0:00:24.767,0:00:29.430
Each week I come across [br]an article or a report

0:00:29.430,0:00:33.476
that asserts that data is the new oil,

0:00:33.476,0:00:36.702
that the use of data will lead[br]to a new era of knowledge,

0:00:37.372,0:00:39.300
or even that it can predict the future.

0:00:39.700,0:00:43.504
This has been particularly true since[br]everyone started talking about big data.

0:00:43.504,0:00:46.804
You know, the use of large-scale data, [br]mega data.

0:00:47.723,0:00:50.673
For example, Sergei Brin, [br]the founder of Google,

0:00:50.673,0:00:53.490
who is focusing on the use of medical data

0:00:53.490,0:00:56.290
to cure Parkinson's disease, [br]for which he is at risk.

0:00:57.044,0:01:00.014
During the World Cup, many people said

0:01:00.014,0:01:03.673
that the German team was able to[br]beat the Brazilian team 7-1

0:01:03.673,0:01:06.033
thanks to the use of match data.

0:01:07.263,0:01:08.263
It's clear

0:01:08.263,0:01:11.525
that there is no field or [br]type of organization

0:01:11.525,0:01:13.082
for which Big Data

0:01:13.082,0:01:16.352
isn't supposed to be[br]a magic wand that will enable

0:01:16.352,0:01:18.642
the resolution of extremely [br]complex problems.

0:01:19.741,0:01:22.741
And I must admit that I feel uneasy

0:01:22.743,0:01:25.537
about these kinds[br]of simplistic statements,

0:01:25.537,0:01:29.187
which I see as overshadowing [br]a number of issues, including the economy,

0:01:29.195,0:01:30.505
the environment,

0:01:30.507,0:01:31.402
politics,

0:01:31.402,0:01:33.802
and the ethics [br]of the massive production of data.

0:01:35.233,0:01:37.638
Please don't think that I am skeptical

0:01:37.638,0:01:39.918
or doubtful about data,

0:01:39.925,0:01:42.505
or that I am opposed [br]to all forms of quantification.

0:01:42.803,0:01:43.803
On the contrary,

0:01:43.803,0:01:45.912
I live surrounded by data.

0:01:45.912,0:01:49.870
During the day, I'm working on a thesis[br]in sociology at Telecom ParisTech

0:01:49.870,0:01:51.102
where I study Open Data.

0:01:51.102,0:01:54.186
The important effort[br]to provide open access to public data.

0:01:54.186,0:01:56.496
And I study the consequences of Open Data

0:01:56.496,0:01:58.772
for the operation of government.

0:01:58.772,0:02:01.627
At night, I am the administrator[br]for an association,

0:02:01.627,0:02:03.653
Open Knowledge Foundation France,

0:02:03.653,0:02:08.310
which campaigns for open knowledge[br]and for data that benefits everyone.

0:02:08.310,0:02:10.227
Today, I would like to persuade you

0:02:10.227,0:02:13.114
that, at this time, when data[br]is becoming obtrusive,

0:02:13.114,0:02:15.583
we need to take a step back.

0:02:15.583,0:02:17.574
This coronation of data [br]that we are witnessing

0:02:17.574,0:02:19.548
during the era of Open Data and Big Data

0:02:19.548,0:02:22.364
demands a new culture of[br]critical thought about data.

0:02:22.364,0:02:26.784
We must be able to understand[br]how it is produced and used,

0:02:26.784,0:02:28.999
and how we can become[br]independent from it.

0:02:28.999,0:02:32.468
I also want to share [br]the results of an experiment

0:02:32.468,0:02:35.279
that we did at[br]Open Knowledge Foundation France

0:02:35.279,0:02:37.246
called "the School of Data."

0:02:37.246,0:02:40.468
I hope to show that, [br]through the use of data,

0:02:40.468,0:02:43.528
we can manage to develop[br]this culture of critical thought

0:02:43.533,0:02:46.513
and that we can develop [br]new checks and balances.

0:02:46.513,0:02:49.291
So, what are the problems with data?

