TEDGlobal 2010
Sugata Mitra: The child-driven education
Filmed:
Readability: 3.4
3,097,850 views
Education scientist Sugata Mitra tackles one of the greatest problems of education -- the best teachers and schools don't exist where they're needed most. In a series of real-life experiments from New Delhi to South Africa to Italy, he gave kids self-supervised access to the web and saw results that could revolutionize how we think about teaching.
Sugata Mitra - Education researcher
Educational researcher Sugata Mitra is the winner of the 2013 TED Prize. His wish: Build a School in the Cloud, where children can explore and learn from one another. Full bio
Educational researcher Sugata Mitra is the winner of the 2013 TED Prize. His wish: Build a School in the Cloud, where children can explore and learn from one another. Full bio
Double-click the English transcript below to play the video.
00:17
Well, that's kind of an obvious statement up there.
0
2000
2000
00:19
I started with that sentence about 12 years ago,
1
4000
3000
00:22
and I started in the context
2
7000
3000
00:25
of developing countries,
3
10000
2000
00:27
but you're sitting here from every corner of the world.
4
12000
3000
00:30
So if you think of a map of your country,
5
15000
3000
00:33
I think you'll realize
6
18000
2000
00:35
that for every country on Earth,
7
20000
2000
00:37
you could draw little circles to say,
8
22000
2000
00:39
"These are places where good teachers won't go."
9
24000
3000
00:43
On top of that,
10
28000
2000
00:45
those are the places from where trouble comes.
11
30000
3000
00:48
So we have an ironic problem --
12
33000
2000
00:50
good teachers don't want to go
13
35000
2000
00:52
to just those places where they're needed the most.
14
37000
3000
00:55
I started in 1999
15
40000
3000
00:58
to try and address this problem with an experiment,
16
43000
3000
01:01
which was a very simple experiment in New Delhi.
17
46000
3000
01:06
I basically embedded a computer
18
51000
3000
01:09
into a wall of a slum in New Delhi.
19
54000
3000
01:13
The children barely went to school, they didn't know any English --
20
58000
3000
01:16
they'd never seen a computer before,
21
61000
2000
01:18
and they didn't know what the internet was.
22
63000
3000
01:21
I connected high speed internet to it -- it's about three feet off the ground --
23
66000
3000
01:24
turned it on and left it there.
24
69000
2000
01:26
After this,
25
71000
2000
01:28
we noticed a couple of interesting things, which you'll see.
26
73000
3000
01:31
But I repeated this all over India
27
76000
3000
01:34
and then through
28
79000
2000
01:36
a large part of the world
29
81000
2000
01:38
and noticed
30
83000
2000
01:40
that children will learn to do
31
85000
2000
01:42
what they want to learn to do.
32
87000
3000
01:45
This is the first experiment that we did --
33
90000
2000
01:47
eight year-old boy on your right
34
92000
2000
01:49
teaching his student, a six year-old girl,
35
94000
3000
01:52
and he was teaching her how to browse.
36
97000
3000
01:56
This boy here in the middle of central India --
37
101000
3000
02:00
this is in a Rajasthan village,
38
105000
2000
02:02
where the children recorded their own music
39
107000
3000
02:05
and then played it back to each other
40
110000
3000
02:08
and in the process,
41
113000
2000
02:10
they've enjoyed themselves thoroughly.
42
115000
2000
02:12
They did all of this in four hours
43
117000
2000
02:14
after seeing the computer for the first time.
44
119000
3000
02:17
In another South Indian village,
45
122000
3000
02:20
these boys here
46
125000
2000
02:22
had assembled a video camera
47
127000
2000
02:24
and were trying to take the photograph of a bumble bee.
48
129000
2000
02:26
They downloaded it from Disney.com,
49
131000
2000
02:28
or one of these websites,
50
133000
2000
02:30
14 days after putting the computer in their village.
51
135000
3000
02:36
So at the end of it,
52
141000
2000
02:38
we concluded that groups of children
53
143000
2000
02:40
can learn to use computers and the internet on their own,
54
145000
3000
02:43
irrespective of who
55
148000
2000
02:45
or where they were.
