Transcript (view)

0:07
my name is
0:08
here are and here your moderator
0:11
for this internist my panelists
0:14
up in in a minute and come from a world which is
0:19
a little bit removed from from this call world for a more of its been really
0:24
exciting for me for the last
0:25
meet so many people so many interesting things that I’m really excited about
0:29
bringing
0:29
other conversation that’s really interested interesting to me
0:33
and data into this at
0:36
into this audience have and the cofounder a company called the office
0:41
for creative research
0:42
we’re halfway in between an artist collective in R&D group
0:45
and we focus on on I’m data base in your city
0:49
and previous to that for two and a half years of the day are
0:53
residents at the New York Times so for the last I would say five years or so
0:58
posted by thinking has been sent their outdated and very specifically the
1:02
intersection between dated him and switch is something that I really
1:05
think it’s valuable for for our discussion today
1:08
have because I’m an artist that I and II wanna stay with the metaphor
1:12
outlets thinking about a about data
1:16
most it’s probably full here in
1:19
into an airplane and and when you re
1:22
when you’re in an airplane and when you’re in that air traffic system I
1:25
think it’s what did they say
1:26
very unusual times in our human lives where we relinquish control
1:29
we’re sorta go into these lineups that her and we see these dense systems
1:34
around us that we only get lessons that will see that
1:37
with that axe disappear entities a
1:40
weird flapping wherever outdoors and then they go out
1:43
who knows where there any systems around the airplane when they’re being field
1:46
and so on and so on
1:47
and we’re we’re partly overfly the vast system and its
1:52
really I don’t think I ever had simply large there are
1:55
a at any given time there are more than a million people in there
1:59
and graphing you’re seeing on the screen is this for aspirants system love
2:03
up airplanes landing and take lock
2:06
fifteen-minute of and and
2:09
so for me this in dade a world
2:13
is is actually really about they systems and
2:16
star to to be able to communicate with the state systems and me
2:21
let I know that a lot of view to work in the type of work
2:25
work that you do are involved with very big and very often
2:28
human systems and so my hope is that well-stocked understand how the state
2:32
data wrote
2:33
translate to a better understanding better engagement with these types of
2:36
systems
2:37
I also want to talk about the experience a bit data is like
2:40
at and David Foster Wallace is one of my favorite writers wrote
2:44
a in this memo to his ETA and if anybody’s ever read the book Infinite
2:49
Jest but the
2:50
Infinite Jest has footnotes and sometimes those footnotes have footnotes
2:53
and occasionally the footnote to the footnotes happened at
2:56
when he was asked about why it if after the vote that way he answered
3:00
partly that he said he wanted to mimic the information flight and data triage
3:04
that he expected to be a
3:05
even bigger part US like 15 years yes it just has written sixteen years ago
3:10
and I know that alive you are probably
3:14
feeling every day with this idea data triage how how to me
3:18
how can we take this baskin outdated you’re being faced with your collecting
3:22
you’re able to use have been you make some type that upset with them
3:26
at the word the word aight as everywhere
3:30
in for me to change for the last five years has been really impressive one
3:34
I did attack I this that has in this workshop
3:38
out and and a I i found out that the new york times has published
3:42
just under a thousand article so far this year that it
3:45
hit the word data and if we look at those articles
3:48
there there really an assessment the types of things that people are talking
3:52
about here
3:53
at exits have a diverse set of hands like this it appears
3:56
aronie have made by all if these data stories here we have stories about
4:01
about health and stories about the environment and stories about education
4:04
where data is really starting to
4:06
up inform all of those things we have a huge a wide swath of territory to deal
4:11
with when where
4:12
what we’re talking that data and I i wud
4:15
with frame the conversation around three at
4:18
pillars if you well and and at
4:21
and note before we get into that I think it’s really important for us to remember
4:26
at all the stories about data that in the New York Times
4:29
not one of them contains a definition I think that it is one of those things
4:33
that everyone
4:34
assumes that we know it is but really when I ask people
4:37
a very very rarely left there in the business that they have an answer that
4:40
make sense
4:41
so when I ask people to close their eyes and think about it there’s two things
4:45
that come out
4:46
every at the first as an Excel spreadsheet and
4:49
the second is that trickle up numbers that was coming down the screen on the
4:52
matrix
4:53
how positive things that people pocket and
4:56
but let’s that’s sharon is the definition then move into the three
4:59
things and that definition it’s a very simple they are measurements at
5:03
I think athlete in that definition are two very important day
5:06
there’s an active measurement there’s something that happens somebody’s
5:09
measure accept it
5:10
and then there’s this thing which in in an hour case in many of the cases here
5:14
those things are humans and so to understand that binding between the data
5:19
and and assistance that that it came from I think are
5:22
very very very important art understanding I think
5:25
so it’s let’s talk first about this term big data because I i allergic
5:30
interdicted and maybe it’s because I’ve heard it so much over the last few years
5:34
and and I want less as we continue this
5:38
it wasn’t going to this conversation to talk a little bit that
5:42
white why are we so sad yesterday data
5:45
so if you may have come into this room with very little engagement with data in
5:49
your work but somebody I know we’re coming into this room with that pete has
5:52
to rethink a job with data
5:53
see you might have thought may be thinking what what is the big deal about
5:56
all this I’ve been using dated my business for years and years and years
5:59
so
6:00
what is really happening to change the way that we engage with it and that’s
6:03
what wanted their
6:04
our focus is and that I also want to focus on small they have little bit
6:08
a let me give you an example and when I worked at the the times we develop its
6:12
to allow
6:12
call cascading cascade is a tool that allows us to
6:16
up to to look at their conversations that are happening around New York Times
6:20
content
6:21
so this is a single conversation around a story called the island where people
6:24
forgot to die
6:25
which is about I’m this island in Greece where people live very long time
6:29
even compared to the neighboring neighborhoods and so we have this system
6:32
where we can take any story and we can model it’s conversation and we can see
6:36
it
6:36
where the conversation started where you the time says the light updated their
6:40
6500 passive
6:42
have content being created every I every month
6:46
and each one of those fastest times have conversation so this is
6:49
help maybe in the day to run this medium size data but it’s fairly big data we
6:53
have millions of data points
6:55
hundreds millions of data points and we can keep track of all those things
6:58
historically
6:59
but but actually what this tool that more than allowing us to understand that
7:03
a message via it was the loudest understand the smallest at the data
7:07
and my favorite cascade that we identified in this and and this whole
7:10
thing with this cascade which I call the rabbi cascade
7:13
and I caught the rabbi casket because everybody talking about this story
7:17
where they were rabbis and the story was about religious workers and why they
7:20
don’t take time off
7:22
because at saturdays and sundays are not good days for religious works the
7:25
tapping time of
7:27
and then they tend to work for the week as well and so what we’re able to find
7:30
with the store was this very small conversation about this story about
7:33
a religious workers not taking time out that with amongst these rabbis
7:37
and one of them has the best Twitter name that I hurts the velvety rabbi
7:42
at could so as we as we’ve talked about this ’cause allowing us to get
7:46
big scale at it let’s also talk about what
7:49
the at the these tools and approaches can help us with the small grain
7:53
higher out %uh everybody’s a
7:57
business and everybody’s college services again or units
8:00
anything in it in a few little bit for me feat in our conversations
8:03
and then the second thing and who is this date in value have happening really
8:07
a lot that
