Kwabena Boahen: Making a computer that works like the brain

96,468 views ใƒป 2008-07-30

TED


ืื ื ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ืœืžื˜ื” ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ.

ืžืชืจื’ื: Roy Freifeld ืžื‘ืงืจ: Sigal Tifferet
00:18
I got my first computer when I was a teenager growing up in Accra,
0
18330
5000
ืงื™ื‘ืœืชื™ ืืช ื”ืžื—ืฉื‘ ื”ืจืืฉื•ืŸ ืฉืœื™ ื›ืฉื”ื™ื™ืชื™ ื ืขืจ ืžืชื‘ื’ืจ ื‘ืืงืจื”,
00:23
and it was a really cool device.
1
23330
3000
ื•ื–ื” ื”ื™ื” ืžื›ืฉื™ืจ ืžืื•ื“ ืžื’ื ื™ื‘
00:26
You could play games with it. You could program it in BASIC.
2
26330
5000
ื™ื›ื•ืœืช ืœืฉื—ืง ืื™ืชื• ืžืฉื—ืงื™ื, ื™ื›ื•ืœืช ืœืชื›ื ืช ืื•ืชื• ื‘"ื‘ื™ื™ืกื™ืง".
00:31
And I was fascinated.
3
31330
2000
ื•ื”ื™ื™ืชื™ ืžืจื•ืชืง.
00:33
So I went into the library to figure out how did this thing work.
4
33330
6000
ืื– ื”ืœื›ืชื™ ืœืกืคืจื™ื” ืœื”ื‘ื™ืŸ ืื™ืš ื”ื“ื‘ืจ ื”ื–ื” ืขื•ื‘ื“.
00:39
I read about how the CPU is constantly shuffling data back and forth
5
39330
5000
ืงืจืืชื™ ืขืœ ืื™ืš ื”ืžืขื‘ื“ ืžืชืžืจืŸ ืžื™ื“ืข ื”ืœื•ืš ื•ืฉื•ื‘ ื›ืœ ื”ื–ืžืŸ
00:44
between the memory, the RAM and the ALU,
6
44330
4000
ื‘ื™ืŸ ื”ื“ื™ืกืง, ื”ื–ื™ื›ืจื•ืŸ ื•ื” ALU,
00:48
the arithmetic and logic unit.
7
48330
2000
ื™ื—ื™ื“ืช ื”ืขื™ื‘ื•ื“ ื”ืžืชืžื˜ื™ืช.
00:50
And I thought to myself, this CPU really has to work like crazy
8
50330
4000
ื•ื—ืฉื‘ืชื™ ืœืขืฆืžื™, ื”ืžืขื‘ื“ ื”ื–ื” ืฆืจื™ืš ืœืขื‘ื•ื“ ื›ืžื• ืžืฉื•ื’ืข
00:54
just to keep all this data moving through the system.
9
54330
4000
ืจืง ื‘ืฉื‘ื™ืœ ืœืฉืžื•ืจ ืขืœ ื›ืœ ื”ืžื™ื“ืข ื”ื–ื” ื–ื•ืจื ื‘ืžืขืจื›ืช.
00:58
But nobody was really worried about this.
10
58330
3000
ืื‘ืœ ืืฃ ืื—ื“ ืœื ื”ื™ื” ืžื•ื“ืื’ ืžื–ื”.
01:01
When computers were first introduced,
11
61330
2000
ื›ืฉื”ืžื—ืฉื‘ื™ื ื”ื•ืฆื’ื• ืœืจืืฉื•ื ื”,
01:03
they were said to be a million times faster than neurons.
12
63330
3000
ื ืืžืจ ืฉื”ื ืžื”ื™ืจื™ื ืคื™ ืžื™ืœื™ื•ืŸ ืžื ื™ื•ืจื•ื ื™ื.
01:06
People were really excited. They thought they would soon outstrip
13
66330
5000
ืื ืฉื™ื ืžืžืฉ ื”ืชืจื’ืฉื•, ื”ื ื—ืฉื‘ื• ืฉื”ื ื‘ืžื”ืจื” ื™ืชืขืœื•
01:11
the capacity of the brain.
14
71330
3000
ืขืœ ื”ื™ื›ื•ืœืช ืฉืœ ื”ืžื•ื—.
01:14
This is a quote, actually, from Alan Turing:
15
74330
3000
ื–ื”ื• ืฆื™ื˜ื•ื˜, ืœืžืขืŸ ื”ืืžืช, ืฉืœ ืืœืŸ ื˜ื™ื•ืจื™ื ื’:
01:17
"In 30 years, it will be as easy to ask a computer a question
16
77330
4000
"ื‘ืขื•ื“ 30 ืฉื ื”, ื™ื”ื™ื” ืงืœ ืœืฉืื•ืœ ืืช ื”ืžื—ืฉื‘ ืฉืืœื” ื›ืžื• ืฉืงืœ
01:21
as to ask a person."
17
81330
2000
ืœืฉืื•ืœ ืื“ื.
01:23
This was in 1946. And now, in 2007, it's still not true.
18
83330
7000
ื–ื” ื”ื™ื” ื‘-1946. ื•ื”ื™ื•ื ื‘-2007, ื–ื” ืขื“ื™ื™ืŸ ืœื ื ื›ื•ืŸ.
01:30
And so, the question is, why aren't we really seeing
19
90330
4000
ื”ืฉืืœื” ื”ื™ื, ืœืžื” ืื ื—ื ื• ืœื ืจื•ืื™ื
01:34
this kind of power in computers that we see in the brain?
20
94330
4000
ื‘ืžื—ืฉื‘ื™ื ืืช ืื•ืชื” ื”ืขื•ืฆืžื” ืฉืื ื—ื ื• ืจื•ืื™ื ื‘ืžื•ื—?
01:38
What people didn't realize, and I'm just beginning to realize right now,
21
98330
4000
ืžื” ืฉืื ืฉื™ื ืœื ื”ื‘ื™ื ื•, ื•ืื ื™ ืžืชื—ื™ืœ ืœื”ื‘ื™ืŸ ืขื›ืฉื™ื•,
01:42
is that we pay a huge price for the speed
22
102330
2000
ื–ื” ืฉืื ื—ื ื• ืžืฉืœืžื™ื ืžื›ื™ืจ ื›ื‘ื“ ื‘ืฉื‘ื™ืœ ื”ืžื”ื™ืจื•ืช,
01:44
that we claim is a big advantage of these computers.
23
104330
4000
ืฉืื ื—ื ื• ื˜ื•ืขื ื™ื ืฉื”ื™ื ื”ื™ืชืจื•ืŸ ื”ื’ื“ื•ืœ ืฉืœ ื”ืžื—ืฉื‘ื™ื.
01:48
Let's take a look at some numbers.
24
108330
2000
ื‘ื•ืื• ื ืกืชื›ืœ ืขืœ ื›ืžื” ืžืกืคืจื™ื.
01:50
This is Blue Gene, the fastest computer in the world.
25
110330
4000
ื–ื”ื• ื‘ืœื• ื’'ื™ืŸ (Blue Gene), ื”ืžื—ืฉื‘ ื”ืžื”ื™ืจ ื‘ืขื•ืœื.
01:54
It's got 120,000 processors; they can basically process
26
114330
5000
ื™ืฉ ืœื• 120,000 ืžืขื‘ื“ื™ื. ื”ื ื™ื›ื•ืœื™ื ื‘ืขื™ืงืจื•ืŸ ืœืขื‘ื“
01:59
10 quadrillion bits of information per second.
27
119330
3000
10 ืงื•ืื“ืจื™ืœื™ื•ืŸ ื‘ื™ื˜ื™ื ืฉืœ ืžื™ื“ืข ื‘ืฉื ื™ื”.
02:02
That's 10 to the sixteenth. And they consume one and a half megawatts of power.
28
122330
7000
ื–ื” 10 ื‘ื—ื–ืงืช 16. ื•ื”ื ืžื›ืœื™ื ื—ืฆื™ ืžื’ื” ื•ื•ืื˜ ืฉืœ ื›ื•ื—.
02:09
So that would be really great, if you could add that
29
129330
3000
ืื– ื–ื” ื”ื™ื” ืžืžืฉ ื ืคืœื ืื ื™ื›ื•ืœื ื• ืœื”ื•ืกื™ืฃ ืืช ื–ื”
02:12
to the production capacity in Tanzania.
30
132330
2000
ืœื™ื›ื•ืœืช ื”ื™ื™ืฆื•ืจ ืฉืœ ื˜ื ื–ื ื™ื”.
02:14
It would really boost the economy.
31
134330
2000
ื–ื” ื”ื™ื” ืžืžืฉ ืžืงืคื™ืฅ ืืช ื”ื›ืœื›ืœื”.
02:16
Just to go back to the States,
32
136330
4000
ืื ื ื—ื–ื•ืจ ืœืืจื”"ื‘,
02:20
if you translate the amount of power or electricity
33
140330
2000
ืื ื ืชืจื’ื ืืช ื›ืžื•ืช ื”ื›ื•ื— ืื• ื”ื—ืฉืžืœ
02:22
this computer uses to the amount of households in the States,
34
142330
3000
ื‘ื”ื ื”ืžื—ืฉื‘ ื”ื–ื” ืžืฉืชืžืฉ ืœื›ืžื•ืช ื‘ืชื™ ืื‘ ื‘ืืจื”"ื‘,
02:25
you get 1,200 households in the U.S.
35
145330
4000
ื ืงื‘ืœ 1,200 ื‘ืชื™ ืื‘,
02:29
That's how much power this computer uses.
36
149330
2000
ื–ื• ื›ืžื•ืช ื”ื›ื•ื— ืฉื”ืžื—ืฉื‘ ื”ื–ื” ืฆื•ืจืš.
02:31
Now, let's compare this with the brain.
37
151330
3000
ืขื›ืฉื™ื•, ื‘ื•ืื• ื ืฉื•ื•ื” ืืช ื–ื” ืœืžื•ื—.
02:34
This is a picture of, actually Rory Sayres' girlfriend's brain.
38
154330
5000
ื–ื• ืชืžื•ื ื” ืœืžืขืฉื”, ืฉืœ ื”ืžื•ื— ืฉืœ ื”ื—ื‘ืจื” ืฉืœ ืจื•ืจื™ ืกื™ื™ืจืก.
