zapravo čine malo, lagano, super brzo računalo.
Kwabena Boahen - Bioengineer
Kwabena Boahen wants to understand how brains work -- and to build a computer that works like the brain by reverse-engineering the nervous system. His group at Stanford is developing Neurogrid, a hardware platform that will emulate the cortex’s inner workings. Full bio
Kwabena Boahen wants to understand how brains work -- and to build a computer that works like the brain by reverse-engineering the nervous system. His group at Stanford is developing Neurogrid, a hardware platform that will emulate the cortex’s inner workings. Full bio
Double-click the English transcript below to play the video.
00:18
I got my first computer when I was a teenager growing up in Accra,
0
0
5000
Dobio sam svoje prvo računalo kada sam bio tinejdžer i odrastao u Akri
00:23
and it was a really cool device.
1
5000
3000
i to je bila zbilja cool naprava.
00:26
You could play games with it. You could program it in BASIC.
2
8000
5000
Mogao si igrati igre na njemu. Mogao si programirati u BASIC-u.
00:31
And I was fascinated.
3
13000
2000
I bio sam fasciniran.
00:33
So I went into the library to figure out how did this thing work.
4
15000
6000
Dakle, otišao sam u knjižnicu da shvatim kako ta stvar radi.
00:39
I read about how the CPU is constantly shuffling data back and forth
5
21000
5000
Čitao sam kako CPU stalno prebacuje podatke naprijed natrag
00:44
between the memory, the RAM and the ALU,
6
26000
4000
kroz memoriju, o RAM-u i o ALU-u,
00:48
the arithmetic and logic unit.
7
30000
2000
o aritmetičkoj i logaritamskoj jedinici.
00:50
And I thought to myself, this CPU really has to work like crazy
8
32000
4000
I mislio sam si: ovaj CPU zbilja mora raditi kao lud
00:54
just to keep all this data moving through the system.
9
36000
4000
samo da održi sve te podatke koji se pomiču kroz sistem.
00:58
But nobody was really worried about this.
10
40000
3000
Ali, nitko se zapravo nije brinuo oko toga.
01:01
When computers were first introduced,
11
43000
2000
Kada su računala prvi puta predstavljena,
01:03
they were said to be a million times faster than neurons.
12
45000
3000
rečeno je da su milijun puta brži od neurona.
01:06
People were really excited. They thought they would soon outstrip
13
48000
5000
Ljudi su zbilja bili uzbuđeni. Mislili su da će uskoro nadmašiti
01:11
the capacity of the brain.
14
53000
3000
kapacitete mozga.
01:14
This is a quote, actually, from Alan Turing:
15
56000
3000
Ovo je zapravo citat Alana Turinga:
01:17
"In 30 years, it will be as easy to ask a computer a question
16
59000
4000
„Za 30 godina biti će jednako lako postaviti računalu pitanje
01:21
as to ask a person."
17
63000
2000
kao što je pitati čovjeka.“
01:23
This was in 1946. And now, in 2007, it's still not true.
18
65000
7000
To je bilo 1946. I sada, u 2007. To još uvjek nije točno.
01:30
And so, the question is, why aren't we really seeing
19
72000
4000
Dakle, pitanje jest, zašto ne vidimo
01:34
this kind of power in computers that we see in the brain?
20
76000
4000
istu vrstu moći u računalima kakvu vidimo u mozgu?
01:38
What people didn't realize, and I'm just beginning to realize right now,
21
80000
4000
Ono što ljudi nisu shvatili, a ja tek sad počinjem shvaćati,
01:42
is that we pay a huge price for the speed
22
84000
2000
jest da plaćamo golemu cijenu za brzinu
01:44
that we claim is a big advantage of these computers.
23
86000
4000
koja je navodno velika prednost tim računalima.
01:48
Let's take a look at some numbers.
24
90000
2000
Pogledajmo neke brojeve.
01:50
This is Blue Gene, the fastest computer in the world.
25
92000
4000
Ovo je Blue Gene, najbrže računalo na svijetu.
01:54
It's got 120,000 processors; they can basically process
26
96000
5000
Ima 120.000 procesora, oni u principu mogu procesuirati
01:59
10 quadrillion bits of information per second.
27
101000
3000
10 kvadrilijon bita informacije po sekundi.
02:02
That's 10 to the sixteenth. And they consume one and a half megawatts of power.
28
104000
7000
To je 10 na šesnaestu. I oni troše jedan i pol megavata snage.
02:09
So that would be really great, if you could add that
29
111000
3000
To bi bilo zbilja dobro, dodati to kapacitetu
02:12
to the production capacity in Tanzania.
30
114000
2000
proizvodnje Tanzanije.
02:14
It would really boost the economy.
31
116000
2000
To bi im povećalo proizvodnju.
02:16
Just to go back to the States,
32
118000
4000
Samo da se vratimo u SAD,
02:20
if you translate the amount of power or electricity
33
122000
2000
ako prevedete količinu elektriciteta
02:22
this computer uses to the amount of households in the States,
34
124000
3000
koji ovo računalo koristi na količini kućanstava u SAD-u,
02:25
you get 1,200 households in the U.S.