8:08
this for it with a better understanding of how you can apply these big data
8:12
concepts and actually get some value
8:14
and I’m gonna break that into two into two parts the first is value
8:17
health and and and there’s a great deal of expertise
8:21
in have that value do it for analysis
8:24
I’m and the data visualizer mostly in so much to my value when I
8:27
when I think about data and data protest comes the visualization
8:31
and I wanna give you just a very small sample about that as well as
8:34
as we move forward so up in nineteen in in that in 2009 I was asked it
8:40
develop a an algorithm to help place that nearly
8:43
four thousand names the 9/11 memorial in that hat
8:46
this is a very difficult job very interstate him and shot them
8:50
and the very first thing I did help understand the system was to build a
8:54
visualization and its
8:55
not that the most beautiful visualization ever
8:59
specialization did made away with go help me understand the problem
9:03
and so we think about now sits as well I i life but on the big screen also on a
9:07
small scale and how
9:08
data visualization these takes a process that we’re gonna think about today
9:11
can’t help myself really difficult problems and
9:15
and and and then the next thing is is value of their story
9:19
this is how how can Houston a data
9:22
help you helped us tell story it’s better
9:25
and this is something that that that I’m personally really excited about being a
9:29
visualization person
9:31
and I think what this process allows us to do with that allows us to take things
9:34
which sometimes can pay
9:36
I try okay britt in really bring them to life
9:39
and and it was after this from from my own work again
9:43
at this is a paper that that nasa published in
9:46
it two years ago when they announce the initial findings at the coupler project
9:50
which is the telescope that’s trying to find
9:52
distant planets that are orbiting a around
9:55
around these far-away stars and and this is a text and graphics that nasa
10:00
publishers
10:00
with with these findings and to astronomers think they’re really excited
10:05
ass maybe last six a day and I think that a lot I i’ve
10:09
you probably have dated information that your in some cases forced to present
10:13
place where it’s going to be report to that for your funders and so on and so
10:17
on and so on
10:18
what I like best thing about this have to me helping me engage with
10:22
with individualization it and and as and something to let
10:26
analytic techniques to help tell their stories better in this case with the
10:29
Kepler project
10:30
I produced this visualization which has the exact same data in a slightly
10:34
different way
10:35
have at and and through a really narrative way we
10:38
we get understand what that system is and what that system looks like
10:43
up hour to play at for for
10:46
for just a moment and and the have a fair
10:50
saying and it’s her thing was added because robert and I took the train
10:53
in from the airport %uh underway in here we had a really interesting conversation
10:58
that and past how how can we use our
11:02
are our data that’s maybe in your organization and combine that data with
11:06
with data from other organizations and other sources
11:10
happen this this date if they actually start to really
11:13
fact change change the world and
11:16
and so the third part of our conversation today is going to be around
11:20
this idea of social dated
11:21
census and rather than project example a minute have any use a quote for this is
11:26
a quote from
11:27
a is actually a network scientist a life a very bassy in
11:31
and in hit this very interesting quote which I’ve been a little bit obsessed
11:35
with
11:36
0 over the last few months workers and do you wanna stop different transmitted
11:40
diseases
11:41
do you want to design better cities do you wanna stop traffic jams the data to
11:44
do so is there in private hands
11:46
and we need to identify some social consensus by which the data
11:49
can be shared with different stakeholders said
11:52
I think there’s a really really exciting opportunity for the people in the room
11:55
to be driver service to understand how we can
11:58
move data a out at a private hands and and into a place where they can help it
12:03
exit efforts
12:04
and the things that we’re working and
12:08
and really that name partners I want people to live with
12:11
and only some inspiration about the type of work the people on the panel but some
12:15
understanding of
12:15
I read there how are we gonna do this so how how is this actually going to happen
12:20
so I’m hoping that when we break into questions will be able to
12:24
to to to really really address that
12:28
and one of the great things about working
12:31
at with it but that with that panel at this is that I for that
12:35
have restrictions with these amazing people and finding out a little bit more
12:38
about that works really well we’re gonna start up its
12:40
is for you to find out a little bit more about this amazing people in their work
12:44
so
12:44
let me is some introductions for civil right next to me is Robert Packer
12:48
and who is the director have a a
12:53
greatest he went well said I i
12:57
at and
13:00
wherever I’ll and and the the UN Global pulse works works very closely with the
13:05
secretary general’s office to
13:06
to at do what I think when actual Roberts work by students as really the
13:10
clear example up
13:12
the rather hitting the road data is there a really really really
13:16
the the best examples that people you say
13:19
capital be dictated to help solve capital be
13:23
problems and and I’m Roberts gonna show us
13:26
a little bit about about its work but first let me introduce the other
13:29
speakers so doctor take
13:30
our way in the middle here yeah is the founder and executive director at the
13:34
group called data kind and I’ll let
13:36
Jake a at explained it a kind in a little bit more
13:39
more detail yes i know i won I won’t do it justice and Inter Milan
13:43
again is the founder and CEO have clear story data
13:47
have Robert
13:50
Robert retro there and and
13:53
and and had his put together a smile
13:57
a small set of 52 talk about some of the issues that he wants to talk about so
14:02
I wanna let rather pick up that that PC controller and will switch over to
14:07
to terror here
14:11
well thank you up so I
14:14
as your noted I am I work and a small team
14:18
inside the secretary general’s office and New York
14:22
called local pols we are essentially allowed
14:25
for the UN system to learn how to figure out what’s happening while it’s still
14:29
happening
14:31
on how to take advantage of this real time information that’s out there
14:34
up the project came out of the global financial crisis when everyone realize
14:38
that they were trying to stand how the crisis was affecting viral populations
14:41
around the world
14:42
and all the data it was available preset you know predated the onset of the
14:46
crisis
14:47
I’ll they are the project itself does a lot of work
14:50
looking at archives a big date of arm
14:53
and seeing what changed in real time a year ago and something happened to a
14:57
population
14:58
and we compare that to the official statistics so
15:02
so you may remember this at this is an IBM at that came out a couple of years
15:05
ago
15:06
with %uh chief data scientist I’ll Jeff Jonas hear from IBM
15:10
and it was i asked a question targeted at the private sector
15:13
would you want to cross the street based on data that was five minutes old
15:18
are you know now if you’re making stock trades your tracking what’s happening on
15:22
the manufacturing floor
15:24
I you’re looking to understand instantly what happened
15:27
but say I’m I in your operations
15:31
you know private sector gets this but and development
15:34
we’re still using three-year-old data to make five your plant
15:38
I’d up we’re living in a hyper-connected world
15:41
where things are changing faster and faster and yet we’re not able to
15:44
understand what’s happening is it happened
15:46
so this is a video out there is essentially a visualization of $300
15:50
million tweets over one day in the US
15:53
from which sentiment has been extracted and you can see overtime
15:57
people get happier and satyr from one coast to another
16:00
I’m people really happy in Florida ensalada sunshine their
16:03
new yorkers not so happy that’s for other than our some colleagues her on
16:07
stage
16:08
are this is a a visualization from the city of Geneva
16:12
with up millions and millions anonymous call records
16:16
you can see every time people make calls or send messages their location distract
16:20
and what this shows is how people commute to work we are a swarming
16:24
species
16:25
those who aren’t aware that we show like fishes
16:28
this is a visualization of financial transactions from hi
16:31