02:39
Rory is a graduate student at Stanford.
39
159330
2000
ืจื•ืจื™ ื”ื•ื ืกื˜ื•ื“ื ื˜ ืœืชื•ืืจ ืžืชืงื“ื ื‘ืกื˜ื ืคื•ืจื“.
02:41
He studies the brain using MRI, and he claims that
40
161330
4000
ื”ื•ื ื—ื•ืงืจ ืืช ื”ืžื•ื— ื‘ืืžืฆืขื•ืช MRI, ื•ื”ื•ื ื˜ื•ืขืŸ
02:45
this is the most beautiful brain that he has ever scanned.
41
165330
3000
ืฉื–ื” ื”ืžื•ื— ื”ื›ื™ ื™ืคื” ืฉื”ื•ื ืื™ ืคืขื ืกืจืง.
02:48
(Laughter)
42
168330
2000
(ืฆื—ื•ืง)
02:50
So that's true love, right there.
43
170330
3000
ืื– ื–ื• ืื”ื‘ืช ืืžืช.
02:53
Now, how much computation does the brain do?
44
173330
3000
ืขื›ืฉื™ื•, ื›ืžื” ื—ื™ืฉื•ื‘ื™ื ื”ืžื•ื— ืขื•ืฉื”?
02:56
I estimate 10 to the 16 bits per second,
45
176330
2000
ืื ื™ ืžืขืจื™ืš ืฉ-10 ื‘ื—ื–ืงืช 16 ื‘ื™ื˜ ืœืฉื ื™ื”
02:58
which is actually about very similar to what Blue Gene does.
46
178330
4000
ืฉื–ื” ื‘ืขืฆื ืžืื•ื“ ื“ื•ืžื” ืœืžื” ืฉื‘ืœื• ื’'ื™ืŸ ืขื•ืฉื”
03:02
So that's the question. The question is, how much --
47
182330
2000
ืื– ื–ื• ื”ืฉืืœื”. ื”ืฉืืœื” ื”ื™ื ื›ืžื” --
03:04
they are doing a similar amount of processing, similar amount of data --
48
184330
3000
ื”ื ืžื‘ืฆืขื™ื ื›ืžื•ืช ื—ื™ืฉื•ื‘ื™ื ื–ื”ื”, ื›ืžื•ืช ืžื™ื“ืข ื–ื”ื” --
03:07
the question is how much energy or electricity does the brain use?
49
187330
5000
ื”ืฉืืœื” ื”ื™ื ื‘ื›ืžื” ืื ืจื’ื™ื” ืื• ื—ืฉืžืœ ืžืฉืชืžืฉ ื”ืžื•ื—?
03:12
And it's actually as much as your laptop computer:
50
192330
3000
ื•ืœืžืขืŸ ื”ืืžืช, ื”ื™ื ื›ืžื• ืฉืœ ื”ืžื—ืฉื‘ ื”ื ื™ื™ื“ ืฉืœื›ื.
03:15
it's just 10 watts.
51
195330
2000
ื”ื™ื ืจืง 10 ื•ื•ืื˜.
03:17
So what we are doing right now with computers
52
197330
3000
ืื– ืžื” ืฉืื ื—ื ื• ืขื•ืฉื™ื ืขื›ืฉื™ื• ืขื ืžื—ืฉื‘ื™ื,
03:20
with the energy consumed by 1,200 houses,
53
200330
3000
ืขื ืื ืจื’ื™ื” ื–ื”ื” ืœื–ื• ืฉืœ 1,200 ื‘ืชื™ื,
03:23
the brain is doing with the energy consumed by your laptop.
54
203330
5000
ื”ืžื•ื— ืขื•ืฉื” ื‘ืขื–ืจืช ื”ืื ืจื’ื™ื” ืฉืœ ื”ืžื—ืฉื‘ ื”ื ื™ื™ื“ ืฉืœื›ื.
03:28
So the question is, how is the brain able to achieve this kind of efficiency?
55
208330
3000
ืื– ื”ืฉืืœื” ื”ื™ื, ืื™ืš ื”ืžื•ื— ืžืฆืœื™ื— ืœื”ื’ื™ืข ืœื™ืขื™ืœื•ืช ื›ื–ื•?
03:31
And let me just summarize. So the bottom line:
56
211330
2000
ื•ืชื ื• ืœื™ ืจืง ืœืกื›ื. ืื– ื”ืฉื•ืจื” ื”ืื—ืจื•ื ื” ื”ื™ื:
03:33
the brain processes information using 100,000 times less energy
57
213330
4000
ื”ืžื•ื— ืžืขื‘ื“ ืื™ื ืคื•ืจืžืฆื™ื” ื‘ืขื–ืจืช 100,000 ืคืขืžื™ื ืคื—ื•ืช ืื ืจื’ื™ื”
03:37
than we do right now with this computer technology that we have.
58
217330
4000
ืžืืฉืจ ื˜ื›ื ื•ืœื•ื’ื™ืช ื”ืžื—ืฉื‘ื™ื ืฉืื ื—ื ื• ืžืฉืชืžืฉื™ื ื‘ื” ื›ืจื’ืข
03:41
How is the brain able to do this?
59
221330
2000
ืื™ืš ื”ืžื•ื— ืžืกื•ื’ืœ ืœืขืฉื•ืช ื–ืืช?
03:43
Let's just take a look about how the brain works,
60
223330
3000
ื‘ื•ืื• ื ืกืชื›ืœ ืขืœ ืื™ืš ืฉื”ืžื•ื— ืขื•ื‘ื“,
03:46
and then I'll compare that with how computers work.
61
226330
4000
ื•ืื– ืื ื™ ืืฉื•ื•ื” ื–ืืช ืขื ืื™ืš ืฉืžื—ืฉื‘ื™ื ืขื•ื‘ื“ื™ื.
03:50
So, this clip is from the PBS series, "The Secret Life of the Brain."
62
230330
4000
ืื– ื”ืงืœื™ืค ื”ื–ื” ื”ื•ื ืžืจืฉืช PBS, "ื”ื—ื™ื™ื ื”ืกื•ื“ื™ื™ื ืฉืœ ื”ืžื•ื—"
03:54
It shows you these cells that process information.
63
234330
3000
ื”ื•ื ืžืจืื” ืœื›ื ืืช ื”ืชืื™ื ื”ืืœื” ืฉืžืขื‘ื“ื™ื ืื™ื ืคื•ืจืžืฆื™ื”
03:57
They are called neurons.
64
237330
1000
ื”ื ื ืงืจืื™ื ื ื•ื™ืจื•ื ื™ื.
03:58
They send little pulses of electricity down their processes to each other,
65
238330
6000
ื”ื ืฉื•ืœื—ื™ื ืคืขื™ืžื•ืช ืงื˜ื ื•ืช ืฉืœ ื—ืฉืžืœ ื—ื™ื‘ื•ืจื™ื ืื—ื“ ืœืฉื ื™,
04:04
and where they contact each other, those little pulses
66
244330
2000
ื•ื‘ื—ื™ื‘ื•ืจ ื‘ื™ื ื™ื”ื, ื”ืคืขื™ืžื•ืช ื”ืงื˜ื ื•ืช ื”ืืœื”
04:06
of electricity can jump from one neuron to the other.
67
246330
2000
ืฉืœ ื—ืฉืžืœ ื™ื›ื•ืœื•ืช ืœืงืคื•ืฅ ืžื ื•ื™ืจื•ืŸ ืื—ื“ ืœืฉื ื™.
04:08
That process is called a synapse.
68
248330
3000
ื—ื™ื‘ื•ืจ ื–ื” ื ืงืจื ืกื™ื ืคืกื”.
04:11
You've got this huge network of cells interacting with each other --
69
251330
2000
ื™ืฉ ืœื›ื ืจืฉืช ืขื ืงื™ืช ืฉืœ ืชืื™ื ื”ืžืชืงืฉืจื™ื ืื—ื“ ืขื ื”ืฉื ื™,
04:13
about 100 million of them,
70
253330
2000
ื‘ืขืจืš 100 ืžื™ืœื™ื•ืŸ,
04:15
sending about 10 quadrillion of these pulses around every second.
71
255330
4000
ืฉื•ืœื—ื™ื ื‘ืขืจืš 10 ืงื•ื•ืื“ืจื™ืœื™ื•ืŸ ืคืขื™ืžื•ืช ื‘ื›ืœ ืฉื ื™ื”.
04:19
And that's basically what's going on in your brain right now as you're watching this.
72
259330
6000
ื•ื–ื” ื‘ืขื™ืงืจื•ืŸ ืžื” ืฉืงื•ืจื” ื‘ืžื•ื— ืฉืœื›ื ื‘ืจื’ืข ื–ื” ื›ืฉืืชื ืฆื•ืคื™ื ื‘ื–ื”.
04:25
How does that compare with the way computers work?
73
265330
2000
ืื™ืš ื–ื” ืžืฉืชื•ื•ื” ืœื“ืจืš ื‘ื” ืžื—ืฉื‘ื™ื ืขื•ื‘ื“ื™ื?
04:27
In the computer, you have all the data
74
267330
2000
ื‘ืžื—ืฉื‘ ื™ืฉ ืœื›ื ืืช ื›ืœ ื”ืžื™ื“ืข
04:29
going through the central processing unit,
75
269330
2000
ืขื•ื‘ืจื™ื ื“ืจืš ื”ืžืขื‘ื“ ื”ืžืจื›ื–ื™,
04:31
and any piece of data basically has to go through that bottleneck,
76
271330
3000
ื•ื›ืœ ืคื™ืกืช ืžื™ื“ืข ื‘ืขื™ืงืจื•ืŸ ืฆืจื™ื›ื” ืœืขื‘ื•ืจ ื“ืจืš ืฆื•ื•ืืจ ื”ื‘ืงื‘ื•ืง ื”ื–ื”.