35
127000
4000
dobijete 1.200 kućanstava.
02:29
That's how much power this computer uses.
36
131000
2000
Toliko koristi ovo računalo.
02:31
Now, let's compare this with the brain.
37
133000
3000
Sada, usporedimo to s mozgom.
02:34
This is a picture of, actually Rory Sayres' girlfriend's brain.
38
136000
5000
Ovo je zapravo slika mozga cure Roryja Sayersa.
02:39
Rory is a graduate student at Stanford.
39
141000
2000
Rory je apsolvent na Stanfordu.
02:41
He studies the brain using MRI, and he claims that
40
143000
4000
On proučava mozak koristeći magnetsku rezonanciju, i tvrdi
02:45
this is the most beautiful brain that he has ever scanned.
41
147000
3000
da je ovo najljepši mozak koji je ikada skenirao.
02:48
(Laughter)
42
150000
2000
(Smijeh)
02:50
So that's true love, right there.
43
152000
3000
Eto, to je prava ljubav.
02:53
Now, how much computation does the brain do?
44
155000
3000
Koliko izračuna radi mozak?
02:56
I estimate 10 to the 16 bits per second,
45
158000
2000
Ja procjenjujem 10 na šesnaestu bita po sekundi,
02:58
which is actually about very similar to what Blue Gene does.
46
160000
4000
što je otprilike slično koliko i Blue Gene.
03:02
So that's the question. The question is, how much --
47
164000
2000
Dakle, to je pitanje. Pitanje je, koliko --
03:04
they are doing a similar amount of processing, similar amount of data --
48
166000
3000
oni rade sličnu količinu procesiranja, sličnu količinu podataka --
03:07
the question is how much energy or electricity does the brain use?
49
169000
5000
pitanje je koliko energije ili elektriciteta mozak koristi?
03:12
And it's actually as much as your laptop computer:
50
174000
3000
I, to je zapravo onoliko koliko troši vaš laptop:
03:15
it's just 10 watts.
51
177000
2000
to je samo 10 W.
03:17
So what we are doing right now with computers
52
179000
3000
Dakle, što radimo s računalima
03:20
with the energy consumed by 1,200 houses,
53
182000
3000
koji troše energije kao 1.200 kućanstava,
03:23
the brain is doing with the energy consumed by your laptop.
54
185000
5000
mozak radi s utroškom energije koji ima vaš laptop.
03:28
So the question is, how is the brain able to achieve this kind of efficiency?
55
190000
3000
Pitanje jest kako mozak uspjeva postići ovu razinu učinkovitosti?
03:31
And let me just summarize. So the bottom line:
56
193000
2000
Dozvolite mi da sažmem. Na kraju krajeva,
03:33
the brain processes information using 100,000 times less energy
57
195000
4000
mozak procesira informacije koristeći 100.000 puta manje energije
03:37
than we do right now with this computer technology that we have.
58
199000
4000
nego što mi trošimo s ovom računalnom tehnologijom.
03:41
How is the brain able to do this?
59
203000
2000
Kako to mozak uspijeva?
03:43
Let's just take a look about how the brain works,
60
205000
3000
Pogledajmo samo način na koji mozak radi,
03:46
and then I'll compare that with how computers work.
61
208000
4000
i onda ću to usporediti s radom računala.
03:50
So, this clip is from the PBS series, "The Secret Life of the Brain."
62
212000
4000
Ovo je video iz PBS-ove serije „Tajni život Mozga“.
03:54
It shows you these cells that process information.
63
216000
3000
Pokazuje vam stanice koje procesiraju informacije.
03:57
They are called neurons.
64
219000
1000
One se nazivaju neuroni.
03:58
They send little pulses of electricity down their processes to each other,
65
220000
6000
Oni odašilju male pulsove elektriciteta niz njihove procesore jedne drugima
04:04
and where they contact each other, those little pulses
66
226000
2000
i na mjestima gdje se dodiruju ovi mali impulsi
04:06
of electricity can jump from one neuron to the other.
67
228000
2000
mogu skočiti s jednog neurona na drugi.
04:08
That process is called a synapse.
68
230000
3000
Taj je proces nazvan sinapsa.
04:11
You've got this huge network of cells interacting with each other --
69
233000
2000
Imate ovu golemu mrežu stanica koje vrše interakcije jedne s drugima --
04:13
about 100 million of them,
70
235000
2000
oko 100 milijuna njih,
04:15
sending about 10 quadrillion of these pulses around every second.
71
237000
4000
koje šalju oko 10 kvadrilijuna pulseva svake sekunde.
04:19
And that's basically what's going on in your brain right now as you're watching this.
72
241000
6000
I to je otprilike što se događa u vašem mozgu sada dok ovo gledate.
04:25
How does that compare with the way computers work?