Easter weekend 2011 and Spain
16:35
the data was provided aggregated anonymous for my Spanish bank
16:39
BBVA this information is out there it’s being generated in real time
16:43
we’re not using it to understand what’s happening to beneficiaries
16:48
our founding hypothesis is that all these digital
16:52
services that are out there the people are accessing for twenty dollar smart
16:55
you know twenty dollar
16:56
feature phones all over the world now are sensors for human well-being
16:59
where people’s household needs change they change how to use the services and
17:03
those changes leaf patterns
17:05
we need to understand what the patterns are so that we can actually track events
17:08
as they occur
17:10
up this is really what local polls is focused on I wanna show you some
17:13
examples
17:14
of the kind of information that’s out there up but this is basically what
17:18
we’re asking
17:19
if private sector can understand its customers and its markets in real time
17:23
how can we began to look at the different sectors that are relevant to
17:26
our work
17:28
so after one example love this this is from john brown stains world
17:32
are and it’s about using this kind of public big data to track disease
17:35
outbreaks
17:36
I’m this is from New York and this is simply people reporting that they are a
17:40
family member have the flu compared to the official US CDC
17:43
um influenza statistics some of you may know about Google Flu Trends and search
17:47
is predicting
17:48
a disease outbreaks Twitter’s even better I’m this case
17:52
people say how can you trust information from people you don’t know well
17:55
there are a lot of people doing this and they have no reason to lie about having
17:59
the flu
18:00
turns out to be very powerful up
18:03
really interesting data here mobile phone carriers
18:06
now they’re not tracking your text message or what you been
18:09
your they’re not recording your conversation but what I do know is who
18:12
called whom and when
18:14
from one location to what location this allows them to track how you spend money
18:18
on your phone
18:19
are they can see the patterns of communication to call only within your
18:23
village a recall and other areas
18:25
and of course they can see as we saw from the video people moving around on
18:28
the map
18:29
all over so using maps we use maps and I worked all across the UN system we have
18:33
thousands of maps what’s missing from the maps
18:35
the people the mobile carriers can see people moving around on a map
18:40
they can tell you sleep here and you go here all day and you come back they can
18:43
see your commute how far you commuting
18:45
are you employed they can make inferences about this
18:48
we’re doing lively arts programs that we have no clue
18:52
turns out artwork in East Africa off by my doctor Nathan Eagle
18:56
Josh Bowman stock a sound that in fact you can protect people’s household
18:59
income simply got the size and frequency their time purchases
19:02
right so here we have somebody who spends ten dollars a month all at once
19:06
on their phone versus somebody who
19:07
buys little increments little scratch cards every couple of days
19:11
turns out that predicts income very very well
19:14
up sometimes the stereotypes are true hear you say differences between men and
19:18
women
19:19
and how they use their phone statistically I love the algorithm
19:22
skinhead eighty percent accuracy on gender prediction
19:25
they can also predict age to plus or minus three years
19:28
oMG very very interesting stuff because now we can start to make inferences
19:32
about where
19:33
a particular population is under straps
19:37
this is an example of how mobility can actually be used potentially this is our
19:40
work done by Telefonica largest carrier in Latin America they look to the h1n1
19:44
outbreak in Mexico City and a look back at the call records and were able to see
19:48
are through a mobility and acts how much people were moving around but their
19:51
farms that after the government said
19:53
stop moving around don’t get on the buses don’t spread this
19:57
are people start moving around you can say I 429
20:00
that was the day of the announcement people quiet the dow the ability to
20:03
evaluate a policy response in real time
20:05
is out there so we see three types of opportunities in the state of
20:09
one better early-warning to
20:12
situational awareness what’s happening right now as I’m my program
20:16
and three perhaps most significantly real time impact evaluation
20:21
the ability to see collective changes in human behavior in the population where
20:24
we’re implementing
20:25
can create a new real-time evidence base that can tell us where what word
20:29
doing is working aware it’s not
20:32
we see a future in which development is actually
20:36
more agile and adaptive if we’re driven by real time feedback we can take course
20:40
corrections
20:41
and learn as we go scale as we go and get away from pilots that last two years
20:46
before we find out it wasn’t working the way we fought
20:50
so our challenge is here how to get access to the data
20:53
how do we combine it and how we anonymizer to protect privacy
20:57
I what are the patterns we should be looking for in the data how can we
21:00
actually turn that into
21:02
action and institutions that are not familiar with risk
21:05
you know faster cheaper but sometimes less accurate information
21:09
but talk quickly about somebody’s opportunities and challenges
21:12
this is how we’re going to gauge with private sector past two years a concept
21:15
called data philanthropy
21:17
up essentially working with these companies to say look you’re sitting on
21:20
a gold mine you’re sitting on information that can be used to help
21:23
people right now
21:25
but we’ve got to find a way for you to share it safely in a way that say for
21:28
the people
21:29
in terms of privacy and safe for your own institution your own organization
21:33
from a competitive standpoint
21:35
the response has been nothing short of overwhelming companies today
21:38
get the power data they say this is an all former corporate social
21:42
responsibility
21:43
and they also understand that if the information they have a how people use
21:46
their
21:47
their services could actually help protect their customers from economic
21:50
harm
21:51
that’s good for business to but it’s not really about philanthropy and
21:56
it’s what that’s what year was touching on here we need to move toward a world
22:01
in which public and private sector share real time information
22:04
in a commons because everybody benefits from
22:10
big data we hear a lot of big data and privacy it’s not just a privacy issue
22:14
it’s a human rights issue
22:16
the massive potential of this information to be you know exploited for
22:20
harm
22:20
is out there part of what we’re doing and the UN is really looking at this
22:25
pace
22:25
increasingly you know how do we do this safely and how do we mitigate the risk
22:29
misuse of this information we’ve been working all night data privacy
22:33
protection framework for the last eight months
22:35
to inform our work through our labs up
22:38
this is the debate today around the stuff right on the one hand you have the
22:42
I you have kind of the the tech industry saying privacy is dead
22:46
on the other side you have the regulators saying any
22:49
any reuse my data as potential misuse and its har
22:53
we want every frame this debate and insert a third pole into it
22:56
namely big data is a raw public good
22:59
but in order to convert it safely and responsibly into a genuine public good
23:04
we have created a sample battery can experiment to learn how to do this
23:08
so we’re setting up labs global possible out over establishing louts up the
23:12
country level
23:13
we partner with organizations out there I love different types organizations
23:17
with technology
23:18
with challenges with data with ideas
23:21
and we do research project to learn where the signals are and we build
23:24
technology foresight types open source prototypes
23:27
and testes in the field to see if we can actually move the needle and where we
23:31
find something that works
23:32
back in scale on we have lost so far too loud
23:36
%uh the first launch last year and Indonesia up indonesia is an amazing
23:40
place to study shows social media
23:42
there are the largest users social media in the world post lab Compal open later
23:46
this year
23:47
I will be opening up the labs future up
23:50
a few examples of what we’ve done just a to wrap up up
23:53
this is a project we worked on I’m and the end of 2011
23:57
actually where we had a hypothesis that before people lose their jobs they know
24:01
something’s wrong at the office
24:03
and that will change how I talk a lot about work and sure enough we found in
24:06
these two countries at least
24:08