04:34
whereas in the brain, what you have is these neurons,
77
274330
4000
ื•ืื™ืœื• ื‘ืžื•ื—, ืžื” ืฉื™ืฉ ื”ื ื”ื ื™ื•ืจื•ื ื™ื ื”ืืœื•
04:38
and the data just really flows through a network of connections
78
278330
4000
ื•ื”ืžื™ื“ืข ื‘ืขืฆื ื–ื•ืจื ื“ืจืš ืจืฉืช ืฉืœ ืงื™ืฉื•ืจื™ื
04:42
among the neurons. There's no bottleneck here.
79
282330
2000
ื‘ื™ืŸ ื”ื ื™ื•ืจื•ื ื™ื, ืื™ืŸ ืฆื•ื•ืืจ ื‘ืงื‘ื•ืง.
04:44
It's really a network in the literal sense of the word.
80
284330
4000
ื–ื• ื‘ืืžืช "ืขื‘ื•ื“ืช ืจืฉืช" (network) ื‘ืžืœื•ื ืžื•ื‘ืŸ ื”ืžื™ืœื”.
04:48
The net is doing the work in the brain.
81
288330
4000
ื”ืจืฉืช ืขื•ืฉื” ืืช ื”ืขื‘ื•ื“ื”.
04:52
If you just look at these two pictures,
82
292330
2000
ืื ืจืง ืชืกืชื›ืœื• ืขืœ ืฉืชื™ ื”ืชืžื•ื ื•ืช ื”ืืœื•,
04:54
these kind of words pop into your mind.
83
294330
2000
ืืœื• ื”ืžื™ืœื™ื ืฉืงื•ืคืฆื•ืช ืœืจืืฉ.
04:56
This is serial and it's rigid -- it's like cars on a freeway,
84
296330
4000
ื–ื” ื˜ื•ืจื™ ื•ื ื•ืงืฉื”. ื›ืžื• ืžื›ื•ื ื™ื•ืช ื‘ื›ื‘ื™ืฉ ืžื”ื™ืจ --
05:00
everything has to happen in lockstep --
85
300330
3000
ื”ื›ืœ ืฆืจื™ืš ืœืงืจื•ืช ื‘ืกื“ืจ ื ื•ืงืฉื”.
05:03
whereas this is parallel and it's fluid.
86
303330
2000
ืœืขื•ืžืช ื–ืืช ื–ื” ืžืงื‘ื™ืœื™ ื•ื–ื•ืจื.
05:05
Information processing is very dynamic and adaptive.
87
305330
3000
ืขื™ื‘ื•ื“ ืžื™ื“ืข ื”ื•ื ืžืื•ื“ ื“ื™ื ืžื™ ื•ื’ืžื™ืฉ.
05:08
So I'm not the first to figure this out. This is a quote from Brian Eno:
88
308330
4000
ืื– ืื ื™ ืœื ื”ืจืืฉื•ืŸ ืฉื”ื‘ื™ืŸ ื–ืืช. ื–ื”ื• ืฆื™ื˜ื•ื˜ ืฉืœ ื‘ืจื™ืืŸ ืื ื•:
05:12
"the problem with computers is that there is not enough Africa in them."
89
312330
4000
"ื”ื‘ืขื™ื” ืขื ืžื—ืฉื‘ื™ื ื”ื™ื ืฉืื™ืŸ ื‘ื”ื ืžืกืคื™ืง ืืคืจื™ืงื”."
05:16
(Laughter)
90
316330
6000
(ืฆื—ื•ืง)
05:22
Brian actually said this in 1995.
91
322330
3000
ื”ืืžืช ืฉื‘ืจื™ืืŸ ืืžืจ ืืช ื–ื” ื‘- 1995.
05:25
And nobody was listening then,
92
325330
3000
ื•ืืฃ ืื—ื“ ืœื ื”ืงืฉื™ื‘ ืื–,
05:28
but now people are beginning to listen
93
328330
2000
ืื‘ืœ ืขื›ืฉื™ื• ืื ืฉื™ื ืžืชื—ื™ืœื™ื ืœื”ืงืฉื™ื‘
05:30
because there's a pressing, technological problem that we face.
94
330330
5000
ื‘ื’ืœืœ ืฉื™ืฉ ืžื•ืœื ื• ื‘ืขื™ื” ื˜ื›ื ื•ืœื•ื’ื™ืช ืœื•ื—ืฆืช,
05:35
And I'll just take you through that a little bit in the next few slides.
95
335330
5000
ื•ืื ื™ ืจืง ืืกื‘ื™ืจ ืœื›ื ืืช ื–ื” ืงืฆืช ื‘ืฉืงื•ืคื™ื•ืช ื”ื‘ืื•ืช.
05:40
This is -- it's actually really this remarkable convergence
96
340330
4000
ื–ื” -- ื–ื” ื‘ืขืฆื ืฉื™ืœื•ื‘ ืžื“ื”ื™ื
05:44
between the devices that we use to compute in computers,
97
344330
5000
ื‘ื™ืŸ ื”ื›ืœื™ ืฉืื ื—ื ื• ืžืฉืชืžืฉื™ื ื‘ื• ืœื—ืฉื‘ ื‘ืžื—ืฉื‘ื™ื,
05:49
and the devices that our brains use to compute.
98
349330
4000
ืœื‘ื™ืŸ ื”ื›ืœื™ ืฉื”ืžื•ื— ืฉืœื ื• ืžืฉืชืžืฉ ื‘ื• ืœื—ืฉื‘.
05:53
The devices that computers use are what's called a transistor.
99
353330
4000
ื”ื›ืœื™ ืฉื”ืžื—ืฉื‘ื™ื ืžืฉืชืžืฉื™ื ื‘ื• ื ืงืจื ื˜ืจื ื–ื™ืกื˜ื•ืจ.
05:57
This electrode here, called the gate, controls the flow of current
100
357330
4000
ื”ืืœืงื˜ืจื•ื“ื” ื”ื–ืืช, ื ืงืจืืช ืฉืขืจ, ืฉื•ืœื˜ืช ื‘ื–ืจื ื”ื—ืฉืžืœื™
06:01
from the source to the drain -- these two electrodes.
101
361330
3000
ืžื”ืžืงื•ืจ ืœืฉืคืš, ืฉืชื™ ื”ืืœืงื˜ืจื•ื“ื•ืช ื”ืืœื”.
06:04
And that current, electrical current,
102
364330
2000
ื•ื”ื–ืจื ื”ื–ื”, ื–ืจื ื—ืฉืžืœื™,
06:06
is carried by electrons, just like in your house and so on.
103
366330
6000
ื ื™ืฉื ืขืœ ื™ื“ื™ ืืœืงื˜ืจื•ื ื™ื, ื‘ื“ื™ื•ืง ื›ืžื• ืืฆืœื›ื ื‘ื‘ื™ืช.
06:12
And what you have here is, when you actually turn on the gate,
104
372330
5000
ื•ืžื” ืฉื™ืฉ ืœื›ื ืคื”, ื›ืฉืืชื ืคื•ืชื—ื™ื ืืช ื”ืฉืขืจ,
06:17
you get an increase in the amount of current, and you get a steady flow of current.
105
377330
4000
ืืชื ืžืงื‘ืœื™ื ืขืœื™ื” ื‘ื›ืžื•ืช ื”ื–ืจื, ื•ืืชื ืžืงื‘ืœื™ื ื–ืจื ืงื‘ื•ืข.
06:21
And when you turn off the gate, there's no current flowing through the device.
106
381330
4000
ื•ื›ืฉืืชื ืกื•ื’ืจื™ื ืืช ื”ืฉืขืจ, ืื™ืŸ ื›ืœืœ ื–ืจื ื‘ืžื›ืฉื™ืจ.
06:25
Your computer uses this presence of current to represent a one,
107
385330
5000
ื”ืžื—ืฉื‘ ืฉืœื›ื ืžืฉืชืžืฉ ื‘ื ื•ื›ื—ื•ืช ื”ื–ืจื ื›ื“ื™ ืœื™ื™ืฆื’ ืื—ื“,
06:30
and the absence of current to represent a zero.
108
390330
4000
ื•ืืช ื”ื™ืขื“ืจ ื”ื–ืจื ื›ื“ื™ ืœื™ื™ืฆื’ ืืคืก.
06:34
Now, what's happening is that as transistors are getting smaller and smaller and smaller,
109
394330
6000
ืขื›ืฉื™ื•, ืžื” ืฉืงื•ืจื” ื–ื” ืฉื›ื›ืœ ืฉื”ื˜ืจื ื–ื™ืกื˜ื•ืจื™ื ืงื˜ื ื™ื ื•ืงื˜ื ื™ื,
06:40
they no longer behave like this.
110
400330
2000
ื”ื ื›ื‘ืจ ืœื ืžืชื ื”ื’ื™ื ื›ื›ื”.
06:42
In fact, they are starting to behave like the device that neurons use to compute,
111
402330
5000
ื‘ืขืฆื, ื”ื ืžืชื—ื™ืœื™ื ืœื”ืชื ื”ื’ ื›ืžื• ื”ื›ืœื™ื ื‘ื”ื ื”ื ื•ื™ืจื•ื ื™ื ืžืฉืชืžืฉื™ื ื›ื“ื™ ืœื—ืฉื‘,
06:47
which is called an ion channel.
112
407330
2000
ืฉื ืงืจืื™ื ืชืขืœื•ืช ื™ื•ื ื™ื.
06:49
And this is a little protein molecule.
113
409330
2000
ื•ื–ื• ืžื•ืœืงื•ืœืช ื—ืœื‘ื•ืŸ ืงื˜ื ื”.
06:51
I mean, neurons have thousands of these.
114
411330
4000
ืื ื™ ืžืชื›ื•ื•ืŸ, ืœื ื•ื™ืจื•ื ื™ื ื™ืฉ ื›ืืœื” ื‘ืืœืคื™ื.
06:55
And it sits in the membrane of the cell and it's got a pore in it.
115
415330
4000
ื•ื–ื” ื™ื•ืฉื‘ ื‘ืžืžื‘ืจื ื” ืฉืœ ื”ืชื ื•ื™ืฉ ื‘ื• ื ืงื‘ื•ื‘ื™ืช.
06:59
And these are individual potassium ions
116
419330
3000
ื•ืืœื• ื™ื•ื ื™ ืืฉืœื’ืŸ ื‘ื•ื“ื“ื™ื,
07:02
that are flowing through that pore.