73
247000
2000
Kako se to može usporediti s načinom na koji radi računalo?
04:27
In the computer, you have all the data
74
249000
2000
U računalu svi podaci
04:29
going through the central processing unit,
75
251000
2000
prolaze kroz centralnu procesorsku jedinicu,
04:31
and any piece of data basically has to go through that bottleneck,
76
253000
3000
i svaki djelić podatka u osnovi mora proći kroz to usko grlo,
04:34
whereas in the brain, what you have is these neurons,
77
256000
4000
dok u mozgu imate neurone,
04:38
and the data just really flows through a network of connections
78
260000
4000
i podatci jednostavno teku kroz mrežu spojeva
04:42
among the neurons. There's no bottleneck here.
79
264000
2000
među neuronima. Nema uskog grla.
04:44
It's really a network in the literal sense of the word.
80
266000
4000
To je zaista mreža u doslovnom smislu riječi.
04:48
The net is doing the work in the brain.
81
270000
4000
Mreža radi posao za vaš mozak.
04:52
If you just look at these two pictures,
82
274000
2000
I ako samo pogledate ove dvije slike,
04:54
these kind of words pop into your mind.
83
276000
2000
ove riječi uskaču u vaš mozak.
04:56
This is serial and it's rigid -- it's like cars on a freeway,
84
278000
4000
Ovo je serijski spojeno, i kruto je, kao auti na autocesti,
05:00
everything has to happen in lockstep --
85
282000
3000
i sve se mora dogoditi u pravilnom razmaku --
05:03
whereas this is parallel and it's fluid.
86
285000
2000
dok je ovo paralelno i fluidno.
05:05
Information processing is very dynamic and adaptive.
87
287000
3000
Procesiranje informacija je jako dinamično i prilagodljivo.
05:08
So I'm not the first to figure this out. This is a quote from Brian Eno:
88
290000
4000
Nisam prvi koji je to shvatio. Ovo je citat Briana Ena:
05:12
"the problem with computers is that there is not enough Africa in them."
89
294000
4000
„Problem s računalima jest to što u njima nema dosta Afrike“.
05:16
(Laughter)
90
298000
6000
(Smijeh)
05:22
Brian actually said this in 1995.
91
304000
3000
Brian je zapravo to rekao 1995.
05:25
And nobody was listening then,
92
307000
3000
i onda ga nitko nije slušao,
05:28
but now people are beginning to listen
93
310000
2000
ali sada ljudi počinju slušati
05:30
because there's a pressing, technological problem that we face.
94
312000
5000
jer postoji veliki tehnološki problem s kojim se suočavamo.
05:35
And I'll just take you through that a little bit in the next few slides.
95
317000
5000
I sada ću vas pomalo provesti kroz to u sljedećih nekoliko slajdova.
05:40
This is -- it's actually really this remarkable convergence
96
322000
4000
Ovo je zapravo izvanredna konvergencija
05:44
between the devices that we use to compute in computers,
97
326000
5000
između naprava koje koristimo da računaju u računalima
05:49
and the devices that our brains use to compute.
98
331000
4000
i naprava koje koriste naši mozgovi.
05:53
The devices that computers use are what's called a transistor.
99
335000
4000
Naprava koju računala koriste naziva se tranzistor.
05:57
This electrode here, called the gate, controls the flow of current
100
339000
4000
Ova elektroda ovdje -- naziva se prekidač i kontrolira tok struje
06:01
from the source to the drain -- these two electrodes.
101
343000
3000
od izvora do potrošača -- ovih dviju elektroda.
06:04
And that current, electrical current,
102
346000
2000
A struja -- električna struja --
06:06
is carried by electrons, just like in your house and so on.
103
348000
6000
je nošena elektronima baš kao u vašoj kući i tako dalje.
06:12
And what you have here is, when you actually turn on the gate,
104
354000
5000
Ono što imamo ovdje jest, kada upalimo prekidač,
06:17
you get an increase in the amount of current, and you get a steady flow of current.
105
359000
4000
povećavamo količinu struje i dobivamo stalan tok struje.
06:21
And when you turn off the gate, there's no current flowing through the device.
106
363000
4000
A kada ugasimo prekidač, nema struje koja teče kroz napravu.
06:25
Your computer uses this presence of current to represent a one,
107
367000
5000
Vaše računalo koristi prisutnost struje da predstavlja jedinicu,
06:30
and the absence of current to represent a zero.
108
372000
4000
a nedostatak struje da predstavi nulu.
06:34
Now, what's happening is that as transistors are getting smaller and smaller and smaller,
109
376000
6000
Ono što se događa jest da tranzistori postaju sve manji, i manji, i manji
06:40
they no longer behave like this.
110
382000
2000
i više se ne ponašaju tako.
06:42
In fact, they are starting to behave like the device that neurons use to compute,
111
384000
5000
U stvari počnu se ponašati kao naprave koje neuroni koriste za računanje,
06:47
which is called an ion channel.