56
150000
3000
02:48
At that point, I became a little more ambitious
57
153000
3000
02:51
and decided to see
58
156000
3000
02:54
what else could children do with a computer.
59
159000
3000
02:57
We started off with an experiment in Hyderabad, India,
60
162000
3000
03:00
where I gave a group of children --
61
165000
3000
03:03
they spoke English with a very strong Telugu accent.
62
168000
3000
03:06
I gave them a computer
63
171000
2000
03:08
with a speech-to-text interface,
64
173000
2000
03:10
which you now get free with Windows,
65
175000
3000
03:13
and asked them to speak into it.
66
178000
2000
03:15
So when they spoke into it,
67
180000
2000
03:17
the computer typed out gibberish,
68
182000
2000
03:19
so they said, "Well, it doesn't understand anything of what we are saying."
69
184000
2000
03:21
So I said, "Yeah, I'll leave it here for two months.
70
186000
2000
03:23
Make yourself understood
71
188000
2000
03:25
to the computer."
72
190000
2000
03:27
So the children said, "How do we do that."
73
192000
2000
03:29
And I said,
74
194000
2000
03:31
"I don't know, actually."
75
196000
2000
03:33
(Laughter)
76
198000
2000
03:35
And I left.
77
200000
2000
03:37
(Laughter)
78
202000
2000
03:40
Two months later --
79
205000
2000
03:42
and this is now documented
80
207000
2000
03:44
in the Information Technology
81
209000
2000
03:46
for International Development journal --
82
211000
2000
03:48
that accents had changed
83
213000
2000
03:50
and were remarkably close to the neutral British accent
84
215000
3000
03:53
in which I had trained the speech-to-text synthesizer.
85
218000
3000
03:56
In other words, they were all speaking like James Tooley.
86
221000
3000
03:59
(Laughter)
87
224000
2000
04:01
So they could do that on their own.
88
226000
2000
04:03
After that, I started to experiment
89
228000
2000
04:05
with various other things
90
230000
2000
04:07
that they might learn to do on their own.
91
232000
2000
04:09
I got an interesting phone call once from Columbo,
92
234000
3000
04:12
from the late Arthur C. Clarke,
93
237000
2000
04:14
who said, "I want to see what's going on."
94
239000
2000
04:16
And he couldn't travel, so I went over there.
95
241000
3000
04:19
He said two interesting things,
96
244000
2000
04:21
"A teacher that can be replaced by a machine should be."
97
246000
5000
04:26
(Laughter)
98
251000
2000
04:28
The second thing he said was that,
99
253000
2000
04:30
"If children have interest,
100
255000
2000
04:32
then education happens."
101
257000
3000
04:35
And I was doing that in the field,
102
260000
2000
04:37
so every time I would watch it and think of him.
103
262000
2000
04:39
(Video) Arthur C. Clarke: And they can definitely
104
264000
3000
04:42
help people,
105
267000
2000
04:44
because children quickly learn to navigate
106
269000
2000
04:46
the web and find things which interest them.
107
271000
3000
04:49
And when you've got interest, then you have education.
108
274000
3000
04:52
Sugata Mitra: I took the experiment to South Africa.
109
277000
3000
04:55
This is a 15 year-old boy.
110
280000
2000
04:57
(Video) Boy: ... just mention, I play games
111
282000
3000
05:00
like animals,
112
285000
3000
05:03
and I listen to music.
113
288000
3000
05:06
SM: And I asked him, "Do you send emails?"
114
291000
2000
05:08
And he said, "Yes, and they hop across the ocean."
115
293000
3000
05:12
This is in Cambodia,
116
297000
2000
05:14
rural Cambodia --
117
299000
3000
05:17
a fairly silly arithmetic game,
118
302000
3000
05:20
which no child would play inside the classroom or at home.
119
305000
2000
05:22
They would, you know, throw it back at you.
120
307000
2000
05:24
They'd say, "This is very boring."
121
309000
2000
05:26
If you leave it on the pavement
122
311000
2000
05:28
and if all the adults go away,
123
313000
2000
05:30
then they will show off with each other
124
315000
2000
05:32
about what they can do.