0:02:50.659,0:02:53.711
The first problem is that[br]data is always right.

0:02:53.711,0:02:56.858
Now, don't believe that this is[br]anything new.

0:02:56.858,0:03:00.511
Historically, the word 'data'[br]comes from the Latin word "datum"

0:03:00.511,0:03:03.791
which, in mathematics and theology

0:03:03.791,0:03:05.534
in the 15th century, referred to

0:03:05.534,0:03:07.689
the facts taken as given in an argument[br]

0:03:07.689,0:03:09.919
and which were not to be [br]called into question.

0:03:09.919,0:03:12.311
Today, as you know, data refers

0:03:12.321,0:03:14.331
to everything that flows [br]in your computer.

0:03:14.331,0:03:17.830
That is to say, the 1's and 0's[br]that pass from USB stick to hard disk

0:03:17.830,0:03:18.830
are considered data.

0:03:19.774,0:03:20.802
On the other hand,

0:03:20.802,0:03:22.432
the sense that data is a given,

0:03:22.432,0:03:23.598
that it is factual,

0:03:23.598,0:03:25.307
that it is not to be questioned,

0:03:25.307,0:03:26.007
has remained.

0:03:27.825,0:03:29.865
The second problem with data

0:03:29.865,0:03:32.407
is that we don't really know[br]where it comes from.

0:03:32.574,0:03:34.885
In general, when someone uses data,

0:03:34.885,0:03:37.418
he or she has very little information[br]about the way

0:03:37.418,0:03:38.918
in which it was produced.

0:03:38.918,0:03:41.068
At best, you will have access to metadata,

0:03:41.068,0:03:43.094
that is to say, data about the data,

0:03:43.094,0:03:46.984
which will tell you the contents[br]of the file and, occasionally,

0:03:46.984,0:03:49.077
how the data was produced.

0:03:49.938,0:03:52.458
However, that data has a long history.

0:03:53.494,0:03:54.744
It was collected.

0:03:54.744,0:03:57.073
It was processed, formatted,

0:03:57.073,0:03:59.871
aggregated, processed by algorithms,

0:03:59.871,0:04:03.083
and visualized before reaching you.

0:04:03.083,0:04:05.383
This is why sociologist [br]Bruno Latour asserts

0:04:05.393,0:04:07.833
that we should say 'obtaineds' [br]instead of data

0:04:07.833,0:04:10.167
to accurately reflect this long history

0:04:10.167,0:04:12.557
which will constrain a number of uses.

0:04:13.268,0:04:15.939
Finally, the third problem with data

0:04:15.939,0:04:17.869
is that we can't really see it.

0:04:17.869,0:04:19.915
Have you ever seen a data center,

0:04:19.915,0:04:22.871
even if only from outside, [br]or from the road?

0:04:22.871,0:04:25.840
Do you have any idea [br]of where your data is stored?

0:04:25.840,0:04:28.521
I mean, physically, where it is stored?

0:04:28.521,0:04:31.512
Do you have any idea what will happen[br]to it in 10 years?

0:04:31.512,0:04:34.483
In any case, I have no answer [br]for these three questions.

0:04:34.483,0:04:38.173
However, even if we can't see our data,[br]we can measure its effects.

0:04:38.173,0:04:41.251
At the individual level,

0:04:41.251,0:04:43.771
when Facebook changes its terms of service

0:04:43.771,0:04:47.685
or modifies its algorithm, it has [br]consequences for your private life

0:04:47.685,0:04:50.720
and for the way in which you present [br]yourself as an individual.

0:04:50.720,0:04:53.620
And on the most macroscopic level,[br]the Snowden affair has shown

0:04:53.620,0:04:56.120
that the massive production of data

0:04:56.120,0:04:58.943
can have consequences [br]for the sovereignty of the State

0:04:58.943,0:05:00.433
or for diplomacy.

0:05:01.509,0:05:03.699
This is why we must develop a culture

0:05:03.699,0:05:05.465
of critical thought about data.

