ABOUT THE SPEAKER
Sebastian Wernicke - Data scientist
After making a splash in the field of bioinformatics, Sebastian Wernicke moved on to the corporate sphere, where he motivates and manages multidimensional projects.

Why you should listen

Dr. Sebastian Wernicke is the Chief Data Scientist of ONE LOGIC, a data science boutique that supports organizations across industries to make sense of their vast data collections to improve operations and gain strategic advantages. Wernicke originally studied bioinformatics and previously led the strategy and growth of Seven Bridges Genomics, a Cambridge-based startup that builds platforms for genetic analysis.

Before his career in statistics began, Wernicke worked stints as both a paramedic and successful short animated filmmaker. He's also the author of the TEDPad app, an irreverent tool for creating an infinite number of "amazing and really bad" and mostly completely meaningless talks. He's the author of the statistically authoritative and yet completely ridiculous "How to Give the Perfect TEDTalk."

More profile about the speaker
Sebastian Wernicke | Speaker | TED.com
TEDxZurich 2011

Sebastian Wernicke: 1,000 TED Talks in six words

Filmed:
702,149 views

Sebastian Wernicke thinks every TED Talk can be summarized in six words. In this talk, he shows how to do just that -- and less.
- Data scientist
After making a splash in the field of bioinformatics, Sebastian Wernicke moved on to the corporate sphere, where he motivates and manages multidimensional projects. Full bio

Double-click the English transcript below to play the video.