125
317000
2000
05:34
This is what these children are doing.
126
319000
2000
05:36
They are trying to multiply, I think.
127
321000
3000
05:39
And all over India,
128
324000
2000
05:41
at the end of about two years,
129
326000
2000
05:43
children were beginning to Google their homework.
130
328000
3000
05:46
As a result, the teachers reported
131
331000
2000
05:48
tremendous improvements in their English --
132
333000
2000
05:50
(Laughter)
133
335000
4000
05:54
rapid improvement and all sorts of things.
134
339000
2000
05:56
They said, "They have become really deep thinkers and so on and so forth.
135
341000
3000
05:59
(Laughter)
136
344000
3000
06:02
And indeed they had.
137
347000
2000
06:04
I mean, if there's stuff on Google,
138
349000
2000
06:06
why would you need to stuff it into your head?
139
351000
3000
06:10
So at the end of the next four years,
140
355000
2000
06:12
I decided that groups of children can navigate the internet
141
357000
3000
06:15
to achieve educational objectives on their own.
142
360000
3000
06:18
At that time, a large amount of money
143
363000
2000
06:20
had come into Newcastle University
144
365000
2000
06:22
to improve schooling in India.
145
367000
3000
06:25
So Newcastle gave me a call. I said, "I'll do it from Delhi."
146
370000
3000
06:28
They said, "There's no way you're going to handle
147
373000
2000
06:30
a million pounds-worth of University money
148
375000
3000
06:33
sitting in Delhi."
149
378000
2000
06:35
So in 2006,
150
380000
2000
06:37
I bought myself a heavy overcoat
151
382000
2000
06:39
and moved to Newcastle.
152
384000
2000
06:42
I wanted to test the limits
153
387000
2000
06:44
of the system.
154
389000
2000
06:46
The first experiment I did out of Newcastle
155
391000
2000
06:48
was actually done in India.
156
393000
2000
06:50
And I set myself and impossible target:
157
395000
3000
06:53
can Tamil speaking
158
398000
3000
06:56
12-year-old children
159
401000
2000
06:58
in a South Indian village
160
403000
3000
07:01
teach themselves biotechnology
161
406000
2000
07:03
in English on their own?
162
408000
2000
07:05
And I thought, I'll test them, they'll get a zero --
163
410000
3000
07:08
I'll give the materials, I'll come back and test them --
164
413000
2000
07:10
they get another zero,
165
415000
2000
07:12
I'll go back and say, "Yes, we need teachers for certain things."
166
417000
4000
07:16
I called in 26 children.
167
421000
2000
07:18
They all came in there, and I told them
168
423000
2000
07:20
that there's some really difficult stuff on this computer.
169
425000
2000
07:22
I wouldn't be surprised if you didn't understand anything.
170
427000
3000
07:25
It's all in English, and I'm going.
171
430000
3000
07:28
(Laughter)
172
433000
2000
07:30
So I left them with it.
173
435000
2000
07:32
I came back after two months,
174
437000
2000
07:34
and the 26 children marched in looking very, very quiet.
175
439000
2000
07:36
I said, "Well, did you look at any of the stuff?"
176
441000
3000
07:39
They said, "Yes, we did."
177
444000
2000
07:41
"Did you understand anything?" "No, nothing."
178
446000
3000
07:44
So I said,
179
449000
2000
07:46
"Well, how long did you practice on it
180
451000
2000
07:48
before you decided you understood nothing?"
181
453000
2000
07:50
They said, "We look at it every day."
182
455000
3000
07:53
So I said, "For two months, you were looking at stuff you didn't understand?"
183
458000
2000
07:55
So a 12 year-old girl raises her hand and says,
184
460000
2000
07:57
literally,
185
462000
2000
08:00
"Apart from the fact that improper replication of the DNA molecule
186
465000
3000
08:03
causes genetic disease,
187
468000
2000
08:05
we've understood nothing else."
188
470000
2000
08:07
(Laughter)
189
472000
2000
08:09
(Applause)
190
474000
7000
08:16
(Laughter)
191
481000
3000
08:19
It took me three years to publish that.
192
484000
2000
08:21
It's just been published in the British Journal of Educational Technology.