117
422330
2000
ืฉื–ื•ืจืžื™ื ื“ืจืš ื”ื ืงื‘ื•ื‘ื™ืช.
07:04
Now, this pore can open and close.
118
424330
2000
ืขื›ืฉื™ื•, ื”ื ืงื‘ื•ื‘ื™ืช ื”ื–ื• ื™ื›ื•ืœื” ืœื”ื™ืกื’ืจ ื•ืœื”ื™ืคืชื—.
07:06
But, when it's open, because these ions have to line up
119
426330
5000
ืื‘ืœ, ื›ืฉื”ื™ื ืคืชื•ื—ื”, ื‘ื’ืœืœ ืฉื”ื™ื•ื ื™ื ื”ืืœื• ืฆืจื™ื›ื™ื ืœื”ืกืชื“ืจ ื‘ืฉื•ืจื”
07:11
and flow through, one at a time, you get a kind of sporadic, not steady --
120
431330
5000
ื•ืœื–ืจื•ื ืื—ื“ ืื—ื“, ืืชื ืžืงื‘ืœื™ื ืื™ ืกื“ืจ, ื—ื•ืกืจ ื™ืฆื™ื‘ื•ืช -
07:16
it's a sporadic flow of current.
121
436330
3000
ื–ืจื™ืžืช ื—ืฉืžืœ ืœื ืกื“ื™ืจื”.
07:19
And even when you close the pore -- which neurons can do,
122
439330
3000
ื•ื’ื ื›ืฉืืชื ืกื•ื’ืจื™ื ืืช ื”ื ืงื‘ื•ื‘ื™ืช -- ืžื” ืฉื”ื ื•ื™ืจื•ื ื™ื ื™ื›ื•ืœื™ื ืœืขืฉื•ืช,
07:22
they can open and close these pores to generate electrical activity --
123
442330
5000
ื”ื ื™ื›ื•ืœื™ื ืœืกื’ื•ืจ ื•ืœืคืชื•ื— ืืช ื”ื ืงื‘ื•ื‘ื™ื•ืช ื”ืืœื” ื›ื“ื™ ืœื™ื™ืฆืจ ืคืขื™ืœื•ืช ื—ืฉืžืœื™ืช --
07:27
even when it's closed, because these ions are so small,
124
447330
3000
ื’ื ื›ืฉื–ื” ืกื’ื•ืจ, ื‘ื’ืœืœ ืฉื”ื™ื•ื ื™ื ื”ืืœื• ื›ืœ ื›ืš ืงื˜ื ื™ื,
07:30
they can actually sneak through, a few can sneak through at a time.
125
450330
3000
ื”ื ื™ื›ื•ืœื™ื ืœื”ืชื’ื ื‘ ืคื ื™ืžื”, ื›ืžื” ื™ื›ื•ืœื™ื ืœื”ืชื’ื ื‘ ืคื ื™ืžื” ื›ืœ ืคืขื.
07:33
So, what you have is that when the pore is open,
126
453330
3000
ืื– ื›ืฉื”ื ืงื‘ื•ื‘ื™ืช ืคืชื•ื—ื”,
07:36
you get some current sometimes.
127
456330
2000
ืืชื ืžืงื‘ืœื™ื ืงืฆืช ื–ืจื ืžื“ื™ ืคืขื.
07:38
These are your ones, but you've got a few zeros thrown in.
128
458330
3000
ืืœื” ื”ืื—ื“ื™ื ืฉืœื›ื, ืื‘ืœ ืืชื ืžืงื‘ืœื™ื ื›ืžื” ืืคืกื™ื ื‘ื›ืœ ื–ืืช.
07:41
And when it's closed, you have a zero,
129
461330
4000
ื•ื›ืฉื–ื” ืกื’ื•ืจ, ืืชื ืžืงื‘ืœื™ื ืืคืก,
07:45
but you have a few ones thrown in.
130
465330
3000
ืื‘ืœ ืืชื ืžืงื‘ืœื™ื ืงืฆืช ืื—ื“ื™ื, ืื•ืงื™.
07:48
Now, this is starting to happen in transistors.
131
468330
3000
ืขื›ืฉื™ื•, ื–ื” ืžืชื—ื™ืœ ืœืงืจื•ืช ื‘ื˜ืจื ื–ื™ืกื˜ื•ืจื™ื.
07:51
And the reason why that's happening is that, right now, in 2007 --
132
471330
5000
ื•ื”ืกื™ื‘ื” ืฉื”ื“ื‘ืจ ื”ื–ื” ืงื•ืจื”, ื”ื™ื ืฉืขื›ืฉื™ื• ื‘ืฉื ืช 2007,
07:56
the technology that we are using -- a transistor is big enough
133
476330
4000
ื”ื˜ื›ื ื•ืœื•ื’ื™ื” ื‘ื” ืื ื—ื ื• ืžืฉืชืžืฉื™ื, ื”ื˜ืจื ื–ื™ืกื˜ื•ืจ ื’ื“ื•ืœ ืžืกืคื™ืง
08:00
that several electrons can flow through the channel simultaneously, side by side.
134
480330
5000
ืฉื›ืžื” ืืœืงื˜ืจื•ื ื™ื ื™ื›ื•ืœื™ื ืœื–ืจื•ื ื“ืจืš ื”ืชืขืœื” ื‘ื• ื–ืžื ื™ืช, ืื—ื“ ืœืฆื“ ื”ืฉื ื™.
08:05
In fact, there's about 12 electrons can all be flowing this way.
135
485330
4000
ืœืžืขืฉื”, ื™ืฉ ื‘ืขืจืš 12 ืืœืงื˜ืจื•ื ื™ื ืฉื™ื›ื•ืœื™ื ืœื–ืจื•ื ื‘ื›ื™ื•ื•ืŸ ื–ื”.
08:09
And that means that a transistor corresponds
136
489330
2000
ื•ื–ื” ืื•ืžืจ ืฉื”ื˜ืจื ื–ื™ืกื˜ื•ืจ ืžืชืื™ื
08:11
to about 12 ion channels in parallel.
137
491330
3000
ืœื‘ืขืจืš 12 ืขืจื•ืฆื™ ื™ื•ื ื™ื ื‘ืžืงื‘ื™ืœ.
08:14
Now, in a few years time, by 2015, we will shrink transistors so much.
138
494330
5000
ืขื›ืฉื™ื•, ื‘ืขื•ื“ ื›ืžื” ืฉื ื™ื, ืขื“ 2015, ืื ื—ื ื• ื ื›ื•ื•ืฅ ืืช ื”ื˜ืจื ื–ื™ืกื˜ื•ืจื™ื ื›ืœ ื›ืš.
08:19
This is what Intel does to keep adding more cores onto the chip.
139
499330
5000
ื–ื” ืžื” ืฉืื™ื ื˜ืœ ืขื•ืฉื” ื›ื“ื™ ืœื”ืžืฉื™ืš ืœื”ื•ืกื™ืฃ ืœื™ื‘ื•ืช ืขืœ ื”ืฉื‘ื‘,
08:24
Or your memory sticks that you have now can carry one gigabyte
140
504330
3000
ืื• ื›ืจื˜ื™ืกื™ ื”ื–ื™ื›ืจื•ืŸ ืฉื™ืฉ ืœื›ื ืขื›ืฉื™ื• ืฉื™ื›ื•ืœื™ื ืœืื—ืกืŸ ื’'ื™ื’ื” ื‘ื™ื™ื˜ ืื—ื“
08:27
of stuff on them -- before, it was 256.
141
507330
2000
ืฉืœ ื ืชื•ื ื™ื ืขืœื™ื”ื -- ืœืคื ื™ ืฉื–ื” ื”ื™ื” 256.
08:29
Transistors are getting smaller to allow this to happen,
142
509330
3000
ื˜ืจื ืกื™ื–ื˜ื•ืจื™ื ื ืขืฉื™ื ืงื˜ื ื™ื ื™ื•ืชืจ, ื•ืžืืคืฉืจื™ื ืœื–ื” ืœืงืจื•ืช,
08:32
and technology has really benefitted from that.
143
512330
3000
ื•ื”ื˜ื›ื ื•ืœื•ื’ื™ื” ืžืžืฉ ื”ืจื•ื•ื™ื—ื” ืžื–ื”.
08:35
But what's happening now is that in 2015, the transistor is going to become so small,
144
515330
5000
ืื‘ืœ ืžื” ืฉืงื•ืจื” ืขื›ืฉื™ื• ื–ื” ืฉืขื“ 2015, ื”ื˜ืจื ื–ื™ืกื˜ื•ืจื™ื ืขื•ืžื“ื™ื ืœื”ื™ื•ืช ื›ืœ ื›ืš ืงื˜ื ื™ื,
08:40
that it corresponds to only one electron at a time
145
520330
3000
ืฉื”ื ื™ืชืื™ืžื• ืœืืœืงื˜ืจื•ืŸ ืื—ื“ ื‘ืœื‘ื“ ื‘ื›ืœ ื–ืžืŸ
08:43
can flow through that channel,
146
523330
2000
ืฉื–ื•ืจื ื“ืจืš ื”ืชืขืœื”,
08:45
and that corresponds to a single ion channel.
147
525330
2000
ื•ื–ื” ืžืชืื™ื ืœืชืขืœืช ื™ื•ื ื™ื ืื—ืช.
08:47
And you start having the same kind of traffic jams that you have in the ion channel.
148
527330
4000
ื•ืืชื ืžืชื—ื™ืœื™ื ืœืงื‘ืœ ืืช ืื•ืชื ืคืงืงื™ ืชื ื•ืขื” ืฉื™ืฉ ืœื›ื ื‘ืชืขืœื•ืช ื”ื™ื•ื ื™ื,
08:51
The current will turn on and off at random,
149
531330
3000
ื”ื–ืจื ื™ื“ืœืง ื•ื™ื›ื‘ื” ืจื ื“ื•ืžืœื™ืช
08:54
even when it's supposed to be on.
150
534330
2000
ืืคื™ืœื• ื›ืฉื”ื•ื ืืžื•ืจ ืœื”ื™ื•ืช ื“ื•ืœืง.