that between four and five months before you lose your job we can actually
24:10
predict
24:12
the increases in unemployment just from looking at blogs and forums
24:15
us
24:18
this is a map of tweets and a car I can tell people spend a lot of time tweeting
24:22
stuck in traffic to cart it today produces more tweaks that any city on
24:25
earth
24:27
I’m really interesting place to study this content now people are tweeting a
24:30
lot about celebrities and sports scores but when you filter that out and that’s
24:33
what machines are good for
24:34
you get a lot of people talking about food prices and
24:38
not being able to afford basic necessities and you know losing their
24:41
jobs
24:42
so interesting signal up HIV
24:45
we spent a lot of time funding programs you’re trying to change perception and
24:48
awareness of the disease
24:50
turns out these technologies for mining big data are really good measuring back
24:53
because they were designed to measure the effectiveness
24:55
%uh at and here we saw when we look at conversation and World AIDS Day
24:59
in two consecutive years a tripling of conversation about HIV
25:03
and a near doubling of conversation about testing and counseling so
25:06
something’s working
25:09
we look at food indonesians talk a lot about who turns out the nation’s tweet a
25:13
lot about
25:13
we found 14 million tweets about food the question was debate went differently
25:17
about
25:17
open they’re not getting food so that’s Ramadan
25:24
up when you compare simply the number of tweets mentioning the word price and the
25:28
word rice
25:29
you start to get correlations with Fiat official consumer price index for food
25:33
and in Asia
25:34
up so you start to see some potential protection here are
25:38
we’ve been using the platform crimson hexagon came out of Gary King’s work at
25:41
Harvard
25:42
are and tracking the online conversation through our
25:45
lab in Jakarta over on the right that’s that little blip
25:49
around Ramadan you see that every year baseline on the left
25:52
that’s what happens when the soybean supply globally crash because the
25:56
drought to the US
25:58
that’s up that’s a real a real for crisis other purple subsequent about
26:02
conversations
26:03
about people rioting are we’ve done some comparisons of
26:07
I around certain topics and moves and a founder we can protect the consumer
26:11
price index around
26:12
of food in Indonesia with eighty-nine percent accuracy
26:15
something I might miss out media up
26:19
this is very recent I’ll back at the end of January
26:23
the team in Jakarta call emerging topic people saying
26:26
vaccines serums made in the US it contains pork products
26:30
it is not whole all good muslims are not to vaccinate their kids
26:34
now vaccines are all but what people do or do not know in the village
26:38
has a huge impact and how how many conversations at a cafe do you think
26:42
there are
26:43
for every tweet you say so we’ve been working closely with UNICEF abetting
26:47
conversation with Islamic authorities and Jakarta
26:50
about getting the word out and particularly looking at those provinces
26:53
where you’re saying this
26:54
this pattern emerge so
26:58
lots of opportunities to get involved in this kind of work others in the spacer
27:01
are beginning to get engaged as well I’m you hear about what
27:05
but Jacob Sean Miller doing I’ll but the lot of opportunities if your data you
27:09
have tools you have ideas
27:11
you have knowledge arm or you just like to get involved in a project
27:16
thanks %ah
27:23
a I have the same question for for taken terminal I because I think it’ll help us
27:27
understand a little bit about
27:29
let here you and you’re doing II less thinking about my panelists a big river
27:33
a data superheroes everything exciting
27:36
most exciting thing about superheroes is there origin story
27:40
and so and any take you want to start with that with that
27:44
but the data Concord story because presley I was lucky enough to be
27:48
be nearby when this thing started and I think there was a
27:51
there’s a really interesting story there so I’ll give it could be a little bit
27:55
I’m
27:56
to episode why yeah
27:59
film rights are gonna be option set actually I think you with my whole name
28:02
to the original project data without Borders
28:05
I describe the subject when we’re working at the New York Times I said
28:08
I’ll
28:08
colic a data without Borders at is ceci
28:12
that’s really get I’m so Alex when I that’s a great guy gets out ago the
28:16
origin story
28:17
I just say it’s it’s so inspiring being on panels like this because it gets
28:21
really bringing home to me
28:23
a special people like this how the world list is changing
28:27
and how all this data that we have found us allows us to see these new things and
28:31
that ties into my origin story because I came from a a data science background
28:36
and I was a statistics kak
28:39
your science guy but long before that I don’t want to get into engineering cuz I
28:42
wanted to build solutions
28:44
to make the world a better place and I’d always notebooks full
28:47
love trying to as a kid ive me building robots to go to Spain ASA and making
28:53
cities in the sky and
28:55
at 8 I didn’t quite get there but you know that was
28:58
there was always this idea that technology could be used to sell this
29:01
really big problems
29:02
and I was super excited to follow a path for I like computers and data
29:07
right as I was coming out with around 2006 this exciting data moment
29:12
where it seems that hone the skills that I did picking up to make sense if data
29:16
looking at technology
29:17
were are finally coming to an age where everyone can make use of them because
29:20
we’re all being flooded by data
29:22
data with we’re all carrying these things are outright cell phones that are
29:26
just a meeting where we are
29:27
like we do what a great opportunity to make sense of that to learn more about
29:31
our world
29:32
and at so I went and you’ve got a job working at sort of me
29:37
the most forward-thinking place I could find a job over the New York Times I
29:42
think white wanna go anyway
29:44
go or face but that’s a great job babe they lack a little bit in the warm fuzzy
29:49
well you know a defending journalism’s a pretty good
29:52
pretty good way to to pay the bills and up but while there
29:56
accounts something really interesting which was
29:59
that data community tech community that was growing
30:02
was super engaged they were in X so excited with what they could do with day
30:07
to the fact that we could all go home and download
30:09
all this government data sport 100 computers to count all the words in
30:12
Wikipedia
30:13
just sitting at home in our underpants incredibly powerful
30:16
have to say that’s always nice to wear but happens
30:20
and and what was cool with outside of like our day jobs
30:23
I would see people gathering is doing the side projects and their
30:27
even these things called hackathons going on Russia curious as I went into
30:30
hackathon
30:32
oh my god I love this audience that is way more than anyone else so
30:36
or amongst friends here arm and i cant if you’re
30:39
as nerdy as I am about these things here prolly super excited
30:43
I’m super excited because I sitting in this room all these guys have a
30:46
amazing computer science skills amazing amazing she learning skills great day to
30:50
scale the thought
30:52
this is how we’re going to make change we don’t need to be doing this at a
30:54
company
30:55
we don’t need government to come into this week at all in this together I
30:58
can’t wait to see what you come up with
31:00
and at the end the weekend we came up with stuff that was sold to
31:03
pressing lean morning Aaron
31:07
people came up with like mobile app’s to find bars
31:10
I by local see all sounds like out man
31:14
we have so many skills are using to make sure that people know what movies to
31:17
watch a good
31:18
where to get the best deal objects like that sucks
31:21
I smaller and the really cool thing and does the same time was this big
31:26
like in big data that big to me means expansive there’s this realization
31:30
everyone’s being touched by data and that now even your smallest clean water
31:33
NGO
31:34
was being inundated with data from surveys their digitizing a random
31:38
overall program they had mobile data even if they can collect any data
31:41
the World Bank and the government’s for opening up data about
31:44
at the demographics in about that region there’s all this potential
31:49
and only necessarily have the skills to work on it so the
31:52
the origin story I’ve Derek and the law is the long lead and decide
31:56
sat down with a buncha friends in a bar in said
31:59
but we know are all doing on the weekend these datasets can we just get like
32:02
Keeva to lend us their data account is work together and see forget
32:06
answer some these questions are built some really cool things around it not