112
389000
2000
koji se nazivaju ionski kanali.
06:49
And this is a little protein molecule.
113
391000
2000
I ovo je mala molekula proteina.
06:51
I mean, neurons have thousands of these.
114
393000
4000
Mislim, neuroni imaju tisuće njih.
06:55
And it sits in the membrane of the cell and it's got a pore in it.
115
397000
4000
I oni leže u membrani stanice koja ima poru u sebi.
06:59
And these are individual potassium ions
116
401000
3000
Ovo su pojedini ioni kalija
07:02
that are flowing through that pore.
117
404000
2000
koji prolaze kroz poru.
07:04
Now, this pore can open and close.
118
406000
2000
Ove se pore mogu zatvoriti i otvoriti.
07:06
But, when it's open, because these ions have to line up
119
408000
5000
Ali, kada su otvorene, ioni prolaze jedan po jedan
07:11
and flow through, one at a time, you get a kind of sporadic, not steady --
120
413000
5000
i zato se moraju poredati da bi prošli, dobijete sporadičnu, neravnomjernu --
07:16
it's a sporadic flow of current.
121
418000
3000
to je sporadičan tok struje.
07:19
And even when you close the pore -- which neurons can do,
122
421000
3000
I čak i kad zatvorite poru -- što neuroni mogu napraviti,
07:22
they can open and close these pores to generate electrical activity --
123
424000
5000
oni mogu otvoriti i zatvoriti pore da bi stvorili električnu aktivnost --
07:27
even when it's closed, because these ions are so small,
124
429000
3000
čak i kad je zatvorena, zato što su ovi ioni toliko mali,
07:30
they can actually sneak through, a few can sneak through at a time.
125
432000
3000
mogu se prikrasti unutra, nekoliko se može prikrasti s vremena na vrijeme.
07:33
So, what you have is that when the pore is open,
126
435000
3000
Dakle, što imamo jest da, kad su pore otvorene
07:36
you get some current sometimes.
127
438000
2000
ponekad dobijemo struju.
07:38
These are your ones, but you've got a few zeros thrown in.
128
440000
3000
To su jedinice, ali dobijete i par nula ubačenih unutra.
07:41
And when it's closed, you have a zero,
129
443000
4000
A kad je zatvoreno, imate nule,
07:45
but you have a few ones thrown in.
130
447000
3000
ali imate i nekoliko jedinica.
07:48
Now, this is starting to happen in transistors.
131
450000
3000
E sada, ovo se počelo događati u tranzistorima.
07:51
And the reason why that's happening is that, right now, in 2007 --
132
453000
5000
I razlog zašto se to događa je što upravo sada, 2007. --
07:56
the technology that we are using -- a transistor is big enough
133
458000
4000
tehnologija koju koristimo, tranzistor, je dovoljno velik
08:00
that several electrons can flow through the channel simultaneously, side by side.
134
462000
5000
da nekoliko elektrona mogu proći kroz kanal istodobno, jedan pored drugoga.
08:05
In fact, there's about 12 electrons can all be flowing this way.
135
467000
4000
Zapravo, otprilike 12 elektrona mogu teći ovuda.
08:09
And that means that a transistor corresponds
136
471000
2000
I to znači da tranzistor odgovara
08:11
to about 12 ion channels in parallel.
137
473000
3000
otprilike 12 paralelnih ionskih kanala.
08:14
Now, in a few years time, by 2015, we will shrink transistors so much.
138
476000
5000
Za nekoliko godina, u 2015-oj smanjit ćemo elektrone za toliko.
08:19
This is what Intel does to keep adding more cores onto the chip.
139
481000
5000
Ovo Intel radi da bi dodao još jezgri na čip,
08:24
Or your memory sticks that you have now can carry one gigabyte
140
486000
3000
ili na USB memorije koje sada nose jedan gigabajt
08:27
of stuff on them -- before, it was 256.
141
489000
2000
stvari na njima -- prije je bilo samo 256 megabajta.
08:29
Transistors are getting smaller to allow this to happen,
142
491000
3000
Tranzistori postaju manji kako bi to omogućili,
08:32
and technology has really benefitted from that.
143
494000
3000
a tehnologija od toga zbilja profitira.
08:35
But what's happening now is that in 2015, the transistor is going to become so small,
144
497000
5000
Ali, ono što se sada događa jest da će 2015. tranzistori postati toliko mali
08:40
that it corresponds to only one electron at a time
145
502000
3000
da će odgovarati samo jednom elektronu
08:43
can flow through that channel,
146
505000
2000
koji može prolaziti kroz taj kanal,
08:45
and that corresponds to a single ion channel.
147
507000
2000
i to će odgovarati jednom ionskom kanalu.
08:47
And you start having the same kind of traffic jams that you have in the ion channel.
148
509000
4000
I počet ćemo imati iste prometne gužve kao i u ionskim kanalima.
08:51
The current will turn on and off at random,
149
513000
3000
Struja će se paliti i gasiti nasumice,
08:54
even when it's supposed to be on.