08:56
And that means your computer is going to get
151
536330
2000
ื•ื–ื” ืื•ืžืจ ืฉื”ืžื—ืฉื‘ ืฉืœื›ื ื”ื•ืœืš ืœืงื‘ืœ
08:58
its ones and zeros mixed up, and that's going to crash your machine.
152
538330
4000
ื‘ืœื‘ื•ืœ ืฉืœ ืื—ื“ื™ื ื•ืืคืกื™ื, ื•ื–ื” ื”ื•ืœืš ืœื’ืจื•ื ืœืžื›ื•ื ื” ืฉืœื›ื ืœืงืจื•ืก.
09:02
So, we are at the stage where we
153
542330
4000
ืื–, ืื ื—ื ื• ื‘ืฉืœื‘ ื‘ื• ืื ื—ื ื•
09:06
don't really know how to compute with these kinds of devices.
154
546330
3000
ืœื ื‘ืืžืช ื™ื•ื“ืขื™ื ืœื—ืฉื‘ ืขื ื›ืœื™ื ืžืกื•ื’ ื›ื–ื”.
09:09
And the only kind of thing -- the only thing we know right now
155
549330
3000
ื•ื”ืกื•ื’ ื”ื™ื—ื™ื“, ื”ื“ื‘ืจ ืฉืื ื—ื ื• ืžื›ื™ืจื™ื ื›ืจื’ืข,
09:12
that can compute with these kinds of devices are the brain.
156
552330
3000
ืฉื™ื•ื“ืข ืœื—ืฉื‘ ื‘ืขื–ืจืช ื›ืœื™ื ื›ืืœื”, ื”ื•ื ื”ืžื•ื—.
09:15
OK, so a computer picks a specific item of data from memory,
157
555330
4000
ืื•ืงื™, ืื– ืžื—ืฉื‘ ืฉื•ืœืš ืคื™ืกืช ืžื™ื“ืข ืžืกื•ื™ื™ืžืช ืžื”ื–ื™ื›ืจื•ืŸ,
09:19
it sends it into the processor or the ALU,
158
559330
3000
ื”ื•ื ืฉื•ืœื— ืืช ื–ื” ืœืžืขื‘ื“ ืื• ืœ ALU,
09:22
and then it puts the result back into memory.
159
562330
2000
ื•ืื– ื”ื•ื ืžื›ื ื™ืก ืืช ื”ืชื•ืฆืื•ืช ื—ื–ืจื” ืœื–ื™ื›ืจื•ืŸ.
09:24
That's the red path that's highlighted.
160
564330
2000
ื–ื” ื”ื ืชื™ื‘ ื”ืื“ื•ื ื”ืžื•ื“ื’ืฉ.
09:26
The way brains work, I told you all, you have got all these neurons.
161
566330
4000
ื”ื“ืจืš ื‘ื” ืžื•ื—ื•ืช ืคื•ืขืœื™ื, ื™ืฉ ืœื›ื ืืช ื›ืœ ื”ื ื™ื•ืจื•ื ื™ื ื”ืืœื”.
09:30
And the way they represent information is
162
570330
2000
ื•ื”ื“ืจืš ื‘ื” ื”ื ืžื™ื™ืฆื’ื™ื ืžื™ื“ืข ื”ื•ื
09:32
they break up that data into little pieces
163
572330
2000
ืฉื”ื ืžืคืจืงื™ื ืืช ื”ืžื™ื“ืข ืœื—ืชื™ื›ื•ืช ืงื˜ื ื•ืช
09:34
that are represented by pulses and different neurons.
164
574330
3000
ืฉืžื™ื•ืฆื’ื•ืช ืขืœ ื™ื“ื™ ืคื•ืœืกื™ื ื•ื ื™ื•ืจื•ื ื™ื ืฉื•ื ื™ื.
09:37
So you have all these pieces of data
165
577330
2000
ืื– ื™ืฉ ืœื›ื ืืช ื›ืœ ืคื™ืกื•ืช ื”ืžื™ื“ืข ื”ื–ื”
09:39
distributed throughout the network.
166
579330
2000
ืžืคื•ื–ืจื•ืช ื‘ื›ืœ ื‘ืจืฉืช.
09:41
And then the way that you process that data to get a result
167
581330
3000
ื•ืื– ื”ื“ืจืš ืฉืืชื ืžืขื‘ื“ื™ื ืžื™ื“ืข ื›ื“ื™ ืœืงื‘ืœ ืชื•ืฆืื”
09:44
is that you translate this pattern of activity into a new pattern of activity,
168
584330
4000
ื”ื™ื ืฉืืชื ืžืชืจื’ืžื™ื ืืช ื”ืชื‘ื ื™ืช ื”ื–ื• ืฉืœ ืคืขื™ืœื•ืช ืœืชื‘ื ื™ืช ื—ื“ืฉื” ืฉืœ ืคืขื™ืœื•ืช,
09:48
just by it flowing through the network.
169
588330
3000
ืจืง ืขืœ ื™ื“ื™ ื›ืš ืฉื”ื™ื ื–ื•ืจืžืช ื‘ืจืฉืช.
09:51
So you set up these connections
170
591330
2000
ืื– ืืชื ืžื™ื™ืฆืจื™ื ืืช ื”ืงืฉืจื™ื ื”ืืœื”,
09:53
such that the input pattern just flows
171
593330
3000
ื›ืš ืฉื”ืžื™ื“ืข ื”ื ื™ื›ื ืก ืจืง ื–ื•ืจื
09:56
and generates the output pattern.
172
596330
2000
ื•ื™ื•ืฆืจ ืืช ืชื‘ื ื™ืช ื”ื™ืฆื™ืื”.
09:58
What you see here is that there's these redundant connections.
173
598330
4000
ืžื” ืฉืืชื ืจื•ืื™ื ื›ืืŸ ื–ื” ืฉื™ืฉ ืืช ื”ืงืฉืจื™ื ื”ืขื•ื“ืคื™ื.
10:02
So if this piece of data or this piece of the data gets clobbered,
174
602330
4000
ื›ืš ืฉืื ืคื™ืกืช ืžื™ื“ืข ื–ื• ืื• ืคื™ืกืช ืžื™ื“ืข ื–ื• ื ืขืœืžืช
10:06
it doesn't show up over here, these two pieces can activate the missing part
175
606330
5000
ื”ื™ื ืœื ืžื•ืคื™ืขื” ื›ืืŸ, ืฉืชื™ ืคื™ืกื•ืช ืืœื” ื™ื›ื•ืœื•ืช ืœืฉื—ื–ืจ ืืช ื”ื—ืœืง ื”ื—ืกืจ
10:11
with these redundant connections.
176
611330
2000
ื‘ืขื–ืจืช ื”ืงืฉืจื™ื ื”ืขื•ื“ืคื™ื.
10:13
So even when you go to these crappy devices
177
613330
2000
ืื– ื’ื ืื ืืชื ื”ื•ืœื›ื™ื ืœื›ืœื™ื ื”ื“ืคื•ืงื™ื ื”ืืœื”
10:15
where sometimes you want a one and you get a zero, and it doesn't show up,
178
615330
3000
ื‘ื”ื ืืชื ืœืคืขืžื™ื ืจื•ืฆื™ื ืื—ื“ ื•ืžืงื‘ืœื™ื ืืคืก,
10:18
there's redundancy in the network
179
618330
2000
ื™ืฉ ืขื•ื“ืคื™ื ื‘ืžืขืจื›ืช
10:20
that can actually recover the missing information.
180
620330
3000
ืฉื™ื›ื•ืœื™ื ืœืฉื—ื–ืจ ืืช ื”ืžื™ื“ืข ื”ื—ืกืจ.
10:23
It makes the brain inherently robust.
181
623330
3000
ื–ื” ืขื•ืฉื” ืืช ื”ืžื•ื— ื—ืกื™ืŸ ืžื˜ื‘ืขื•.
10:26
What you have here is a system where you store data locally.
182
626330
3000
ืžื” ืฉื™ืฉ ืœื›ื ืคื” ื”ื™ื ืžืขืจื›ืช ื‘ื” ืืชื ืžืื—ืกื ื™ื ืžื™ื“ืข ืžืงื•ืžื™.
10:29
And it's brittle, because each of these steps has to be flawless,
183
629330
4000
ื•ื”ื™ื ืฉื‘ืจื™ืจื™ืช, ื›ื™ ื›ืœ ืื—ื“ ืžื”ืฉืœื‘ื™ื ืฆืจื™ืš ืœื”ื™ื•ืช ืžื•ืฉืœื,
10:33
otherwise you lose that data, whereas in the brain, you have a system
184
633330
3000
ืื—ืจืช ืืชื ืžืื‘ื“ื™ื ืืช ื”ืžื™ื“ืข. ืœืขื•ืžืช ื–ืืช ื‘ืžื•ื—, ื™ืฉ ืœื›ื ืžืขืจื›ืช
10:36
that stores data in a distributed way, and it's robust.
185
636330
4000
ืฉืžืื›ืกื ืช ืžื™ื“ืข ื‘ืฆื•ืจื” ืžืคื•ื–ืจืช, ื•ื”ื™ื ื—ืกื™ื ื”.
10:40
What I want to basically talk about is my dream,
186
640330
4000
ืžื” ืฉืื ื™ ืจื•ืฆื” ื‘ืขืฆื ืœื“ื‘ืจ ืขืœื™ื• ื”ื•ื ื”ื—ืœื•ื ืฉืœื™,
10:44
which is to build a computer that works like the brain.
187
644330
3000
ืฉื”ื•ื ืœื‘ื ื•ืช ืžื—ืฉื‘ ืฉืขื•ื‘ื“ ื›ืžื• ื”ืžื•ื—.
10:47
This is something that we've been working on for the last couple of years.
188
647330
4000
ื–ื” ืžืฉื”ื• ืฉืื ื—ื ื• ืขื•ื‘ื“ื™ื ืขืœื™ื• ื‘ืžืฉืš ื”ืฉื ืชื™ื™ื ื”ืื—ืจื•ื ื•ืช.
10:51
And I'm going to show you a system that we designed
189
651330
3000
ื•ืื ื™ ืขื•ืžื“ ืœื”ืจืื•ืช ืœื›ื ืžืขืจื›ืช ืฉืชื™ื›ื ื ื•
10:54
to model the retina,
190
654330
3000
ืœืžื“ืœ ืืช ื”ืจืฉืชื™ืช,
10:57
which is a piece of brain that lines the inside of your eyeball.