32:09
reporting stuff but
32:10
really answer tough problems and I would probably oppose to say hey anybody in
32:14
New York gonna get together
32:16
more on data for good CI cos and before we knew it we had
32:20
responses from around the world saying yes let’s do this let’s get it I want to
32:23
do that
32:24
I i’m in a hedge fund I need to get some karma points I love you xoxoxo
32:28
this I know it made NGOs and yes according to Thomson data how to get
32:31
people look at it
32:32
and with that the originally Day Without Borders
32:36
now data kind was born and so so ever since we’ve been combining pro bono data
32:40
scientist social organizations
32:42
to build teams is a how can we solve a problem together people getting together
32:45
their skills
32:47
people who’d I are doing good with their skills and maybe could use data better
32:51
coming together to make the world a better place stop there thanks Jake
32:55
and I i love I love this idea like
32:58
harnessing take care of well
33:02
here like work for companies in there and advertising department and they are
33:05
they’re these
33:06
really a PhD in thread and secretly they do want to be in the world
33:10
and Inc what you’re doing is giving them an opportunity to do so so I
33:14
I’ll ask questions for you but let’s move and
33:17
down to Sharmila up I i’ve been really
33:20
really excited to be talking to similar because she is this
33:23
company which i think is doing something really amazing and their little under
33:27
their
33:27
under their radar right now and I i heard them
33:31
a I just for the first and the last little while so
33:34
shamelle you wanna tell us the origin story them secret word Astoria
33:38
a clear some so just %uh by I the start of this story is quite different than
33:44
check story here that it did not start with a hackathon of putting a blog up
33:49
but what to I was doing pop ride is starting her story as though we built a
33:54
Big Data Platform to allow organizations to store data upscale
34:00
and this was a company call Aster Data a
34:03
quite a bit about five years ago and the purpose of that was to
34:07
allow organizations to basically be able to bring data together from multiple
34:12
sources but store
34:14
upscale at about you know it ninety percent cheaper
34:17
what that what traditional data storage platforms while
34:21
a what happened so was actually quite interesting as our
34:25
tad originally I was literally to offer a new Big Data Platform that was more
34:31
effective in terms of cost
34:33
and more effective it tells of share performance
34:37
it turned out that everywhere we went whether it was Angie oo
34:40
as a Chivas up large government organization
34:44
and commercial enterprise what was happening is that
34:48
organizations we’re actually starting to use the solution
34:52
to poll data and from heading to a friend private and public sources that
34:57
actually make it data hub
34:59
up because it was just a cheaper way to actually store a lot of data
35:03
scale so why not start pulling data and from various different sources a start
35:08
right a so we had a organization of that would just basically a mass
35:14
tons and tons of data into our platform
35:17
and what happened at that point plus there whether it gets top was okay down
35:22
what do I do with that
35:23
right so we it’s been it’s great that we can pull it all together it’s great that
35:28
one can store the
35:29
ninety percent cheaper costs a but once it’s all in the
35:33
in there how do you actually harness all the state what do you do with that how
35:36
to extract insights
35:38
and that’s where things are what became of very obvious that
35:42
there were any solutions out there to make it easy
35:45
to actually bring and converged data together from multiple sources that make
35:49
sense of it
35:50
so a lot of what you know Robert just talked about is exactly the problem that
35:56
this story was found it to solve which is really three things one as
36:01
very focused on enabling you to bring data together
36:05
from private and public sources and what we call this is literally data
36:10
convergence
36:11
and it doesn’t matter if the data is big in size small size
36:15
what matters are as the ability to Ashley
36:18
source data from trusted public and web sources
36:22
bring it into the mix with your private data and drive what we call a conversion
36:26
else is so a lot of the examples that use our Leo
36:30
I think some of those examples were actually examples of sources that you
36:34
were bringing together
36:36
and some other sources of data that robert was talking about were more seidl
36:40
views so for instance you can look at Twitter but whether in itself is
36:44
interesting
36:45
however it’s far more interesting when you bring that Twitter data into the mix
36:48
with other private data that you have access to
36:51
or other sources a public tale so if you look at the
36:55
very first thing that you know her story set out to solve the problem are out
37:00
public and private data convergence whether it be Big Data small tit it
37:03
doesn’t matter
37:04
the the situation everyone is facing right now is that data is living and
37:08
many many just for it places
37:11
ranging from large Excel spreadsheets to
37:14
open data that’s available over the web to social sorceress
37:18
to private data repositories and looking at the state %uh silos
37:23
is not a viable solution lot of it’s become critical
37:26
to be able to pull data together from public and private sources and make it
37:30
easier
37:31
to drive what we called it a courageous so picked focus of class story is the
37:36
process of convergence and the intelligence around data convergence
37:41
the second aspect if what we’re doing is
37:45
want to bring data together from multiple different sources and bubble up
37:49
up good trusted sources of public and web and social data
37:53
allow you to combine it with private data the process of doing so and being
37:57
able to then drive a an alice is out of it
38:00
should not take months to do right
38:03
and the reality is in the past and with the pre-existing solutions
38:08
it takes a very long time you have to hire a lot update architects you have to
38:13
have the ability to hire data scientists
38:15
they need to know how to convert all the stay together how to normalize it how to
38:19
bring it all together
38:20
so above the other focus of class story is to accelerate
38:24
the whole process accessing data and driving to a converged for you
38:30
%uh in order to do that again you need basically
38:33
get a new technologies such as the technologies are building
38:37
to be able to accelerate and shrink basically the process from the point of
38:41
accessing data
38:43
to the point of values as being able to see it so basically shrinking the
38:46
pipeline
38:47
and why why would you want to do this it’s exactly for
38:51
the up for the the results that Robert just talked about is that it’s becoming
38:55
more more critical
38:57
to be able to see insights and real-time I N real time can mean many different
39:02
things we actually like to call it
39:03
right time purses real time because real-time
39:07
on the notion of real-time is that you’re seeing it you know
39:10
right there a second as it up false but really what we’re after and this new
39:15
walled-off just for a day though
39:16
is being able to give users the view into the data and view into the insights
39:21
at the right time
39:22
that could mean that some of the data has to be delivered
39:26
in a real-time way it could mean that some other data is not coming in
39:29
real-time
39:30
however you need the ability to see it when you need to see it
39:33
right so if there’s a crisis or there’s an opportunity
39:37
or there’s a flu outbreak you don’t necessarily it doesn’t necessarily to
39:42
mount seeing it in real time it
39:43
avail seeing it at the right time so that you can take action
39:47
so we believe that we need to move to
39:51
solutions that on the stand and enable the concept of right time
39:55
embedded in there could be some sources that happened to deliver real-time data
39:58
as well
39:59
%uh the third thing bad we are focused on and this was a
40:03
glaring issue when I was at you doubt my loss company and we were
40:09
in the possessive delivering a big data platform
40:12
is this notion of Jesus consume ability
40:15
there’s almost nothing out there today that makes it
40:18
easy for a user that doesn’t have a technical skill set
40:22
to comprehend the data and arrive at an insight and know what to do with that
40:27
insight so how do you actually
40:29
basically on accelerate the process of converting data from multiple sources
40:35
and then driving to insights and presenting in size
40:39
that users can actually take action on a so we are changing the whole model or
40:45