150
516000
2000
čak i kad bi trebala biti upaljena.
08:56
And that means your computer is going to get
151
518000
2000
A to znači da će računalo pobrkati
08:58
its ones and zeros mixed up, and that's going to crash your machine.
152
520000
4000
jedinice i nule i to će srušiti vaše računalo.
09:02
So, we are at the stage where we
153
524000
4000
Dakle, mi smo u fazi kada
09:06
don't really know how to compute with these kinds of devices.
154
528000
3000
ne znamo zapravo kako računati s takvim napravama.
09:09
And the only kind of thing -- the only thing we know right now
155
531000
3000
A jedina stvar -- jedina stvar za koju za sada znamo
09:12
that can compute with these kinds of devices are the brain.
156
534000
3000
da može raditi s takvom vrstom naprave, jest mozak.
09:15
OK, so a computer picks a specific item of data from memory,
157
537000
4000
Ok, znači računalo izabere određenu jedinicu podatka iz memorije,
09:19
it sends it into the processor or the ALU,
158
541000
3000
pošalje je u procesor ili ALU
09:22
and then it puts the result back into memory.
159
544000
2000
i onda vrati rezultat natrag u memoriju.
09:24
That's the red path that's highlighted.
160
546000
2000
To je crveno označeni put.
09:26
The way brains work, I told you all, you have got all these neurons.
161
548000
4000
Način na koji mozak radi, rekao sam vam, imate puno takvih neurona.
09:30
And the way they represent information is
162
552000
2000
I način na koji predstavljaju informacije je
09:32
they break up that data into little pieces
163
554000
2000
da podjele te podatke u male dijelove,
09:34
that are represented by pulses and different neurons.
164
556000
3000
koji su predstavljeni impulsima i drugim neuronima.
09:37
So you have all these pieces of data
165
559000
2000
I sada imate sve te dijelove podataka
09:39
distributed throughout the network.
166
561000
2000
podijeljene kroz mrežu.
09:41
And then the way that you process that data to get a result
167
563000
3000
I onda način na koji obrađujete te podatke kako biste dobili rezultate
09:44
is that you translate this pattern of activity into a new pattern of activity,
168
566000
4000
jest da prevedete taj uzorak aktivnosti u novi uzorak aktivnosti,
09:48
just by it flowing through the network.
169
570000
3000
prateći samo njegov tok kroz mrežu.
09:51
So you set up these connections
170
573000
2000
Dakle spojite te veze tako
09:53
such that the input pattern just flows
171
575000
3000
da kao ulazni uzorak samo teče
09:56
and generates the output pattern.
172
578000
2000
i stvara izlazni uzorak.
09:58
What you see here is that there's these redundant connections.
173
580000
4000
Ono što vidite ovdje su ove redundantne veze.
10:02
So if this piece of data or this piece of the data gets clobbered,
174
584000
4000
Dakle, ako ovaj ili onaj dio podatka postane izmiješan do neprepoznatljivosti
10:06
it doesn't show up over here, these two pieces can activate the missing part
175
588000
5000
onda se ne pojavljuje ovdje, i ova dva dijela mogu aktivirati dio koji nedostaje
10:11
with these redundant connections.
176
593000
2000
preko ovih redundantnih veza.
10:13
So even when you go to these crappy devices
177
595000
2000
Čak i ako kroz ove loše naprave
10:15
where sometimes you want a one and you get a zero, and it doesn't show up,
178
597000
3000
gdje ponekad želite dobiti jedan, a dobijete nulu i to se ne pokaže,
10:18
there's redundancy in the network
179
600000
2000
postoji redundancija u mreži
10:20
that can actually recover the missing information.
180
602000
3000
koja može vratiti izgubljene informacije.
10:23
It makes the brain inherently robust.
181
605000
3000
To čini mozak nevjerojatno postojanim.
10:26
What you have here is a system where you store data locally.
182
608000
3000
Ono što imamo ovdje jest sistem koji pohranjuje podatke lokalno.
10:29
And it's brittle, because each of these steps has to be flawless,
183
611000
4000
Krhak je, jer svaki korak mora biti savršen
10:33
otherwise you lose that data, whereas in the brain, you have a system
184
615000
3000
inače gubite podatke, dok u mozgu imate sistem
10:36
that stores data in a distributed way, and it's robust.
185
618000
4000
koji pohranjuje podatke na distribuirani način, i to ga čini postojanim.
10:40
What I want to basically talk about is my dream,
186
622000
4000
Ono o čemu zapravo želim pričati jest moj san,
10:44
which is to build a computer that works like the brain.
187
626000
3000
a to je da napravim računalo koje radi kao mozak.
10:47
This is something that we've been working on for the last couple of years.
188
629000
4000
To je nešto na čemu radimo zadnjih nekoliko godina.