191
657330
5000
ืฉื”ื™ื ื—ืœืง ืžื”ืžื•ื— ืฉืžื›ืกื” ืืช ืคื ื™ื ื’ืœื’ืœ ื”ืขื™ืŸ.
11:02
We didn't do this by actually writing code, like you do in a computer.
192
662330
6000
ืœื ืขืฉื™ื ื• ื–ืืช ื‘ืขื–ืจืช ื›ืชื™ื‘ืช ืงื•ื“, ื›ืžื• ืฉืืชื ืขื•ืฉื™ื ื‘ืžื—ืฉื‘.
11:08
In fact, the processing that happens
193
668330
3000
ืœืžืขืฉื”, ื”ืชื”ืœื™ืš ืฉืžืชืจื—ืฉ
11:11
in that little piece of brain is very similar
194
671330
2000
ื‘ื—ืœืง ื”ืžื•ื— ื”ืงื˜ืŸ ื”ื–ื” ืžืื•ื“ ื“ื•ืžื”
11:13
to the kind of processing that computers
195
673330
1000
ืœืกื•ื’ ื”ื—ื™ืฉื•ื‘ื™ื ืฉืขื•ืฉื™ื ืžื—ืฉื‘ื™ื
11:14
do when they stream video over the Internet.
196
674330
4000
ื›ืฉื”ื ืฉื•ืœื—ื™ื ื•ื™ื“ืื• ื“ืจืš ื”ืื™ื ื˜ืจื ื˜.
11:18
They want to compress the information --
197
678330
1000
ื”ื ืจื•ืฆื™ื ืœื›ื•ื•ืฅ ืืช ื”ืžื™ื“ืข --
11:19
they just want to send the changes, what's new in the image, and so on --
198
679330
4000
ื”ื ืจืง ืจื•ืฆื™ื ืœืฉืœื•ื— ืืช ื”ืฉื™ื ื•ื™ื™ื ื”ืžืชื—ื“ืฉื™ื ื‘ืชืžื•ื ื” ื•ื›ืš ื”ืœืื” --
11:23
and that is how your eyeball
199
683330
3000
ื•ื›ื›ื” ื’ืœื’ืœ ื”ืขื™ืŸ ืฉืœื›ื
11:26
is able to squeeze all that information down to your optic nerve,
200
686330
3000
ืžืกื•ื’ืœ ืœื›ื•ื•ืฅ ืืช ื›ืœ ื”ืื™ื ืคื•ืจืžืฆื™ื” ื”ื–ืืช ืืœ ืขืฆื‘ ื”ืจืื™ื™ื”,
11:29
to send to the rest of the brain.
201
689330
2000
ื›ื“ื™ ืœืฉืœื•ื— ืื•ืชื” ืืœ ืฉืืจ ื”ืžื•ื—.
11:31
Instead of doing this in software, or doing those kinds of algorithms,
202
691330
3000
ื‘ืžืงื•ื ืœืขืฉื•ืช ื–ืืช ื‘ืขื–ืจืช ืชื•ื›ื ื”, ืื• ืœืขืฉื•ืช ืืช ื”ืกื•ื’ื™ื ื”ืืœื” ืฉืœ ืืœื’ื•ืจื™ืชืžื™ื,
11:34
we went and talked to neurobiologists
203
694330
3000
ื”ืœื›ื ื• ื•ื“ื™ื‘ืจื ื• ืขื ื ื•ื™ืจื•-ื‘ื™ื•ืœื•ื’ื™ื
11:37
who have actually reverse engineered that piece of brain that's called the retina.
204
697330
4000
ืฉื‘ืขืฆื ื‘ื™ืฆืขื• ื”ื ื“ืกื” ื”ืคื•ื›ื” ืœื—ืœืง ื”ื–ื” ืฉืœ ื”ืžื•ื— ืฉื ืงืจื ืจืฉืชื™ืช.
11:41
And they figured out all the different cells,
205
701330
2000
ื•ื”ืค ืคื™ืขื ื—ื• ืืช ื›ืœ ื”ืชืื™ื ื”ืฉื•ื ื™ื,
11:43
and they figured out the network, and we just took that network
206
703330
3000
ื•ื”ื ืคื™ืขื ื—ื• ืืช ื”ืจืฉืช, ื•ืื ื—ื ื• ืคืฉื•ื˜ ืœืงื—ื ื• ืืช ื”ืจืฉืช ื”ื–ืืช
11:46
and we used it as the blueprint for the design of a silicon chip.
207
706330
4000
ื•ื”ืฉืชืžืฉื ื• ื‘ื” ื›ืชืฉืชื™ืช ืœืขื™ืฆื•ื‘ ืฉืœ ืฉื‘ื‘ ืกื™ืœื™ืงื•ืŸ.
11:50
So now the neurons are represented by little nodes or circuits on the chip,
208
710330
6000
ืื– ืขื›ืฉื™ื• ื”ื ื•ื™ืจื•ื ื™ื ืžื™ื•ืฆื’ื™ื ืขืœ ื™ื“ื™ ืฆืžืชื™ื ืื• ืžืขื’ืœื™ื ืขืœ ื”ืฉื‘ื‘,
11:56
and the connections among the neurons are represented, actually modeled by transistors.
209
716330
5000
ื•ื”ืงื™ืฉื•ืจื™ื ื‘ื™ืŸ ื”ื ื•ื™ืจื•ื ื™ื ื‘ืขืฆื ืžืžื•ื“ืœื™ื ืขืœ ื™ื“ื™ ื˜ืจื ื–ื™ืกื˜ื•ืจื™ื
12:01
And these transistors are behaving essentially
210
721330
2000
ื•ื”ื˜ืจื ื–ื™ืกื˜ื•ืจื™ื ื”ืืœื” ืžืชื ื”ื’ื™ื ื‘ืขืฆื
12:03
just like ion channels behave in the brain.
211
723330
3000
ื‘ื“ื™ื•ืง ื›ืžื• ืฉืชืขืœื•ืช ื”ื™ื•ื ื™ื ืžืชื ื”ื’ื•ืช ื‘ืžื•ื—.
12:06
It will give you the same kind of robust architecture that I described.
212
726330
5000
ื–ื” ื™ืชืŸ ืœื›ื ื‘ื“ื™ื•ืง ืืช ืื•ืชื• ืกื•ื’ ืฉืœ ืืจื›ื™ื˜ืงื˜ื•ืจื” ื—ืกื™ื ื” ืฉืชื™ืืจืชื™.
12:11
Here is actually what our artificial eye looks like.
213
731330
4000
ื›ืš ืœืžืขืฉื” ื ืจืื™ืช ื”ืขื™ืŸ ื”ืžืœืื›ื•ืชื™ืช ืฉืœื ื•.
12:15
The retina chip that we designed sits behind this lens here.
214
735330
5000
ืฉื‘ื‘ ื”ืจืฉืชื™ืช ืฉืขื™ืฆื‘ื ื• ื™ื•ืฉื‘ ืžืื—ื•ืจื™ ื”ืขื“ืฉื” ื”ื–ื• ืคื”.
12:20
And the chip -- I'm going to show you a video
215
740330
2000
ื•ื”ืฉื‘ื‘ -- ืื ื™ ืขื•ืžื“ ืœื”ืจืื•ืช ืœื›ื ื•ื™ื“ืื•
12:22
that the silicon retina put out of its output
216
742330
3000
ืฉืจืฉืชื™ืช ื”ืกื™ืœื™ืงื•ืŸ ื”ื•ืฆื™ืื” ื‘ืคืœื˜
12:25
when it was looking at Kareem Zaghloul,
217
745330
3000
ื›ืฉื”ื™ื ื”ืกืชื›ืœื” ืขืœ ืงืืจื™ื ื–ืื’ื•ืœ,
12:28
who's the student who designed this chip.
218
748330
2000
ืฉื”ื•ื ื”ืกื˜ื•ื“ื ื˜ ืฉืขื™ืฆื‘ ืืช ื”ืฉื‘ื‘ ื”ื–ื”.
12:30
Let me explain what you're going to see, OK,
219
750330
2000
ืชื ื• ืœื™ ืœื”ืกื‘ื™ืจ ืžื” ืืชื ืขื•ื‘ื“ื™ื ืœืจืื•ืช, ืื•ืงื™.
12:32
because it's putting out different kinds of information,
220
752330
3000
ื›ื™ ื–ื” ืžื•ืฆื™ื ืกื•ื’ื™ื ืฉื•ื ื™ื ืฉืœ ืžื™ื“ืข,
12:35
it's not as straightforward as a camera.
221
755330
2000
ื–ื” ืœื ื™ืฉื™ืจ ื›ืžื• ืžืฆืœืžื”.
12:37
The retina chip extracts four different kinds of information.
222
757330
3000
ืฉื‘ื‘ ืจืฉืชื™ืช ืžื—ืœืฅ ืืจื‘ืข ืกื•ื’ื™ื ืฉืœ ืžื™ื“ืข.
12:40
It extracts regions with dark contrast,
223
760330
3000
ื”ื•ื ืžื—ืœืฅ ืื–ื•ืจื™ื ืขื ื ื™ื’ื•ื“ื™ื•ืช ื›ื”ื”,
12:43
which will show up on the video as red.
224
763330
3000
ืฉื™ื•ืคื™ืขื• ืขืœ ื”ื•ื™ื“ืื• ื‘ืื“ื•ื.
12:46
And it extracts regions with white or light contrast,
225
766330
4000
ื•ื”ื•ื ืžื—ืœืฅ ืื–ื•ืจื™ื ืœื‘ื ื™ื ืื• ื‘ืขืœื™ ื ื™ื’ื•ื“ื™ื•ืช ื‘ื”ื™ืจื”,
12:50
which will show up on the video as green.
226
770330
2000
ืฉื™ื•ืคื™ืขื• ื‘ื•ื™ื“ืื• ื‘ื™ืจื•ืง.
12:52
This is Kareem's dark eyes
227
772330
2000
ืืœื• ื”ืŸ ื”ืขื™ื ื™ื™ื ื”ื›ื”ื•ืช ืฉืœ ืงืืจื™ื
12:54
and that's the white background that you see here.