on that day and to use our side of things as well in terms of
40:49
delivering a new use a bottle that makes it far more intuitive
40:53
and farm all guided in terms of how you actually deliver a
40:57
insight to use art how they actually drive to it in sight
41:01
and that’s where this notion and that’s basically the
41:04
up the roots of the company name so class story
41:07
is truly about and now also as our data that’s being brought together from
41:12
multiple different sources to deliver a story
41:15
but that story has to be into for a user that it doesn’t necessarily have a
41:19
technical skill set
41:21
a the other thing that we recognize and the New Years a model for
41:25
for just for a date out is that they’re far more people
41:29
%uh situated across distributed teams
41:32
that need to see data they used to be you know looking back long time ago
41:36
a lot of the data experts were sitting in the same room
41:39
and they will huddle together in the same location then you can
41:42
you could basically you know exchange ideas and pick each other’s brains
41:46
and understand basically how to drive site so the data
41:49
where they’re very different world now where people are bringing data together
41:53
from many different sources they located in many different locations
41:57
andy is distributed data teams need and users
42:00
need a better way to be able to see you highlights
42:03
and share insights and understand basically what to draw out of it
42:08
so they’re really three things like a subclass stories very focused on
42:13
is this notion of public and private data convergence in fact to the public
42:16
realm and well
42:17
what web data out well there’s about
42:21
8,000 Open Data APIs and if you look back around eight years ago there are
42:25
about 50 Open Data APIs right
42:28
if you look at the sources a up
42:32
a social data that i’ve got social indicators they’re about 370
42:36
options out there right now for good trusted sources of data
42:40
%uh we recognize that nobody can Ashley’s filter through thousands and
42:44
thousands of sources themselves and that you need to be able to bubble up
42:49
the trusted sources of data the high quality data
42:52
and then try met that make it possible to try this gorgeous
42:56
Excel writing process to do that as I just talked about and finally
42:59
really delivering any news a model some more users can actually work with date
43:04
out
43:04
that what we’ve been dealing with in the past which is just relying on data
43:09
experts or small groups who are
43:11
who basically have the expertise to do sell up and delivering them to use a
43:15
model we basically
43:16
bring into it a lot of I’ll concepts from the consumer world
43:21
and what what what has happened in terms of consumer applications and as you know
43:26
many consumer
43:26
many applications have become a whole lot easier to use because of certain
43:29
innovation around it
43:31
bringing a lot of those ideas into the world of data and making it easier for
43:36
users who have non technical skill sets to actually work with data
43:40
grape okay 7 I’m gonna cross my legs this way to
43:44
10 that to mean moderator yeah
43:49
rather you what you work with you work with data that that
43:53
is collected through social networks that is collected through
43:57
I various various ways in which most often the person has provided data isn’t
44:02
aware of what you’re doing with the data
44:04
so you know there’s a term for that which we call opportunistic sensing read
44:08
it’s like
44:08
here year you’re here your using people stay there
44:12
and somewhere the nine-page User Agreement that they signed said that
44:16
their data become
44:16
public but I’m not sure let the people who are using the system’s kinda
44:19
understand that their date is being used in that way
44:22
so I just want to ask that how you guys negotiate
44:25
as a as a governmental are the as a big big global organization
44:29
have you negotiate that that and the ethical quandary
44:34
as you’re using people’s data without them knowing that your use
44:37
why I think we’re in in the at the beginnings of a paradigm shift that will
44:42
ultimately be about
44:44
not companies giving putting
44:47
other putting customers data into a common individuals being empowered
44:51
to actually decide how their data can be reused
44:56
that’s the world we want to get to know soon as possible up we don’t
45:00
in our research we don’t actually ever receive personally identifiable
45:04
information
45:05
any kind and we’re looking at the policy dimensions so we’re looking primarily an
45:08
aggregated data
45:10
we’re looking we only receive data that’s been on my list
45:14
I so we’re not trying to to find out that somebody lost his job had to sell
45:17
the family hell
45:19
the idea is really ok it has been a 300 percent
45:22
increase attempts to sell livestock in this district
45:25
that’s where we need to go to the next household survey right now
45:28
so we’re really looking for patterns and Anopheles an aggregated data
45:32
up you know you can do a lot with a day at that has identity and at
45:36
up but the researchers huge today the reason why I ask is because I think that
45:41
you guys
45:41
a and are doing such great things but if
45:45
if Google were to say they were being very similar things people
45:48
kinda there’s an outcry that happens there and what they really have you see
45:52
the difference there
45:52
well there’s research on this I mean it’s interesting if Facebook tells you
45:56
you’re seventy percent likely to break up with your girlfriend next tuesday
45:59
people feel very invaded
46:00
I there are others that do this are but if Amazon says
46:04
you know people who bought this book also bought that book we like it
46:08
if Google searches protect flu outbreaks we see the public good
46:11
and other words you know individuals as individuals as producers
46:15
data are we may not all be as educated as we should be about how our date is
46:20
being sold behind our backs and moved around the world
46:23
used to make a lot of money but what we do understand is that
46:26
you know the data has potential and if I see the benefit of it
46:30
that changes how I weigh the risks and benefits
46:33
reuse what we’re trying to get to is to say you know
46:36
there’s this huge opportunity in the state it’s not just about
46:39
commercial profit we need to be more educated and we need to generate
46:43
evidence that you can use it for good because that will get to a more nuanced
46:46
conversation our application
46:49
and and actually that ties into question for shame lol itself
46:53
up did as property we we
46:57
a I’m very interested in the idea data ownership and the idea that
47:01
the me using my cellphone producing data on
47:05
on that cell phone is illegal the legal boundaries are fuzzy but they
47:09
they are more often than not in the in their favor if the user and that that
47:14
that information that’s being created as the user’s data however there are many
47:17
many companies
47:19
many of which you probably partner with who who
47:22
hold that date and they counted as as property above the company itself right
47:26
I guess I have
47:27
I have a question about about in your interactions with that companies that
47:32
you partner with
47:33
is this something that those companies are thinking about first about and the
47:36
second about
47:36
how do you think we could or do you think it’s worthwhile to even get to
47:40
that point %uh
47:41
Roberts and at date data Commons is there going to be a point in which these
47:45
companies will be more open to share their data
47:48
with the knowledge that that data that can help my the world
47:53
so I’m great questions so what we’re actually very suave pleasantly surprised
47:57
to see is that a lot of the
47:59
companies that have been in the up this also provide a high value to it out
48:04
a very interested in seeing that day that process followed a the
48:08
with other data sources because they’re appreciate
48:11
that if you could take that data and cross pollinate the update out
48:16
with other sources of data you’re gonna drive to rate recite
48:19
and they’re all driving to what’s trying to find out how to get
48:22
out if just the data program business into the business of insights
48:27
and to get into the business for insights that means opening up their
48:30
data and making it more conceivable an easy to cross holiday
48:33
now on one of the things that’s putting pressure are
48:37
on these but these companies to do that’s where previously they were sort
48:42
of protecting that data closely
48:44
is that what’s happened over across the web
48:48
and social sources is that look if you are not going to open up your data out
48:52
there are other places I could go that didn’t exist
48:55
before right so previously you want to store
48:59