193
486000
3000
08:24
One of the referees who refereed the paper said,
194
489000
3000
08:27
"It's too good to be true,"
195
492000
3000
08:30
which was not very nice.
196
495000
2000
08:32
Well, one of the girls had taught herself
197
497000
2000
08:34
to become the teacher.
198
499000
2000
08:36
And then that's her over there.
199
501000
2000
08:46
Remember, they don't study English.
200
511000
2000
09:01
I edited out the last bit when I asked, "Where is the neuron?"
201
526000
3000
09:04
and she says, "The neuron? The neuron,"
202
529000
2000
09:06
and then she looked and did this.
203
531000
3000
09:09
Whatever the expression, it was not very nice.
204
534000
3000
09:12
So their scores had gone up from zero to 30 percent,
205
537000
3000
09:15
which is an educational impossibility under the circumstances.
206
540000
3000
09:18
But 30 percent is not a pass.
207
543000
3000
09:21
So I found that they had a friend,
208
546000
2000
09:23
a local accountant, a young girl,
209
548000
2000
09:25
and they played football with her.
210
550000
2000
09:27
I asked that girl, "Would you teach them
211
552000
2000
09:29
enough biotechnology to pass?"
212
554000
2000
09:31
And she said, "How would I do that? I don't know the subject."
213
556000
2000
09:33
I said, "No, use the method of the grandmother."
214
558000
2000
09:35
She said, "What's that?"
215
560000
2000
09:37
I said, "Well, what you've got to do
216
562000
2000
09:39
is stand behind them
217
564000
2000
09:41
and admire them all the time.
218
566000
3000
09:44
Just say to them, 'That's cool. That's fantastic.
219
569000
2000
09:46
What is that? Can you do that again? Can you show me some more?'"
220
571000
3000
09:49
She did that for two months.
221
574000
2000
09:51
The scores went up to 50,
222
576000
2000
09:53
which is what the posh schools of New Delhi,
223
578000
2000
09:55
with a trained biotechnology teacher were getting.
224
580000
3000
09:58
So I came back to Newcastle
225
583000
2000
10:00
with these results
226
585000
2000
10:02
and decided
227
587000
2000
10:04
that there was something happening here
228
589000
2000
10:06
that definitely was getting very serious.
229
591000
3000
10:10
So, having experimented in all sorts of remote places,
230
595000
3000
10:13
I came to the most remote place that I could think of.
231
598000
3000
10:16
(Laughter)
232
601000
2000
10:19
Approximately 5,000 miles from Delhi
233
604000
3000
10:22
is the little town of Gateshead.
234
607000
2000
10:24
In Gateshead, I took 32 children
235
609000
3000
10:27
and I started to fine-tune the method.
236
612000
3000
10:30
I made them into groups of four.
237
615000
3000
10:33
I said, "You make your own groups of four.
238
618000
2000
10:35
Each group of four can use one computer and not four computers."
239
620000
3000
10:38
Remember, from the Hole in the Wall.
240
623000
3000
10:41
"You can exchange groups.
241
626000
2000
10:43
You can walk across to another group,
242
628000
2000
10:45
if you don't like your group, etc.
243
630000
2000
10:47
You can go to another group, peer over their shoulders, see what they're doing,
244
632000
3000
10:50
come back to you own group and claim it as your own work."
245
635000
3000
10:53
And I explained to them
246
638000
2000
10:55
that, you know, a lot of scientific research is done using that method.
247
640000
3000
10:58
(Laughter)
248
643000
2000
11:00
(Applause)
249
645000
5000
11:07
The children enthusiastically got after me and said,
250
652000
2000
11:09
"Now, what do you want us to do?"
251
654000
2000
11:11
I gave them six GCSE questions.
252
656000
3000
11:14
The first group -- the best one --
253
659000
2000
11:16
solved everything in 20 minutes.
254
661000
2000
11:18
The worst, in 45.
255
663000
3000
11:21
They used everything that they knew --
256
666000
2000
11:23
news groups, Google, Wikipedia,
257
668000
2000
11:25
Ask Jeeves, etc.
258
670000
2000
11:27
The teachers said, "Is this deep learning?"