228
774330
3000
ื•ื–ื” ื”ืจืงืข ื”ืœื‘ืŸ ืฉืืชื ืจื•ืื™ื ื›ืืŸ.
12:57
And then it also extracts movement.
229
777330
2000
ื•ืื– ื”ื•ื ื’ื ืžื—ืœืฅ ืชื ื•ืขื”.
12:59
When Kareem moves his head to the right,
230
779330
2000
ื›ืฉืงืืจื™ื ืžื–ื™ื– ืืช ืจืืฉื• ืœื™ืžื™ืŸ,
13:01
you will see this blue activity there;
231
781330
2000
ืืชื ืชืจืื• ืืช ื”ืคืขื™ืœื•ืช ื”ื›ื—ื•ืœื” ืคื”,
13:03
it represents regions where the contrast is increasing in the image,
232
783330
3000
ื–ื” ืžื™ื™ืฆื’ ืื–ื•ืจื™ื ื‘ื”ื ื”ื ื™ื’ื•ื“ื™ื•ืช ื’ื“ืœื” ื‘ืชืžื•ื ื”,
13:06
that's where it's going from dark to light.
233
786330
3000
ื–ื” ื”ื™ื›ืŸ ืฉื–ื” ืžืฉืชื ื” ืžื›ื”ื” ืœื‘ื”ื™ืจ.
13:09
And you also see this yellow activity,
234
789330
2000
ื•ืืชื ื’ื ืจื•ืื™ื ืืช ื”ืคืขื™ืœื•ืช ื”ืฆื”ื•ื‘ื” ื”ื–ืืช,
13:11
which represents regions where contrast is decreasing;
235
791330
4000
ืฉืžื™ื™ืฆื’ืช ืื–ื•ืจื™ื ื‘ื”ื ื”ื ื™ื’ื•ื“ื™ื•ืช ืงื˜ื ื”,
13:15
it's going from light to dark.
236
795330
2000
ื–ื” ืžืฉืชื ื” ืžื‘ื”ื™ืจ ืœื›ื”ื”.
13:17
And these four types of information --
237
797330
3000
ื•ืืจื‘ืขื” ืกื•ื’ื™ ืžื™ื“ืข ืืœื” --
13:20
your optic nerve has about a million fibers in it,
238
800330
4000
ื‘ืขืฆื‘ ื”ืจืื™ื™ื” ืฉืœื›ื ื™ืฉ ื‘ืขืจืš ืžื™ืœื™ื•ืŸ ืกื™ื‘ื™ื,
13:24
and 900,000 of those fibers
239
804330
3000
ื•-900,000 ืžืชื•ืš ื”ืกื™ื‘ื™ื ื”ืืœื”
13:27
send these four types of information.
240
807330
2000
ืฉื•ืœื—ื™ื ืืช ืืจื‘ืข ื”ืกื•ื’ื™ื ื”ืืœื” ืฉืœ ืื™ื ืคื•ืจืžืฆื™ื”.
13:29
So we are really duplicating the kind of signals that you have on the optic nerve.
241
809330
4000
ืื– ืื ื—ื ื• ืœืžืขืฉื” ืžืฉื›ืคืœื™ื ืืช ื”ืื•ืชื•ืช ืฉื™ืฉ ืœื›ื ื‘ืขืฆื‘ ื”ืจืื™ื”.
13:33
What you notice here is that these snapshots
242
813330
3000
ืžื” ืฉืืชื ืฉืžื™ื ืืœื™ื• ืœื‘ ื›ืืŸ ื”ื•ื ืฉื”ืชืžื•ื ื•ืช
13:36
taken from the output of the retina chip are very sparse, right?
243
816330
4000
ืฉื ืœืงื—ื• ืžื”ืคืœื˜ ืฉืœ ืฉื‘ื‘ ื”ืจืฉืชื™ืช ื”ืŸ ืžืื•ื“ ื“ืœื™ืœื•ืช.
13:40
It doesn't light up green everywhere in the background,
244
820330
2000
ื–ื” ืœื ืฆื‘ื•ืข ื‘ื™ืจื•ืง ื‘ื›ืœ ืžืงื•ื ื‘ืจืงืข,
13:42
only on the edges, and then in the hair, and so on.
245
822330
3000
ืืœื ืจืง ื‘ืงืฆื•ื•ืช, ื•ื›ืš ื”ืœืื”.
13:45
And this is the same thing you see
246
825330
1000
ื•ื–ื” ืื•ืชื• ื”ื“ื‘ืจ ืฉืืชื ืจื•ืื™ื
13:46
when people compress video to send: they want to make it very sparse,
247
826330
4000
ื›ืฉืื ืฉื™ื ืžื›ื•ื•ืฆื™ื ื•ื™ื“ืื• ืœืฉืœื™ื—ื”: ื”ื ืจื•ืฆื™ื ืœืขืฉื•ืช ืื•ืช ื–ื” ืžืื•ื“ ื“ืœื™ืœ,
13:50
because that file is smaller. And this is what the retina is doing,
248
830330
3000
ื›ื™ื•ื•ืŸ ืฉื”ืงื•ื‘ืฅ ื”ื–ื” ืงื˜ืŸ ื™ื•ืชืจ. ื•ื–ื” ืžื” ืฉื”ืจืฉืชื™ืช ืขื•ืฉื”,
13:53
and it's doing it just with the circuitry, and how this network of neurons
249
833330
4000
ื•ื”ื™ื ืขื•ืฉื” ื–ืืช ืจืง ื‘ืืžืฆืขื•ืช ื”ืžืขื’ืœื™ื, ื•ืื™ืš ืฉื”ืจืฉืช ื”ื–ืืช ืฉืœ ื ื™ื•ืจื•ื ื™ื
13:57
that are interacting in there, which we've captured on the chip.
250
837330
3000
ืฉืžืชืงืฉืจื™ื ืฉื ื‘ืคื ื™ื, ืฉืื ื—ื ื• ืชืคืกื ื• ืขืœ ื”ืฉื‘ื‘.
14:00
But the point that I want to make -- I'll show you up here.
251
840330
3000
ืื‘ืœ ื”ื ืงื•ื“ื” ืฉืื ื™ ืจื•ืฆื” ืœื”ืขื‘ื™ืจ, ืื ื™ ืืจืื” ืœื›ื ื›ืืŸ ืœืžืขืœื”.
14:03
So this image here is going to look like these ones,
252
843330
3000
ืื– ื”ืชืžื•ื ื” ื”ื–ื• ื›ืืŸ ืขื•ืžื“ืช ืœื”ื™ืจืื•ืช ื›ืžื• ืืœื”,
14:06
but here I'll show you that we can reconstruct the image,
253
846330
2000
ืื‘ืœ ื›ืืŸ ืื ื™ ืืจืื” ืœื›ื ืฉืื ื—ื ื• ื™ื›ื•ืœื™ื ืœื”ืจื›ื™ื‘ ืžื—ื“ืฉ ืืช ื”ืชืžื•ื ื”,
14:08
so, you know, you can almost recognize Kareem in that top part there.
254
848330
5000
ืื–, ืืชื ื™ื•ื“ืขื™ื, ืืชื ื›ืžืขื˜ ื™ื›ื•ืœื™ื ืœื–ื”ื•ืช ืืช ืงืืจื™ื ื‘ื—ืœืง ื”ืขืœื™ื•ืŸ ื”ื–ื” ืฉื.
14:13
And so, here you go.
255
853330
2000
ื”ื ื”.
14:24
Yes, so that's the idea.
256
864330
3000
ื›ืŸ, ืื– ื–ื” ื”ืจืขื™ื•ืŸ.
14:27
When you stand still, you just see the light and dark contrasts.
257
867330
2000
ื›ืฉืืชื ืœื ื–ื–ื™ื, ืืชื ืจืง ืจื•ืื™ื ืืช ื”ื ื™ื’ื•ื“ื™ื•ืช ื”ื›ื”ื” ื•ื”ื‘ื”ื™ืจื”.
14:29
But when it's moving back and forth,
258
869330
2000
ืื‘ืœ ื›ืฉืืชื ื–ื–ื™ื ื”ืœื•ืš ื•ืฉื•ื‘,
14:31
the retina picks up these changes.
259
871330
3000
ื”ืจืฉืชื™ืช ืงื•ืœื˜ืช ืืช ื”ืฉื™ื ื•ื™ื™ื ื”ืืœื”.
14:34
And that's why, you know, when you're sitting here
260
874330
1000
ื•ื–ื• ื”ืกื™ื‘ื”, ืืชื ื™ื•ื“ืขื™ื, ื›ืฉืืชื ื™ื•ืฉื‘ื™ื ื›ืืŸ
14:35
and something happens in your background,
261
875330
2000
ื•ืžืฉื”ื• ืงื•ืจื” ื‘ืจืงืข ืฉืœื›ื,
14:37
you merely move your eyes to it.
262
877330
2000
ืืชื ืจืง ืžื–ื™ื–ื™ื ืืช ื”ืขื™ื ื™ื™ื ืืœื™ื•.
14:39
There are these cells that detect change
263
879330
2000
ื™ืฉ ืืช ื”ืชืื™ื ื”ืืœื” ืฉืžื–ื”ื™ื ืฉื™ื ื•ื™ื™ื
14:41
and you move your attention to it.
264
881330
2000
ื•ืืชื ืžืกื™ื˜ื™ื ืืช ืชืฉื•ืžืช ื”ืœื‘ ืฉืœื›ื ืœื–ื”.
14:43
So those are very important for catching somebody
265
883330
2000
ืื– ืืœื• ืžืื•ื“ ื—ืฉื•ื‘ื™ื ื‘ืฉื‘ื™ืœ ืœืชืคื•ืก ืžื™ืฉื”ื•
14:45
who's trying to sneak up on you.
266
885330
2000
ืฉืžื ืกื” ืœื”ืชื’ื ื‘ ืžืื—ื•ืจื™ื›ื.
14:47
Let me just end by saying that this is what happens
267
887330
3000
ืชื ื• ืœื™ ืจืง ืœืกื™ื™ื ื•ืœื”ื’ื™ื“ ืœื›ื ืฉื–ื” ืžื” ืฉืงื•ืจื”
14:50
when you put Africa in a piano, OK.