10:51
And I'm going to show you a system that we designed
189
633000
3000
I pokazat ću vam sistem koji smo dizajnirali
10:54
to model the retina,
190
636000
3000
po modelu mrežnice,
10:57
which is a piece of brain that lines the inside of your eyeball.
191
639000
5000
a to je dio mozga koji oblaže unutrašnjost vaših očnih jabučica.
11:02
We didn't do this by actually writing code, like you do in a computer.
192
644000
6000
Nismo to napravili pišući kod, kao što to radite na računalu.
11:08
In fact, the processing that happens
193
650000
3000
Zapravo, procesiranje koje se zbiva
11:11
in that little piece of brain is very similar
194
653000
2000
u tom malom dijelu mozga jest vrlo slično
11:13
to the kind of processing that computers
195
655000
1000
procesiranju koje računala
11:14
do when they stream video over the Internet.
196
656000
4000
vrše dok šalju video preko interneta.
11:18
They want to compress the information --
197
660000
1000
Žele sažeti informacije --
11:19
they just want to send the changes, what's new in the image, and so on --
198
661000
4000
žele poslati samo promjene, što je novo na slici, i tako dalje --
11:23
and that is how your eyeball
199
665000
3000
a ovo je kako vaše oko
11:26
is able to squeeze all that information down to your optic nerve,
200
668000
3000
uspjeva sažeti sve te informacije kroz vidni živac
11:29
to send to the rest of the brain.
201
671000
2000
i poslati ih ostatku mozga.
11:31
Instead of doing this in software, or doing those kinds of algorithms,
202
673000
3000
Umjesto da ovo napravimo u softveru, ili da radimo algoritme,
11:34
we went and talked to neurobiologists
203
676000
3000
otišli smo i razgovarali s neurobiolozima
11:37
who have actually reverse engineered that piece of brain that's called the retina.
204
679000
4000
koji su zapravo sastavili to po modelu mrežnice.
11:41
And they figured out all the different cells,
205
683000
2000
I oni su uspjeli razumjeti sve te različite stanice,
11:43
and they figured out the network, and we just took that network
206
685000
3000
i oni su uspjeli razumijeti mrežu, a mi smo ju samo uzeli
11:46
and we used it as the blueprint for the design of a silicon chip.
207
688000
4000
kao nacrt za dizajn silikonskog čipa.
11:50
So now the neurons are represented by little nodes or circuits on the chip,
208
692000
6000
I sada su neuroni predstavljeni malim čvorićima, ili krugovima na čipu,
11:56
and the connections among the neurons are represented, actually modeled by transistors.
209
698000
5000
a spojevi između neurona su predstavljeni tranzistorima.
12:01
And these transistors are behaving essentially
210
703000
2000
I ovi tranzistori se u osnovi ponašaju
12:03
just like ion channels behave in the brain.
211
705000
3000
baš kao što se ponašaju ionski kanali u mozgu.
12:06
It will give you the same kind of robust architecture that I described.
212
708000
5000
To će vam dati istu vrstu postojane arhitekture koju sam opisao.
12:11
Here is actually what our artificial eye looks like.
213
713000
4000
Ovako zapravo naše umjetno oko izgleda.
12:15
The retina chip that we designed sits behind this lens here.
214
717000
5000
Mrežnični čip koji smo dizajnirali se nalazi ovdje iza leće.
12:20
And the chip -- I'm going to show you a video
215
722000
2000
I čip -- pokazati ću vam video
12:22
that the silicon retina put out of its output
216
724000
3000
koji silikonska mrežnica šalje kroz izlaznu jedinicu
12:25
when it was looking at Kareem Zaghloul,
217
727000
3000
kada gleda Kareema Zaghloula,
12:28
who's the student who designed this chip.
218
730000
2000
studenta koji je dizajnirao ovaj čip.
12:30
Let me explain what you're going to see, OK,
219
732000
2000
Dopustite mi da objasnim što će te vidjeti, OK,
12:32
because it's putting out different kinds of information,
220
734000
3000
zato što to pokazuje različite vrste informacija,
12:35
it's not as straightforward as a camera.
221
737000
2000
nije posve neposredno kao kamera.
12:37
The retina chip extracts four different kinds of information.
222
739000
3000
Mrežnični čip izvlači četiri različite vrste informacija.
12:40
It extracts regions with dark contrast,
223
742000
3000
Izvlači regije s tamnim kontrastom
12:43
which will show up on the video as red.
224
745000
3000
koje će se prikazati u ovom videu kao crvene.
12:46
And it extracts regions with white or light contrast,
225
748000
4000
I izvlači regije s bijelim ili svijetlim kontrastom
12:50
which will show up on the video as green.
226
752000
2000
koje će se prikazati kao zelene.
12:52
This is Kareem's dark eyes
227
754000
2000
Ovo je Kareemovo tamno oko,
12:54
and that's the white background that you see here.
228
756000
3000
a ovo je bijela pozadina koju vidite ovdje.
12:57
And then it also extracts movement.
229
759000
2000
I onda također izvlači pokrete.