00:15
There's currently over a thousand TEDTalks on the TED website.
0
0
4000
00:19
And I guess many of you here
1
4000
3000
00:22
think that this is quite fantastic --
2
7000
2000
00:24
except for me. I don't agree with this.
3
9000
2000
00:26
I think we have a situation here.
4
11000
2000
00:28
Because if you think about it, 1,000 TEDTalks,
5
13000
3000
00:31
that's over 1,000 ideas worth spreading.
6
16000
3000
00:34
How on earth
7
19000
2000
00:36
are you going to spread a thousand ideas?
8
21000
2000
00:38
Even if you just try to get all of those ideas into your head
9
23000
2000
00:40
by watching all those thousand TED videos,
10
25000
2000
00:42
it would actually currently take you
11
27000
3000
00:45
over 250 hours to do so.
12
30000
2000
00:47
And I did a little calculation of this.
13
32000
2000
00:49
The damage to the economy for each one who does this
14
34000
3000
00:52
is around $15,000.
15
37000
2000
00:54
So having seen this danger to the economy,
16
39000
3000
00:57
I thought, we need to find a solution to this problem.
17
42000
3000
01:00
Here's my approach to it all.
18
45000
2000
01:02
If you look at the current situation,
19
47000
2000
01:04
you have a thousand TEDTalks.
20
49000
2000
01:06
Each of those TEDTalks has an average length
21
51000
2000
01:08
of about 2,300 words.
22
53000
2000
01:10
Now take this together
23
55000
2000
01:12
and you end up with 2.3 million words of TEDTalks,
24
57000
3000
01:15
which is about three Bibles-worth of content.
25
60000
3000
01:18
The obvious question here is,
26
63000
2000
01:20
does a TEDTalk really need 2,300 words?
27
65000
3000
01:23
Isn't there something shorter?
28
68000
2000
01:25
I mean, if you have an idea worth spreading,
29
70000
2000
01:27
surely you can put it into something shorter
30
72000
2000
01:29
than 2,300 words.
31
74000
2000
01:31
The only question is, how short can you get?
32
76000
2000
01:33
What's the minimum amount of words
33
78000
2000
01:35
you would need to do a TEDTalk?
34
80000
2000
01:37
While I was pondering this question,
35
82000
2000
01:39
I came across this urban legend about Ernest Hemingway,
36
84000
3000
01:42
who allegedly said that these six words here:
37
87000
3000
01:45
"For sale: baby shoes, never worn,"
38
90000
3000
01:48
were the best novel he had ever written.
39
93000
3000
01:51
And I also encountered a project called Six-Word Memoirs
40
96000
2000
01:53
where people were asked,
41
98000
2000
01:55
take your whole life and please sum this up into six words, such as these here:
42
100000
3000
01:58
"Found true love, married someone else."
43
103000
2000
02:00
Or "Living in existential vacuum; it sucks."
44
105000
3000
02:03
I actually like that one.
45
108000
2000
02:05
So if a novel can be put into six words
46
110000
3000
02:08
and a whole memoir can be put into six words,
47
113000
3000
02:11
you don't need more than six words for a TEDTalk.
48
116000
3000
02:14
We could have been done by lunch here.
49
119000
2000
02:16
I mean ...
50
121000
3000
02:19
And if you did this for all thousand TEDTalks,
51
124000
2000
02:21
you would get from 2.3 million words down to 6,000.
52
126000
3000
02:24
So I thought this was quite worthwhile.
53
129000
2000
02:26
So I started asking all my friends,
54
131000
2000
02:28
please take your favorite TEDTalk and put that into six words.
55
133000
3000
02:31
So here are some of the results that I received. I think they're quite nice.
56
136000
2000
02:33
For example, Dan Pink's talk on motivation,
57
138000
2000
02:35
which was pretty good if you haven't seen it:
58
140000
2000
02:37
"Drop carrot. Drop stick. Bring meaning."
59
142000
2000
02:39
It's what he's basically talking about in those 18 and a half minutes.
60
144000
3000
02:42
Or some even included references to the speakers,
61
147000
2000
02:44
such as Nathan Myhrvold's speaking style,
62
149000
2000
02:46
or the one of Tim Ferriss,
63
151000
2000
02:48
which might be considered a bit strenuous at times.
64
153000
3000
02:51
The challenge here is, if I try to systematically do this,
65
156000
3000
02:54
I would probably end up with a lot of summaries,
66
159000
2000
02:56
but not with many friends in the end.
67
161000
2000
02:58
So I had to find a different method,
68
163000
2000
03:00
preferably involving total strangers.
69
165000
2000
03:02
And luckily there's a website for that called Mechanical Turk,
70
167000
3000
03:05
which is a website where you can post tasks
71
170000
2000
03:07
that you don't want to do yourself,
72
172000
2000
03:09
such as "Please summarize this text for me in six words."
73
174000
3000
03:12
And I didn't allow any low-cost countries to work on this,
74
177000
3000
03:15
but I found out I could get a six-word summary for just 10 cents,
75
180000
4000
03:19
which I think is a pretty good price.
76
184000
2000
03:21
Even then, unfortunately,
77
186000
2000
03:23
it's not possible to summarize each TEDTalk individually.
78
188000
3000
03:26
Because if you do the math, you have a thousand TEDTalks,
79
191000
2000
03:28
the pay 10 cents each;
80
193000
2000
03:30
you have to do more than one summary for each of those talks,
81
195000
3000
03:33
because some of them will probably be, or are, really bad.
82
198000
3000
03:36
So I would end up paying hundreds of dollars.
83
201000
3000
03:39
So I thought of a different way
84
204000
2000
03:41
by thinking, well, the talks revolve around certain themes.
85
206000
3000
03:44
So what if I don't let people summarize
86
209000
2000
03:46
individual TEDTalks to six words,
87
211000
2000
03:48
but give them 10 TEDTalks at the same time
88
213000
2000
03:50
and say, "Please do a six-word summary for that one."
89
215000
3000
03:53
I would cut my costs by 90 percent.
90
218000
2000
03:55
So for $60,
91
220000
3000
03:58
I could summarize a thousand TEDTalks
92
223000
2000
04:00
into just 600 summaries,
93
225000
2000
04:02
which would actually be quite nice.
94
227000
2000
04:04
Now some of you might actually right now be thinking,
95
229000
2000
04:06
It's downright crazy to have 10 TEDTalks summarized into just six words.
96
231000
3000
04:09
But it's actually not,
97
234000
2000
04:11
because there's an example by statistics professor, Hans Rosling.
98
236000
3000
04:14
I guess many of you have seen one or more of his talks.
99
239000
2000
04:16
He's got eight talks online,
100
241000
2000
04:18
and those talks can basically be summed up into just four words,
101
243000
3000
04:21
because that's all he's basically showing us,
102
246000
2000
04:23
our intuition is really bad.
103
248000
2000
04:25
He always proves us wrong.
104
250000
2000
04:27
So people on the Internet, some didn't do so well.
105
252000
3000
04:30
I mean, when I asked them to summarize the 10 TEDTalks at the same time,
106
255000
2000
04:32
some took the easy route out.
107
257000
2000
04:34
They just had some general comment.
108
259000
3000
04:37
There were others, and I found this quite cheeky.
109
262000
3000
04:40
They used their six words to talk back to me
110
265000
2000
04:42
and ask me if I'd been too much on Google lately.
111
267000
4000
04:46
And finally also, I never understood this,
112
271000
3000
04:49
some people really came up with their own version of the truth.
113
274000
3000
04:52
I don't know any TEDTalk that contains this.
114
277000
3000
04:55
But, oh well. In the end, however,
115
280000
2000
04:57
and this is really amazing,
116
282000
2000
04:59
for each of those 10 TEDTalk clusters that I submitted,
117
284000
2000
05:01
I actually received meaningful summaries.
118
286000
2000
05:03
Here are some of my favorites.
119
288000
2000
05:05
For example, for all the TEDTalks around food,
120
290000
2000
05:07
someone summed this up into: "Food shaping body, brains and environment,"
121
292000
2000
05:09
which I think is pretty good.
122
294000
2000
05:11
Or happiness: "Striving toward happiness =
123
296000
2000
05:13
moving toward unhappiness."
124
298000
2000
05:15
So here I was.
125
300000
2000
05:17
I had started out with a thousand TEDTalks
126
302000
2000
05:19
and I had 600 six-word summaries for those.
127
304000
3000
05:22
Actually it sounded nice in the beginning,
128
307000
2000
05:24
but when you look at 600 summaries, it's quite a lot.
129
309000
2000
05:26
It's a huge list.
130
311000
2000
05:28
So I thought, I probably have to take this one step further here
131
313000
4000
05:32
and create summaries of the summaries -- and this is exactly what I did.
132
317000
3000
05:35
So I took the 600 summaries that I had,
133
320000
2000
05:37
put them into nine groups
134
322000
2000
05:39
according to the ratings that the talks had originally received on TED.com
135
324000
4000
05:43
and asked people to do summaries of those.
136
328000
3000
05:46
Again, there were some misunderstandings.
137
331000
2000
05:48
For example, when I had a cluster of all the beautiful talks,
138
333000
2000
05:50
someone thought I was just trying to find the ultimate pick-up line.
139
335000
3000
05:53
But in the end, amazingly,
140
338000
3000
05:56
again, people were able to do it.
141
341000
2000
05:58
For example, all the courageous TEDTalks:
142
343000
2000
06:00
"People dying," or "People suffering," was also one,
143
345000
2000
06:02
"with easy solutions around."
144
347000
2000
06:04
Or the recipe for the ultimate jaw-dropping TEDTalk:
145
349000
2000
06:06
"Flickr photos of intergalactic classical composer."
146
351000
3000
06:09
I mean that's the essence of it all.
147
354000
3000
06:12
Now I had my nine groups,
148
357000
2000
06:14
but, I mean, it's already quite a reduction.
149
359000
3000
06:17
But of course, once you are that far, you're not really satisfied.
150
362000
3000
06:20
I wanted to go all the way, all the way down the distillery,
151
365000
3000
06:23
starting out with a thousand TEDTalks.
152
368000
2000
06:25
I wanted to have a thousand TEDTalks summarized into just six words --
153
370000
3000
06:28
which would be a 99.9997 percent reduction in content.
154
373000
4000
06:32
And I would only pay $99.50 --
155
377000
3000
06:35
so stay even below a hundred dollars for it.
156
380000
3000
06:38
So I had 50 overall summaries done.
157
383000
2000
06:40
This time I paid 25 cents
158
385000
2000
06:42
because I thought the task was a bit harder.
159
387000
3000
06:45
And unfortunately when I first received the answers --
160
390000
2000
06:47
and here you'll see six of the answers --
161
392000
2000
06:49
I was a bit disappointed.
162
394000
2000
06:51
Because I think you'll agree, they all summarize some aspect of TED,
163
396000
3000
06:54
but to me they felt a bit bland,
164
399000
2000
06:56
or they just had a certain aspect of TED in them.
165
401000
3000
06:59
So I was almost ready to give up
166
404000
3000
07:02
when one night I played around with these sentences
167
407000
2000
07:04
and found out that there's actually a beautiful solution in here.
168
409000
4000
07:08
So here it is,
169
413000
3000
07:11
a crowd-sourced, six-word summary of a thousand TEDTalks
170
416000
4000
07:15
at the value of $99.50:
171
420000
3000
07:18
"Why the worry? I'd rather wonder."
172
423000
2000
07:20
Thank you very much.
173
425000
2000
07:22
(Applause)
174
427000
5000