268
890330
3000
ื›ืฉืืชื” ืžื›ื ื™ืก ืืช ืืคืจื™ืงื” ืœืคืกื ืชืจ, ืื•ืงื™.
14:53
This is a steel drum here that has been modified,
269
893330
3000
ื–ื”ื• ืชื•ืฃ ื‘ืจื–ืœ ื›ืืŸ ืฉืขื‘ืจ ืฉื™ื ื•ื™,
14:56
and that's what happens when you put Africa in a piano.
270
896330
3000
ื•ื–ื” ืžื” ืฉืงื•ืจื” ื›ืฉืืชื” ืžื›ื ื™ืก ืืช ืืคืจื™ืงื” ืœืคืกื ืชืจ.
14:59
And what I would like us to do is put Africa in the computer,
271
899330
4000
ื•ืžื” ืฉืื ื™ ื”ื™ื™ืชื™ ืจื•ืฆื” ืฉื ืขืฉื”, ื–ื” ืœื”ื›ื ื™ืก ืืช ืืคืจื™ืงื” ืœืžื—ืฉื‘,
15:03
and come up with a new kind of computer
272
903330
2000
ื•ืœื”ืžืฆื™ื ืกื•ื’ ื—ื“ืฉ ืฉืœ ืžื—ืฉื‘
15:05
that will generate thought, imagination, be creative and things like that.
273
905330
3000
ืฉื™ื™ืฆืจ ืžื—ืฉื‘ื•ืช, ื“ืžื™ื•ืŸ, ื™ืฆื™ืจืชื™ื•ืช ื•ื“ื‘ืจื™ื ื›ืืœื”.
15:08
Thank you.
274
908330
2000
ืชื•ื“ื”.
15:10
(Applause)
275
910330
2000
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)
15:12
Chris Anderson: Question for you, Kwabena.
276
912330
2000
ื›ืจื™ืก ืื ื“ืจืกื•ืŸ: ืฉืืœื” ืืœื™ื™ืš, ืงื•ื•ืื‘ื ื”.
15:14
Do you put together in your mind the work you're doing,
277
914330
4000
ื”ืื ืืชื” ืžื—ื‘ืจ ื‘ื“ืžื™ื•ื ืš ืืช ื”ืขื‘ื•ื“ื” ืฉืืชื” ืขื•ืฉื”,
15:18
the future of Africa, this conference --
278
918330
3000
ื”ืขืชื™ื“ ืฉืœ ืืคืจื™ืงื”, ื”ื›ื ืก ื”ื–ื” --
15:21
what connections can we make, if any, between them?
279
921330
3000
ืื™ื–ื” ืงื™ืฉื•ืจื™ื ืื ื—ื ื• ื™ื›ื•ืœื™ื ืœืขืฉื•ืช, ืื ื‘ื›ืœืœ, ื‘ื™ื ื™ื”ื?
15:24
Kwabena Boahen: Yes, like I said at the beginning,
280
924330
2000
ืงื•ื•ืื‘ื ื” ื‘ื•ืื”ืŸ: ื›ืŸ. ื›ืžื• ืฉืืžืจืชื™ ื‘ื”ืชื—ืœื”.
15:26
I got my first computer when I was a teenager, growing up in Accra.
281
926330
4000
ืงื™ื‘ืœืชื™ ืืช ื”ืžื—ืฉื‘ ื”ืจืืฉื•ืŸ ืฉืœื™ ื›ืฉื”ื™ื™ืชื™ ื ืขืจ ืžืชื‘ื’ืจ ื‘ืืงืจื”.
15:30
And I had this gut reaction that this was the wrong way to do it.
282
930330
4000
ื•ื”ื™ื™ืชื” ืœื™ ืชื—ื•ืฉืช ื‘ื˜ืŸ ืฉื–ื• ืœื ื”ื“ืจืš ื”ื ื›ื•ื ื” ืœืขืฉื•ืช ื–ืืช.
15:34
It was very brute force; it was very inelegant.
283
934330
3000
ื–ื” ื”ื™ื” ืžืื•ื“ ื›ื•ื—ื ื™, ื–ื” ื”ื™ื” ืžืื•ื“ ืœื ืืœื’ื ื˜ื™.
15:37
I don't think that I would've had that reaction,
284
937330
2000
ืื ื™ ืœื ื—ื•ืฉื‘ ืฉื”ื™ืชื” ืœื™ ื”ืชื’ื•ื‘ื” ื”ื–ืืช,
15:39
if I'd grown up reading all this science fiction,
285
939330
3000
ืื ื”ื™ื™ืชื™ ื’ื“ืœ ื‘ืงืจื™ืืช ื›ืœ ื”ืžื“ืข ื”ื‘ื“ื™ื•ื ื™ ื”ื–ื”,
15:42
hearing about RD2D2, whatever it was called, and just -- you know,
286
942330
4000
ืฉื•ืžืข ืขืœ RD2D2, ืื™ืš ืฉืœื ืงืจืื• ืœื–ื”, ื•ืจืง -- ืืชื” ื™ื•ื“ืข,
15:46
buying into this hype about computers.
287
946330
1000
ื ืฉืื‘ ืœืชื•ืš ื”ื”ื’ื–ืžื” ื”ื–ืืช ืขืœ ืžื—ืฉื‘ื™ื.
15:47
I was coming at it from a different perspective,
288
947330
2000
ืื ื™ ื”ืกืชื›ืœืชื™ ืขืœ ื–ื” ืžืคืจืกืคืงื˜ื™ื‘ื” ืื—ืจืช,
15:49
where I was bringing that different perspective
289
949330
2000
ื•ื”ื‘ืืชื™ ืืช ื”ืคืจืกืคืงื˜ื™ื‘ื” ื”ืื—ืจืช ื”ื–ืืช
15:51
to bear on the problem.
290
951330
2000
ืœื”ืชื‘ื•ื ื ื•ืช ืขืœ ื”ื‘ืขื™ื”.
15:53
And I think a lot of people in Africa have this different perspective,
291
953330
3000
ื•ืื ื™ ื—ื•ืฉื‘ ืฉืœื”ืจื‘ื” ืื ืฉื™ื ื‘ืืคืจื™ืงื” ื™ืฉ ืืช ื”ืคืจืกืคืงื˜ื™ื‘ื” ื”ืื—ืจืช ื”ื–ืืช,
15:56
and I think that's going to impact technology.
292
956330
2000
ื•ืื ื™ ื—ื•ืฉื‘ ืฉื–ื” ื”ื•ืœืš ืœื”ืฉืคื™ืข ืขืœ ื”ื˜ื›ื ื•ืœื•ื’ื™ื”.
15:58
And that's going to impact how it's going to evolve.
293
958330
2000
ื•ื–ื” ื”ื•ืœืš ืœื”ืฉืคื™ืข ืขืœ ืฆื•ืจืช ื”ื”ืชืคืชื—ื•ืช ืฉืœื”.
16:00
And I think you're going to be able to see, use that infusion,
294
960330
2000
ื•ืื ื™ ื—ื•ืฉื‘ ืฉืืชื ืชื•ื›ืœื• ืœืจืื•ืช, ืœื”ืฉืชืžืฉ ื‘ื”ื™ืชื•ืš ื”ื–ื”,
16:02
to come up with new things,
295
962330
2000
ื›ื“ื™ ืœื”ืžืฆื™ื ื“ื‘ืจื™ื ื—ื“ืฉื™ื,
16:04
because you're coming from a different perspective.
296
964330
3000
ื›ื™ ืืชื ืžื’ื™ืขื™ื ืžืคืจืกืคืงื˜ื™ื‘ื” ืื—ืจืช.
16:07
I think we can contribute. We can dream like everybody else.
297
967330
4000
ืื ื™ ื—ื•ืฉื‘ ืฉืื ื—ื ื• ื™ื›ื•ืœื™ื ืœืชืจื•ื, ืื ื—ื ื• ื™ื›ื•ืœื™ื ืœื—ืœื•ื ื›ืžื• ื›ืœ ืื—ื“ ืื—ืจ.
16:11
CA: Thanks Kwabena, that was really interesting.
298
971330
2000
ื›ืจื™ืก ืื ื“ืจืกื•ืŸ: ืชื•ื“ื” ืงื•ื•ืื‘ื ื”, ื–ื” ื”ื™ื” ืžืื•ื“ ืžืขื ื™ื™ืŸ.
16:13
Thank you.
299
973330
1000
ืชื•ื“ื” ืœืš.
16:14
(Applause)
300
974330
2000
(ืžื—ื™ืื•ืช ื›ืคื™ื™ื)
ืขืœ ืืชืจ ื–ื”

ืืชืจ ื–ื” ื™ืฆื™ื’ ื‘ืคื ื™ื›ื ืกืจื˜ื•ื ื™ YouTube ื”ืžื•ืขื™ืœื™ื ืœืœื™ืžื•ื“ ืื ื’ืœื™ืช. ืชื•ื›ืœื• ืœืจืื•ืช ืฉื™ืขื•ืจื™ ืื ื’ืœื™ืช ื”ืžื•ืขื‘ืจื™ื ืขืœ ื™ื“ื™ ืžื•ืจื™ื ืžื”ืฉื•ืจื” ื”ืจืืฉื•ื ื” ืžืจื—ื‘ื™ ื”ืขื•ืœื. ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ื”ืžื•ืฆื’ื•ืช ื‘ื›ืœ ื“ืฃ ื•ื™ื“ืื• ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ ืžืฉื. ื”ื›ืชื•ื‘ื™ื•ืช ื’ื•ืœืœื•ืช ื‘ืกื ื›ืจื•ืŸ ืขื ื”ืคืขืœืช ื”ื•ื•ื™ื“ืื•. ืื ื™ืฉ ืœืš ื”ืขืจื•ืช ืื• ื‘ืงืฉื•ืช, ืื ื ืฆื•ืจ ืื™ืชื ื• ืงืฉืจ ื‘ืืžืฆืขื•ืช ื˜ื•ืคืก ื™ืฆื™ืจืช ืงืฉืจ ื–ื”.

https://forms.gle/WvT1wiN1qDtmnspy7