trusted sources update on they were the only places you could go if you want to
49:03
wanted it and now there are many more options as I mentioned there are
49:07
thousands of places that you could go to across the web to call interesting data
49:11
out now grounded you have to
49:13
look for the data quality and the trusted sources
49:16
there are sources now like social data sources where you can get very
49:20
interesting insights out of it
49:22
and it’s not you know pretty a tightly held right
49:26
so we are at a place now where a lot of the
49:29
phenomena that has happened over the web and social networks as placing pressure
49:34
are
49:35
on other providers update it to open up the data and make it more readily usable
49:40
so that you can cross pollinated with other sources and drive to the inside so
49:44
I will say that this transition as you know
49:47
completely happened yet it’s in it’s happening and we’re seeing it unfold in
49:52
front of us
49:53
but I we believe that it will happen and we believe that the phenomenon across
49:58
the web and got social networks
50:00
actually accelerating the the move it from
50:04
you know companies protecting their data very closely to making it more open
50:08
the other thing I would add to this is that ass %uh whether it’s NGOs
50:12
government organizations
50:13
commercial enterprise as they look at data from third party sources and social
50:18
sources
50:19
they themselves are very interested in a mechanism to share data of
50:24
of interest right so I find a source of data I mash it up with another source
50:29
update out
50:30
and I’ve doubt creative basically a mash tater product if there’s something
50:33
interesting in that I should be able to they’ll share that
50:37
and a data marketplace essentially with others that might find out of value
50:42
so I’m sure actually with you know what what you do
50:45
at a lot of the data that should look at this interest in actually making its
50:50
share of a lot opening up a marketplace where more people can get to it
50:54
is probably becoming have high interest sourcing two phenomena assist the
50:58
Paris you know data data providers are transitioning over
51:03
to make that data more easily available see cross pollinated and their
51:07
from the use around the fat there’s a big poll two words
51:11
making data sharable in the marketplace as more people fight
51:14
high-value source of data thanks and
51:18
we can open up to to to questions for the room in a second but i wanna and
51:21
with with
51:22
a question for Jake because and I know I know that many people
51:26
in this audience have organizations that
51:29
are probably very eager to to to to
51:32
get working with data kind this idea as these
51:36
really world-class they decided to spare shooting in
51:39
I like some their Strikeforce have some kind tale like help you solve your data
51:43
problems
51:44
how do you balance of how do you balance this
51:47
I think this really accident excellent mechanism you have
51:50
with the very I think problematic question domain expertise
51:54
thread so there’s all these people who have been working on these problems for
51:58
a year
51:58
the same an accountancy or team love
52:02
love data ninjas what what do you do to like
52:06
to make that excelerate that like here’s have learning
52:09
and Taylor into like a very for that something I’ve been very interested
52:13
about that said
52:15
we’ve never met before yeah that’s a great question
52:18
an ass I’ve never heard of data scientists referred to as ninjas are
52:22
like strike forces it seems like I 93 reading then yeah exactly happened here
52:27
s
52:27
but i i think it’s a awesome question and and one that we do address
52:30
very early on because what you’ll see is that there’s almost a geek hubris
52:35
in the data science community you’ll hear things like let’s hack
52:38
education and let’s hack government yes and that this flip idea that technology
52:43
can solve everything technologies have the solutions
52:46
and we recognize very early on that while we may be very good with
52:49
technology and data
52:50
we don’t know all the problem areas where working at I don’t know the issues
52:54
in poverty or clean water
52:56
and so we started very early on with real what we feel ur bro collaborations
53:01
between organizations and a scientist to say that just out of this your data and
53:05
we’ll tell you what’s in it and give it back but really sitting down with people
53:07
in saying
53:08
what are the big issues here what what’s really
53:11
you know troubling you and there’s this amazing
53:15
I don’t buy interactions is almost two-step happens
53:18
when these projects start where years ago something like this
53:22
organization walks in and says I have all this data
53:26
what do I do of it’s easy where starts in the very first thing you wanna say is
53:30
released a science is there so you know it’s funny it’s it’s actually not at all
53:34
about the data russians are not at all but it’s not really the place to start
53:37
actually starts with the questions
53:39
you want to answer so what would you go to really ask it what sucks most
53:43
about what you do every day was the high I’m sure everyone has an answer that
53:48
like
53:48
up I hate that I have to look for all these 100 files by hand or hate that I
53:53
have to find
53:54
you know that I’m always wondering how people use in our programs I’ve no idea
53:58
until next survey comes out next year
54:00
on and so from there this is funny little back and forth with a data
54:03
scientist will take the date and say okay
54:05
doing what you said I’m going to XY and z. with this dataset
54:09
and out that’s that’s fine I don’t think I’ve action Y and Z are important that
54:13
we think this
54:14
other thing that we have data for and goes back and forth this really great
54:17
way to where you get this
54:18
kinda melding on the subject expertise what really matters and has to be used
54:23
the data scientists who can bring in certain the ideas and creativity about
54:26
what’s possible so that’s something we have satellite is
54:29
that partnership andress thanks okay so let’s open up the questions and and
54:34
and I just ask that that but when I
54:37
pointy for a question like the first question we’re gonna take from here you
54:40
this way for the maker found because we’re recording and
54:43
and I’ll here at will hear it more loudly so the first question will go to
54:47
the
54:48
in the red I’m hired Carol come from edleman
54:52
arm there’s over a hundred and thirty one billion dollar spent advertising
54:58
and there’s always is jerking off fifty percent of advertising works in fifty
55:01
percent doesn’t
55:02
I spent my entire career trying to show two companies
55:06
that if they adopt social issues and if they have the right connections and
55:10
communications
55:12
that they can have a deeper relationship with their employees their communities
55:15
current customers a future customers
55:17
so my question to you is this a lot of it data
55:21
on is your help by manufacturers or retailers are Neil sinners such
55:26
is there some way not that will answer today but is there a way
55:30
that you can take data if a company is standing for something
55:35
and it is begun to be on package or communicated or socially
55:40
that it is more powerful and so they can stop spending their 131 billion
55:45
advertising
55:47
put into more of a social issues engagement and have much more effective
55:51
relationships business plus social purpose
55:57
yeah I got worse we’re certainly seeing
56:00
that kind of consciousness emerge in the organizations that were engaging with
56:05
partly because you know their customers a more data savvy today
56:09
they understand data other people that they’re recruiting or Millennials
56:14
right so they’re coming and they’re saying I want to do well but i wanna do
56:16
good to
56:18
I’ll the at the business model that’s driving a lot of companies today
56:22
is the data produced audio business model which is more has higher margins
56:25
than
56:26
no business model I’ll and you know we’ve we’ve seen for example in our
56:30
discussions with other telcos and Indonesia
56:33
we have all 10 telcos the table and discussions about cooling
56:37
all data from past years into a comment and serving out the company name and
56:40
making it safe to work with
56:42
are and they’re saying you know we want our data scientists
56:45
not we don’t give you our data just you know let it go
56:49
we were actually Co designed the research projects and get our people
56:52
involved because it’s a huge morale boost
56:55
and if the you know if our if our subscribers know that we’re actually
56:58
using aggregate data
56:59
to understand you know migration to cities and the impact of rising food
57:04
prices that’s a win for the company
57:06
so there’s PR here and there’s more out here
57:10