12:59
When Kareem moves his head to the right,
230
761000
2000
Kada Kareem pomakne svoju glavu prema desno,
13:01
you will see this blue activity there;
231
763000
2000
vidite ovu plavu aktivnost ovdje.
13:03
it represents regions where the contrast is increasing in the image,
232
765000
3000
To predstavlja regije gdje se kontrast na slici povećava,
13:06
that's where it's going from dark to light.
233
768000
3000
gdje prelazi iz tamnog u svijetlo.
13:09
And you also see this yellow activity,
234
771000
2000
I vidite ovu žutu aktivnost,
13:11
which represents regions where contrast is decreasing;
235
773000
4000
koja predstavlja regije gdje se kontrast smanjuje,
13:15
it's going from light to dark.
236
777000
2000
ide od svijetlog prema tamnom.
13:17
And these four types of information --
237
779000
3000
I ove četiri vrste informacija --
13:20
your optic nerve has about a million fibers in it,
238
782000
4000
vaš optički živac ima oko milijun vlakana,
13:24
and 900,000 of those fibers
239
786000
3000
a 900.000 od njih
13:27
send these four types of information.
240
789000
2000
šalju ove četiri vrste informacija.
13:29
So we are really duplicating the kind of signals that you have on the optic nerve.
241
791000
4000
Dakle, zapravo dupliciramo ove vrste signala optičkim živcem.
13:33
What you notice here is that these snapshots
242
795000
3000
Primjetit ćete da su ove snimke
13:36
taken from the output of the retina chip are very sparse, right?
243
798000
4000
uzete iz izlazne jedinice mrežnice vrlo oskudne, zar ne?
13:40
It doesn't light up green everywhere in the background,
244
802000
2000
Zeleno se ne pojavljuje svuda na pozadini,
13:42
only on the edges, and then in the hair, and so on.
245
804000
3000
samo na rubovima, i u kosi, i tako dalje.
13:45
And this is the same thing you see
246
807000
1000
Istu stvar imate
13:46
when people compress video to send: they want to make it very sparse,
247
808000
4000
kada ljudi sažimlju videe kako bi ih mogli slati. Žele da budu što oskudnije
13:50
because that file is smaller. And this is what the retina is doing,
248
812000
3000
kako bi datoteka bila što manja. To isto radi i mrežnica,
13:53
and it's doing it just with the circuitry, and how this network of neurons
249
815000
4000
samo sa sklopovima, i tako radi i ova mreža neurona
13:57
that are interacting in there, which we've captured on the chip.
250
819000
3000
koji vrše interakciju koju smo mi ugradili u čip.
14:00
But the point that I want to make -- I'll show you up here.
251
822000
3000
Ali ono što želim reći jest -- pokazat ću vam to ovdje.
14:03
So this image here is going to look like these ones,
252
825000
3000
Dakle, ova slika ovdje će izgledati kao ove,
14:06
but here I'll show you that we can reconstruct the image,
253
828000
2000
ali ovdje ću vam pokazati kako možemo rekonstruirati slike
14:08
so, you know, you can almost recognize Kareem in that top part there.
254
830000
5000
tako da, znate, skoro možete prepoznati Kareema u gornjem djelu.
14:13
And so, here you go.
255
835000
2000
Tako.
14:24
Yes, so that's the idea.
256
846000
3000
Da, dakle to je ideja.
14:27
When you stand still, you just see the light and dark contrasts.
257
849000
2000
Kada stojite mirno, vidite samo svijetle i tamne kontraste.
14:29
But when it's moving back and forth,
258
851000
2000
Ali kada se miče naprijed i natrag
14:31
the retina picks up these changes.
259
853000
3000
mrežnica primjećuje te promjene.
14:34
And that's why, you know, when you're sitting here
260
856000
1000
I to je odgovor na zašto samo pomaknete oči
14:35
and something happens in your background,
261
857000
2000
kada sjedite ovdje i nešto
14:37
you merely move your eyes to it.
262
859000
2000
se dogodi u pozadini.
14:39
There are these cells that detect change
263
861000
2000
Postoje stanice koje primjećuju te promjene
14:41
and you move your attention to it.
264
863000
2000
i skreću vam pozornost na to.
14:43
So those are very important for catching somebody
265
865000
2000
Dakle, to je jako bitno kako bi uhvatili nekoga
14:45
who's trying to sneak up on you.
266
867000
2000
tko vam se pokušava prikrasti.
14:47
Let me just end by saying that this is what happens
267
869000
3000
Dopustite mi da završim tako da kažem da je ovo ono što se dogodi
14:50
when you put Africa in a piano, OK.
268
872000
3000
kada stavite Afriku u klavir. OK.
14:53
This is a steel drum here that has been modified,
269
875000
3000
Ovo je modificirani čelični bubanj,
14:56
and that's what happens when you put Africa in a piano.
270
878000
3000
i to se dogodi kada stavite Afriku u klavir.