0:05:05.465,0:05:06.715
To encourage myself,

0:05:06.715,0:05:11.675
I was inspired by a book [br]called "Statactivism."

0:05:11.675,0:05:13.811
Statactivism is a neologism

0:05:13.811,0:05:16.811
proposed by researchers and artists

0:05:16.811,0:05:21.273
that refers to those experiences [br]that permit one to liberate oneself

0:05:21.273,0:05:22.989
from the power of data.

0:05:22.989,0:05:25.268
The fundamental basis of statactivism

0:05:25.268,0:05:27.456
is that data controls us,

0:05:27.456,0:05:30.379
and that it imposes on us[br]like an argument from authority.

0:05:30.379,0:05:33.787
The goal of statactivism [br]is almost revolutionary.

0:05:33.787,0:05:36.293
It asserts that other kinds of data[br]must be possible.

0:05:36.293,0:05:38.697
It is not necessary to be opposed[br]to all data.

0:05:38.697,0:05:41.197
Instead, we should use the power of data

0:05:41.197,0:05:42.728
to propose other realities

0:05:42.730,0:05:44.557
to critique data more effectively,

0:05:44.557,0:05:46.744
or to propose other measures.

0:05:46.744,0:05:49.054
In short, to propose other data.

0:05:49.056,0:05:52.516
There is a motif in the book[br]which I find particularly meaningful,

0:05:52.520,0:05:53.950
that of the judoka.

0:05:54.947,0:05:58.617
Judoka use the strength of their opponents[br]in order to turn it back on them.

0:05:59.490,0:06:02.273
That is what I want to invite you [br]to do today:

0:06:02.273,0:06:06.723
think about how to use data [br]to better analyze it.

0:06:08.372,0:06:11.529
I think, precisely at this moment[br]in the development of Open Data,

0:06:11.529,0:06:14.413
the need to develop a culture [br]of critical thought about data

0:06:14.413,0:06:16.633
is increasingly crucial.

0:06:17.462,0:06:20.982
Don't be misled: Open Data represents[br]an extraordinary opportunity.

0:06:20.987,0:06:23.027
The volume of data is exploding

0:06:23.027,0:06:25.906
and data is no longer [br]the privilege of the powerful.

0:06:25.906,0:06:28.141
Today, you can use data

0:06:28.141,0:06:30.119
without asking anyone's permission.

0:06:30.119,0:06:33.727
And this is a good idea, [br]because public data is available.

0:06:33.727,0:06:35.381
But I think that there is a risk

0:06:35.381,0:06:38.066
to thinking [br]that the simple diffusion of data

0:06:38.066,0:06:40.106
will be enough to emancipate society,

0:06:40.112,0:06:43.539
that individuals can emancipate [br]themselves from the power of data

0:06:43.539,0:06:45.510
just because they have access to data.

0:06:46.390,0:06:49.253
There is a Canadian sociologist[br]named Michael Gurstein

0:06:49.253,0:06:53.473
who has proposed an expression [br]that sums up a risk of Open data,

0:06:53.477,0:06:56.587
namely, "Empower the Empowered,"

0:06:56.587,0:06:59.475
meaning to give more power [br]to those who already have it.

0:06:59.475,0:07:02.938
That is why it's crucial[br]to develop a culture of critical thought

0:07:02.938,0:07:05.578
to be able to understand how data[br]is produced,

0:07:05.578,0:07:09.267
used, and how you can use it [br]to take a step back.

0:07:10.005,0:07:11.360
Well, that's the theory.

0:07:11.360,0:07:14.550
I would like to share with you[br]the first results from an experiment

0:07:14.550,0:07:18.254
that we did in my association: [br]Open Knowledge Foundation France.

0:07:18.254,0:07:22.028
We are part of a worldwide network [br]dedicated to open knowledge and open data.

0:07:22.028,0:07:24.005
We have groups in more than 50 countries.