259
672000
3000
11:30
I said, "Well, let's try it.
260
675000
2000
11:32
I'll come back after two months.
261
677000
2000
11:34
We'll give them a paper test --
262
679000
2000
11:36
no computers, no talking to each other, etc."
263
681000
2000
11:38
The average score when I'd done it with the computers and the groups
264
683000
2000
11:40
was 76 percent.
265
685000
2000
11:42
When I did the experiment, when I did the test,
266
687000
2000
11:44
after two months, the score
267
689000
3000
11:47
was 76 percent.
268
692000
3000
11:50
There was photographic recall
269
695000
2000
11:52
inside the children,
270
697000
2000
11:54
I suspect because they're discussing with each other.
271
699000
3000
11:57
A single child in front of a single computer
272
702000
2000
11:59
will not do that.
273
704000
2000
12:01
I have further results,
274
706000
2000
12:03
which are almost unbelievable,
275
708000
2000
12:05
of scores which go up with time.
276
710000
2000
12:07
Because their teachers say
277
712000
2000
12:09
that after the session is over,
278
714000
2000
12:11
the children continue to Google further.
279
716000
3000
12:14
Here in Britain, I put out a call
280
719000
2000
12:16
for British grandmothers,
281
721000
2000
12:18
after my Kuppam experiment.
282
723000
2000
12:20
Well, you know,
283
725000
2000
12:22
they're very vigorous people, British grandmothers.
284
727000
2000
12:24
200 of them volunteered immediately.
285
729000
2000
12:26
(Laughter)
286
731000
2000
12:28
The deal was that they would give me
287
733000
3000
12:31
one hour of broadband time,
288
736000
2000
12:33
sitting in their homes,
289
738000
2000
12:35
one day in a week.
290
740000
2000
12:37
So they did that,
291
742000
2000
12:39
and over the last two years,
292
744000
2000
12:41
over 600 hours of instruction
293
746000
2000
12:43
has happened over Skype,
294
748000
2000
12:45
using what my students call the granny cloud.
295
750000
3000
12:48
The granny cloud sits over there.
296
753000
3000
12:51
I can beam them to whichever school I want to.
297
756000
3000
13:00
(Video) Teacher: You can't catch me.
298
765000
2000
13:02
You say it.
299
767000
3000
13:05
You can't catch me.
300
770000
3000
13:08
Children: You can't catch me.
301
773000
3000
13:11
Teacher: I'm the gingerbread man.
302
776000
3000
13:14
Children: I'm the gingerbread man.
303
779000
2000
13:16
Teacher: Well done. Very good ...
304
781000
2000
13:24
SM: Back at Gateshead,
305
789000
2000
13:26
a 10-year-old girl gets into the heart of Hinduism
306
791000
2000
13:28
in 15 minutes.
307
793000
2000
13:30
You know, stuff which I don't know anything about.
308
795000
3000
13:36
Two children watch a TEDTalk.
309
801000
2000
13:38
They wanted to be footballers before.
310
803000
2000
13:40
After watching eight TEDTalks,
311
805000
2000
13:42
he wants to become Leonardo da Vinci.
312
807000
3000
13:45
(Laughter)
313
810000
3000
13:48
(Applause)
314
813000
3000
13:51
It's pretty simple stuff.
315
816000
2000
13:53
This is what I'm building now --
316
818000
2000
13:55
they're called SOLEs: Self Organized Learning Environments.
317
820000
3000
13:58
The furniture is designed
318
823000
2000
14:00
so that children can sit in front of big, powerful screens,
319
825000
3000
14:03
big broadband connections, but in groups.
320
828000
3000
14:06
If they want, they can call the granny cloud.
321
831000
3000
14:09
This is a SOLE in Newcastle.
322
834000
2000
14:11
The mediator is from Pune, India.
323
836000
2000
14:13
So how far can we go? One last little bit and I'll stop.
324
838000
3000
14:16
I went to Turin in May.
325
841000
3000
14:20
I sent all the teachers away from my group of 10 year-old students.
326
845000
3000
14:24
I speak only English, they speak only Italian,
327
849000
3000
14:27
so we had no way to communicate.