▲Back to top

ABOUT THE SPEAKER
Sebastian Wernicke - Data scientist
After making a splash in the field of bioinformatics, Sebastian Wernicke moved on to the corporate sphere, where he motivates and manages multidimensional projects.

Why you should listen

Dr. Sebastian Wernicke is the Chief Data Scientist of ONE LOGIC, a data science boutique that supports organizations across industries to make sense of their vast data collections to improve operations and gain strategic advantages. Wernicke originally studied bioinformatics and previously led the strategy and growth of Seven Bridges Genomics, a Cambridge-based startup that builds platforms for genetic analysis.

Before his career in statistics began, Wernicke worked stints as both a paramedic and successful short animated filmmaker. He's also the author of the TEDPad app, an irreverent tool for creating an infinite number of "amazing and really bad" and mostly completely meaningless talks. He's the author of the statistically authoritative and yet completely ridiculous "How to Give the Perfect TEDTalk."

More profile about the speaker
Sebastian Wernicke | Speaker | TED.com

Data provided by TED.

This site was created in May 2015 and the last update was on January 12, 2020. It will no longer be updated.

We are currently creating a new site called "eng.lish.video" and would be grateful if you could access it.

If you have any questions or suggestions, please feel free to write comments in your language on the contact form.

Privacy Policy

Developer's Blog

Buy Me A Coffee