0:07:24.005,0:07:26.926
And the idea of our association[br]and of this worldwide movement

0:07:26.926,0:07:29.483
is that each person can benefit,[br]can profit,

0:07:29.483,0:07:33.483
from works, scientific articles, [br]and content,

0:07:33.483,0:07:37.344
to create, play, educate, [br]or to start up a business.

0:07:38.225,0:07:40.913
Open Knowledge has [br]a large number of projects.

0:07:40.913,0:07:43.909
I'm going to talk about one project,[br]the "School of Data."

0:07:43.909,0:07:47.433
We participated together [br]in the translation of this project,

0:07:47.433,0:07:48.885
this "School of Data."

0:07:48.885,0:07:51.744
The School of Data consists [br]of online resources

0:07:51.744,0:07:53.834
that are free and accessible to all,

0:07:53.834,0:07:55.339
and also events.

0:07:56.306,0:07:58.097
We first proposed classes.

0:07:58.603,0:08:02.016
In these classes, you do not even have [br]to know what data is.

0:08:02.016,0:08:05.162
Or how to use a spreadsheet,[br]which is really the tool of choice.

0:08:05.162,0:08:07.444
You will be taught about that[br]in our class.

0:08:07.931,0:08:10.159
No expertise is required,

0:08:10.159,0:08:14.519
you are guided step by step[br]in the use of data.

0:08:14.519,0:08:17.739
We also use another format[br]which is particularly educational,

0:08:17.739,0:08:19.279
namely, the recipe.

0:08:19.287,0:08:22.290
Recipes are just like in cooking - [br]you have ingredients

0:08:22.290,0:08:23.293
and steps.

0:08:23.293,0:08:25.284
The ingredients will be data,

0:08:25.284,0:08:27.994
software - free if possible,

0:08:27.994,0:08:31.329
so that you can use data.

0:08:31.329,0:08:34.440
The idea is that making a map[br]of electoral results,

0:08:34.440,0:08:36.879
or a graph of results [br]of the French soccer team

0:08:36.879,0:08:40.142
should be as easy to do as making[br]a tarte Tatin or Bechamel sauce.

0:08:40.142,0:08:41.895
You find the resources online

0:08:41.895,0:08:44.595
and we walk you through the project[br]step by step.

0:08:44.595,0:08:48.040
We also have tried to develop[br]another format for in-person sessions,

0:08:48.040,0:08:49.741
which we call expeditions.

0:08:50.117,0:08:52.644
For expeditions, it's like[br]mountain climbing:

0:08:52.644,0:08:55.103
you have a guide, a "data sherpa,"

0:08:55.103,0:08:56.993
who will accompany you,

0:08:56.993,0:08:58.233
attached by a rope.

0:08:58.233,0:09:01.432
There will be 10 or 20 participants

0:09:01.432,0:09:05.267
who work together during a weekend[br]or sometimes for a few hours.

0:09:05.769,0:09:07.631
Our first data expedition

0:09:07.631,0:09:10.991
focused on the question of air pollution[br]in Île-de-France.

0:09:10.991,0:09:12.675
I don't know if you have seen

0:09:12.675,0:09:15.325
these images of Paris [br]with black clouds of pollution.

0:09:15.325,0:09:18.060
They left their mark on us, [br]and we said to ourselves:

0:09:18.060,0:09:21.729
"Well, let's dig into this set of data."

0:09:22.458,0:09:25.469
The first step, when we undertook[br]this data expedition,

0:09:25.469,0:09:27.741
was to identify the available data.

0:09:27.749,0:09:31.041
We realized that there is[br]no available data

0:09:31.041,0:09:34.801
that is freely reusable, that is to say,[br]that you have the right to reuse

0:09:34.801,0:09:37.466
without asking for permission,[br]on this crucial question.

0:09:37.466,0:09:40.679
Therefore, we had to extract data[br]from websites,

0:09:40.679,0:09:43.299
reports, or even from graphics.

0:09:43.299,0:09:47.127
Imagine what a mess it is to expose[br]data that is in a graphic.