14:59
And what I would like us to do is put Africa in the computer,
271
881000
4000
A ono što bih sada volio da napravite jest da stavite Afriku u računala
15:03
and come up with a new kind of computer
272
885000
2000
i smislite novu vrstu računala
15:05
that will generate thought, imagination, be creative and things like that.
273
887000
3000
koje će generirati misao, maštu, biti kreativno i raditi takve stvari.
15:08
Thank you.
274
890000
2000
Hvala vam.
15:10
(Applause)
275
892000
2000
(Pljesak)
15:12
Chris Anderson: Question for you, Kwabena.
276
894000
2000
Chris Anderson: Pitanje za tebe, Kwabena.
15:14
Do you put together in your mind the work you're doing,
277
896000
4000
Spajaš li u svom umu posao koji radiš,
15:18
the future of Africa, this conference --
278
900000
3000
budućnost Afrike, ovu konferenciju --
15:21
what connections can we make, if any, between them?
279
903000
3000
kakve veze mi imamo, ako ikakve veze ima među njima?
15:24
Kwabena Boahen: Yes, like I said at the beginning,
280
906000
2000
Kwabena Boahen: Da, kao što sam rekao na početku,
15:26
I got my first computer when I was a teenager, growing up in Accra.
281
908000
4000
dobio sam svoje prvo računalo kao tinejdžer, odrastajući u Akri.
15:30
And I had this gut reaction that this was the wrong way to do it.
282
912000
4000
I imao sam predosjećaj da je to krivi način.
15:34
It was very brute force; it was very inelegant.
283
916000
3000
Bila je to vrlo gruba sila, nije bilo elegantno.
15:37
I don't think that I would've had that reaction,
284
919000
2000
Mislim da ne bih imao taj osjećaj
15:39
if I'd grown up reading all this science fiction,
285
921000
3000
da sam odrastao čitajući svu tu znanstvenu fantastiku,
15:42
hearing about RD2D2, whatever it was called, and just -- you know,
286
924000
4000
slušajući o R2D2, kako ga već zovu, i samo -- znaš,
15:46
buying into this hype about computers.
287
928000
1000
živio u toj pomami za računalima.
15:47
I was coming at it from a different perspective,
288
929000
2000
Ja sam to doživio iz druge perspektive,
15:49
where I was bringing that different perspective
289
931000
2000
i donosim tu perspektivu
15:51
to bear on the problem.
290
933000
2000
kako bi se nosio s problemom.
15:53
And I think a lot of people in Africa have this different perspective,
291
935000
3000
I mislim da puno ljudi u Africi ima tu drugačiju perspektivu,
15:56
and I think that's going to impact technology.
292
938000
2000
i mislim da će to utjecati na tehnologiju.
15:58
And that's going to impact how it's going to evolve.
293
940000
2000
I da će to utjecati na način na koji mi evoluiramo.
16:00
And I think you're going to be able to see, use that infusion,
294
942000
2000
I mislim da ćete moći vidjeti, koritstiti tu infuziju
16:02
to come up with new things,
295
944000
2000
kako bi došli do novih stvari
16:04
because you're coming from a different perspective.
296
946000
3000
jer one dolaze iz drugog kuta.
16:07
I think we can contribute. We can dream like everybody else.
297
949000
4000
Mislim da možemo pridonijeti. I mi možemo sanjati kao svi drugi.
16:11
CA: Thanks Kwabena, that was really interesting.
298
953000
2000
CA: Hvala Kwabena, to je bilo zaista zanimljivo.
16:13
Thank you.
299
955000
1000
Hvala.
16:14
(Applause)
300
956000
2000
(Pljesak)
ABOUT THE SPEAKER
Kwabena Boahen - BioengineerKwabena Boahen wants to understand how brains work -- and to build a computer that works like the brain by reverse-engineering the nervous system. His group at Stanford is developing Neurogrid, a hardware platform that will emulate the cortex’s inner workings.
Why you should listen
Kwabena Boahen is the principal investigator at the Brains in Silicon lab at Stanford. He writes of himself:
Being a scientist at heart, I want to understand how cognition arises from neuronal properties. Being an engineer by training, I am using silicon integrated circuits to emulate the way neurons compute, linking the seemingly disparate fields of electronics and computer science with neurobiology and medicine.
My group's contributions to the field of neuromorphic engineering include a silicon retina that could be used to give the blind sight and a self-organizing chip that emulates the way the developing brain wires itself up. Our work is widely recognized, with over sixty publications, including a cover story in the May 2005 issue of Scientific American.
My current research interest is building a simulation platform that will enable the cortex's inner workings to be modeled in detail. While progress has been made linking neuronal properties to brain rhythms, the task of scaling up these models to link neuronal properties to cognition still remains. Making the supercomputer-performance required affordable is the goal of our Neurogrid project. It is at the vanguard of a profound shift in computing, away from the sequential, step-by-step Von Neumann machine towards a parallel, interconnected architecture more like the brain.
Kwabena Boahen | Speaker | TED.com