328
852000
2000
14:29
I started writing English questions on the blackboard.
329
854000
3000
14:33
The children looked at it and said, "What?"
330
858000
2000
14:35
I said, "Well, do it."
331
860000
2000
14:37
They typed it into Google, translated it into Italian,
332
862000
3000
14:40
went back into Italian Google.
333
865000
2000
14:42
Fifteen minutes later --
334
867000
3000
14:52
next question: where is Calcutta?
335
877000
3000
14:57
This one, they took only 10 minutes.
336
882000
3000
15:04
I tried a really hard one then.
337
889000
3000
15:07
Who was Pythagoras, and what did he do?
338
892000
3000
15:12
There was silence for a while,
339
897000
2000
15:14
then they said, "You've spelled it wrong.
340
899000
2000
15:16
It's Pitagora."
341
901000
3000
15:23
And then,
342
908000
2000
15:25
in 20 minutes,
343
910000
2000
15:27
the right-angled triangles began to appear on the screens.
344
912000
2000
15:29
This sent shivers up my spine.
345
914000
3000
15:32
These are 10 year-olds.
346
917000
2000
15:47
Text: In another 30 minutes they would reach the Theory of Relativity. And then?
347
932000
3000
15:50
(Laughter)
348
935000
2000
15:52
(Applause)
349
937000
9000
16:01
SM: So you know what's happened?
350
946000
2000
16:03
I think we've just stumbled across
351
948000
2000
16:05
a self-organizing system.
352
950000
2000
16:07
A self-organizing system is one
353
952000
2000
16:09
where a structure appears
354
954000
2000
16:11
without explicit intervention from the outside.
355
956000
3000
16:14
Self-organizing systems also always show emergence,
356
959000
3000
16:17
which is that the system starts to do things,
357
962000
2000
16:19
which it was never designed for.
358
964000
2000
16:21
Which is why you react the way you do,
359
966000
2000
16:23
because it looks impossible.
360
968000
3000
16:26
I think I can make a guess now --
361
971000
3000
16:29
education is self-organizing system,
362
974000
2000
16:31
where learning is an emergent phenomenon.
363
976000
2000
16:33
It'll take a few years to prove it, experimentally,
364
978000
2000
16:35
but I'm going to try.
365
980000
2000
16:37
But in the meanwhile, there is a method available.
366
982000
3000
16:40
One billion children, we need 100 million mediators --
367
985000
3000
16:43
there are many more than that on the planet --
368
988000
2000
16:45
10 million SOLEs,
369
990000
2000
16:47
180 billion dollars and 10 years.
370
992000
3000
16:51
We could change everything.
371
996000
2000
16:53
Thanks.
372
998000
2000
16:55
(Applause)
373
1000000
11000
ABOUT THE SPEAKER
Sugata Mitra - Education researcherEducational researcher Sugata Mitra is the winner of the 2013 TED Prize. His wish: Build a School in the Cloud, where children can explore and learn from one another.
Why you should listen
In 1999, Sugata Mitra and his colleagues dug a hole in a wall bordering an urban slum in New Delhi, installed an Internet-connected PC and left it there, with a hidden camera filming the area. What they saw: kids from the slum playing with the computer and, in the process, learning how to use it -- then teaching each other. These famed “Hole in the Wall” experiments demonstrated that, in the absence of supervision and formal teaching, children can teach themselves and each other -- if they’re motivated by curiosity. Mitra, now a professor of educational technology at Newcastle University, called it "minimally invasive education."
Mitra thinks self-organized learning will shape the future of education. At TED2013, he made a bold TED Prize wish: Help me build a School in the Cloud where children can explore and learn on their own -- and teach one another -- using resouces from the worldwide cloud.
The School in the Cloud now includes seven physical locations -- five in India and two in the UK. At the same time, the School in the Cloud online platform lets students participate anywhere, with partner learning labs and programs in countries like Colombia, Pakistan and Greece. In 2016, Mitra held the first School in the Cloud conference in India. He shared that more than 16,000 SOLE sessions had taken place so far, with kids all around the world dipping their toes in this new education model.
Sugata Mitra | Speaker | TED.com