0:09:47.811,0:09:50.402
We also realized that Airparif,

0:09:50.402,0:09:53.161
the organization responsible [br]for the production of data

0:09:53.161,0:09:56.411
relevant to the question [br]of air pollution in Île-de-France

0:09:56.411,0:09:59.581
does not allow you to use[br]its data freely.

0:09:59.581,0:10:02.127
One must ask permission, or pay.

0:10:03.130,0:10:05.164
We were able to overcome[br]these constraints

0:10:05.164,0:10:07.701
and to conduct this expedition

0:10:07.701,0:10:10.571
guided by our sherpa, Pierre.

0:10:10.571,0:10:13.987
During this data expedition

0:10:13.987,0:10:18.227
we broke into small groups,[br]and each group was assigned an angle.

0:10:18.227,0:10:22.257
One of the principles of the expeditions:[br]you have an angle, like in journalism,

0:10:22.257,0:10:25.316
we ask ourselves questions that could be[br]the title of an article.

0:10:25.316,0:10:26.996
The first group asked itself

0:10:26.999,0:10:31.484
if bicycle riding had led to a decrease[br]in air pollution in Paris.[br]

0:10:31.484,0:10:32.524
The second group,

0:10:32.524,0:10:34.254
since it was during a strike,

0:10:34.254,0:10:37.464
asked itself if public transport strikes

0:10:37.464,0:10:40.616
cause air pollution in Île-de-France [br]to increase.

0:10:40.984,0:10:44.384
And the third group asked if[br]all regions are equal

0:10:44.384,0:10:48.424
with regard to air pollution,[br]or if geography and environment

0:10:48.424,0:10:51.438
could have an effect, and if so, [br]could be seen in the data.

0:10:53.617,0:10:55.173
The results of this expedition,

0:10:55.173,0:10:57.373
I am sorry to say, will be a bit [br]disappointing.

0:10:57.373,0:11:01.440
We did not find any correlation[br]or causal connection

0:11:01.440,0:11:04.320
with nice data points, [br]a fitting curve, or a straight line,

0:11:04.320,0:11:06.546
that proves that our hypotheses[br]are correct.

0:11:06.546,0:11:10.186
We did not succeed at that,[br]but we worked for four hours.

0:11:10.186,0:11:12.196
What we did manage to show,[br]on the other hand,

0:11:12.196,0:11:14.007
is that it is extremely difficult

0:11:14.007,0:11:17.838
to use data concerning a question[br]as crucial as air pollution,

0:11:17.838,0:11:20.438
to understand how it is produced,

0:11:20.438,0:11:23.228
extremely difficult to use it,

0:11:23.228,0:11:26.084
that the most simple measurements[br]are not accessible,

0:11:26.084,0:11:29.336
and that you do not necessarily have[br]the right to reuse them.

0:11:29.336,0:11:32.346
That is just what we tried [br]to do at this event:

0:11:32.346,0:11:35.426
to develop a culture of critical thought[br]on the way in which data

0:11:35.426,0:11:38.491
is used concerning the question[br]of air pollution.

0:11:38.491,0:11:43.359
We also tried to develop this format [br]of expeditions and training events

0:11:43.359,0:11:45.149
with another group

0:11:45.149,0:11:46.239
that is less expected,

0:11:46.239,0:11:48.131
that of children.

0:11:48.131,0:11:53.061
We asked ourselves the question[br]during an event that we did with Etalab,

0:11:53.061,0:11:56.661
the government institution[br]in charge of data.gouv.fr,

0:11:56.661,0:11:59.500
the open data portal[br]of the French government.

0:11:59.500,0:12:03.737
We suggested the idea [br]of radically different open data portals.

0:12:03.737,0:12:07.767
They were fictional projects,[br]just prototypes.

0:12:07.767,0:12:12.951
There is a group that has come out[br]with a prototype called Tada.gouv.fr.

0:12:12.951,0:12:17.132
Tada.gouv.fr is a fictional portal,[br]a bit idealistic, destined for children.

0:12:17.775,0:12:21.898
The data is presented[br]not by government department or minister,

0:12:21.898,0:12:24.798
but by discipline, that is to say[br]that you have data

0:12:24.798,0:12:27.517
about history and geography,[br]physics and chemistry,

0:12:27.517,0:12:29.477
or life and Earth sciences.

0:12:29.477,0:12:32.473
On this occasion, we realized[br]that open data

0:12:32.473,0:12:34.995
can be a fantastic resource[br]for school

0:12:34.995,0:12:38.295
because it allows the development[br]of inter-disciplinary work,

0:12:38.305,0:12:41.370
and this culture of critical thought[br]about data I have mentioned.

0:12:41.916,0:12:43.877
We did not leave things at observation.

0:12:43.877,0:12:46.217
We tried to do a first experiment

0:12:46.217,0:12:48.751
and I would like to tell you [br]about the first results.

0:12:49.745,0:12:51.668
We joined with Silicon Banlieue,

0:12:51.668,0:12:54.058
which is a site dedicated[br]to data in Argenteuil,

0:12:54.058,0:12:56.089
and we proposed to do an event

0:12:56.089,0:12:58.479
with children between 8 and 14 years old

0:12:58.479,0:13:00.258
who came to the Open World Forum,

0:13:00.258,0:13:02.546
an event dedicated [br]to open computing in Paris.

0:13:02.546,0:13:05.388
There, you can see me from the back.

0:13:05.388,0:13:08.828
With the 8 to 14 year old children,[br]we worked on the question of cinema,

0:13:08.828,0:13:12.152
because that interested them,[br]and it is a simple enough subject.

0:13:12.152,0:13:13.779
First we collected data,

0:13:13.779,0:13:16.619
nothing very complicated,[br]it was just a paper form.

0:13:16.619,0:13:20.026
We asked them how many times a month[br]they go to the cinema,

0:13:20.026,0:13:23.808
which movies they saw from a list;[br]then we compared that with data

0:13:23.808,0:13:27.368
that is available from the survey[br]of French cultural practices,

0:13:27.368,0:13:30.243
on which you have [br]exactly the same type of data.

0:13:30.243,0:13:33.873
With the children, we produced[br]an infographic at this time.

0:13:33.873,0:13:37.510
Now, I am really bad at math,[br]I got a 7,5 on the Bac,

0:13:37.510,0:13:40.590
I found myself explaining[br]the concept and calculation

0:13:40.590,0:13:43.512
of averages using a spreadsheet,[br]which was rather surprising.

0:13:43.512,0:13:45.362
I explained how it works.

0:13:45.362,0:13:49.505
We emerged with an infographic[br]and we were able on this occasion,

0:13:49.505,0:13:53.135
I think that this is the important point,[br]to develop a culture of critical thought.

0:13:53.135,0:13:56.521
I explained to them about data,[br]how it is used,

0:13:56.521,0:13:58.304
how they can use it,

0:13:58.304,0:14:00.914
how it controls us in a certain way,

0:14:00.914,0:14:03.726
but that we can also take back[br]the power over data.

0:14:03.726,0:14:06.731
I assure you that with a topic[br]as attractive as cinema

0:14:06.731,0:14:08.695
we can deliver this kind of message

0:14:08.695,0:14:10.798
and have a discussion on these questions.

0:14:12.058,0:14:15.198
I hope that I have convinced you[br]that it is necessary today

0:14:15.198,0:14:17.217
to take a step back with regard to data,

0:14:17.217,0:14:20.523
to develop a culture of critical thought,[br]to understand

0:14:20.523,0:14:23.574
how it is produced [br]and how you can use it,

0:14:23.574,0:14:26.806
to prevent data from being forced on you.

0:14:26.806,0:14:29.554
So from today, [br]get your hands dirty,

0:14:29.554,0:14:32.294
find a sherpa,[br]all of the resources are online,

0:14:32.294,0:14:34.123
and go on a data expedition.

0:14:34.123,0:14:35.313
Thank you.

0:14:35.313,0:14:36.765
(Applause)