have the I I always say that there’s a there’s a good space for companies to
57:14
to brand themselves as data ethical the same way that we saw happen with her
57:18
with the serbs green FX and and and companies who branded themselves that
57:22
way and I
57:22
I’ve been wondering for awhile would be the first to do that track my name’s
57:26
Chris however for the Salesforce foundation sells for stock on
57:29
so I mean we wrestle with this because we have eighteen thousand
57:32
not-for-profits that you Salesforce
57:34
in with that’s a big data set so we wrestle with how do you get them
57:38
connected
57:39
and then their stories interesting I have was going the technology pictures
57:42
to see how you can possibly
57:43
that’s about a sales force in you know connected therewith radian6 and stuff
57:47
but
57:48
that’s one of the things that we saw Russell with we love to see more of that
57:50
we’re just not sure how
57:52
we actually you know actually need to get the not-for-profits to agree to
57:55
share their data
57:56
to then run the report so I’m actually seeing the opposite right now as I work
58:00
with big not-for-profits they don’t want to share their data
58:02
because it’s very proprietary provided as property again right yeah the only
58:06
example that I see just a specific example
58:08
is we work really closely with homeless Link so it’s a membership association of
58:12
homelessness organisations in the UK
58:14
and the Big Lottery funded this project to do data analysis around
58:19
you know how most people homeless services and they’re gonna get outta
58:22
Salesforce
58:22
other sources as well and then they’re doing a research project at the Big
58:27
Lottery Fund here
58:27
in the UK sponsored but you know that’s you know one of the few examples I can
58:31
think up with
58:32
share data pools with you know specifics you know dr. anonymized customer data
58:37
but you know we’ve got a lot of customers apply today to be called
58:40
you note the not-for-profit to share so that he did not say reports
58:44
I want to just a address the particular question their
58:48
your question on a the effectiveness of
58:51
you know advertising and showing that if it’s put to words
58:55
yes social issues that consumers actually respond
58:58
that are so there was some so we have actually
59:02
with class stories have shown companies that
59:06
if their ad dollars up put to what’s that cause there is actually a better
59:10
market
59:11
responsive consumer adoption and
59:14
interestingly there was what was one very large conceive are packed with
59:18
companies that did a pic
59:19
campaign and instead of using the traditional advertising channels Ashley
59:23
use the social channels
59:25
to promote the campaign around yes saving water in certain locations at so
59:30
it took on
59:31
it was a very very well designed campaign of course driven by there are
59:35
lots of creativity around it
59:37
but the response to it was phenomenal and
59:41
phenomenon that response actually could be measured
59:44
through the social channels as people were across the world
59:49
starting to respond to this campaign which then
59:52
got you know certain products flying off the shelf that will vary so the Pecos at
59:57
me
59:57
a friendly products and that until I think
60:01
you know it close story was seen as a platform that lets you look at all this
60:04
data from multiple sources and be able to measure our
60:08
what’s working where right %uh and that is and
60:12
in the other thing you doing in that particular example is looking at
60:15
everything geo to GL
60:16
and by population by demographics was responding to what
60:20
on and so that’s an example if being able to demonstrate
60:24
to organizations that if they do put that dollars to what’s these types of
60:29
campaigns and advertising that way
60:31
it likely as though it does have a very positive affect
60:35
but again this is back to measuring
60:38
right you off one if you’re up three points as you know not just about
60:42
collecting the data
60:43
but measuring the effectiveness right and down to
60:46
every location geo zip code population to see how people respond day
60:53
and weekly shockingly are running out of time in our
60:56
in our first period if this were going to continue but i wanna take one one
60:59
more question
61:00
on camera and then and and then we’ll
61:03
I’ll loosen up and and and
61:06
get more rowdy and stuff cameras to at
61:10
I don’t know who’s here hand at first I saw you for simply go to the gym but the
61:14
microphone to the top
61:16
high and Marc Choma for macro and I wondered
61:19
I wanted get a sense the palace at what proportion at the smartest invites
61:24
people
61:25
most active in this space County working under and
61:29
advertising funded business models
61:33
and I i wondered what we can do to I mean obviously
61:36
that might be okay but I suspect to be better if we had
61:39
I loved working under alternative business models but i just wondered
61:43
what proportion you think it is in your thoughts on out
61:46
chase the I A don’t know an exact number and fried
61:49
hesitate to put a date out that I couldn’t back up up but i wasnt vast
61:53
majority
61:54
and the questionnaire back actually is
61:57
you mean company like companies that employ data scientists that are
62:00
make the revenue around advertising base models and how would we change
62:04
like Google’s model to be of advertising or how do we get
62:07
people who currently get six-figure jobs in advertising to
62:11
go elsewhere now I mean yeah anywhere
62:14
organizations that your interest this but ultimately
62:17
me whereas come for pair its current
62:22
his hour or so we’re almost there
62:25
interesting I am are you may have very different there’s a
62:29
a data scientist a chose toria
62:32
his talk all the time grey hat data science
62:36
and black that they decide to do we have seen like the really catastrophic uses
62:40
data science yet for harm but
62:42
ever almost everybody working on MySpace is an assorted ethically murky
62:46
think field because at you know at best you’re trying to change human behavior
62:50
to get people to spend more money
62:52
this is why you know certainly in the UN we’re getting
62:56
some of the you note the people are typically recovered particle physicists
63:00
I’m um at their hugely interest in doing this because like you can figure out how
63:04
to get people to click on out
63:06
or you can fight hunger poverty and disease I mean we can’t pay with the
63:09
googles pay
63:10
but there’s a lot of interest out there its profits because the vast majority
63:14
is solely focused on art now agree with that
63:18
and I think we have time for one more question so before we go
63:21
for we switch the camera of several will go over here
63:27
red-haired they work for our government Development Agency
63:32
cult ok the overseas Private Investment Corporation
63:35
and for years I’ve thought about it an engineer
63:39
packed I’m Strikeforce
63:43
Ninja hack up and spanked I he got that
63:48
especially as it pertains to the work week do here when we come together its
63:52
goal
63:54
and that is that I perceive this great number of people that wanna give
63:59
or invest money products and services
64:04
and themselves people and they want to connect to things that are
64:09
good and vetted yet simple
64:14
and they are cross impact factors
64:17
agriculture health care all the things that we care about and development
64:21
but they don’t know how to connect and if there were to be a Google Map
64:28
to the the block
64:32
and if it was not so heavily laden with data that people with fees up at the
64:37
thought
64:39
the rather someone anywhere anytime today
64:43
wants to give eight hundred dollars famine church group
64:48
to fight ha my own children
64:52
in their village in India
64:56
how do we connect those two dots with just
64:59
a simple transporting
65:03
doc to another website
65:07
how hard would that be and
65:11
I that’s a great question and I know and related
65:14
answer and akiva people are sitting here in the audience who I know are trying to
65:18
match up donors directly to people who need that around the world so
65:21
it clearly possible and more than possible successful
65:26
I I think that challenge and intermittent sugar other
65:29
giving group for people in this room us know that allow you to find
65:32
group in 1982 here you know
65:35
work either how do you do that name
65:39
so that it has one may have been
65:43
there’s no one site that ties all the other sites together if you don’t know
65:47
the word
65:48
capital you not gon and that fail to get to the smallholder farmer
65:52
and even this is the big secret of all time
65:55
if you work in a major g8 development finance institution
66:01
you don’t know necessarily who from government
66:05
who from the private sector and who from the world not-for-profits
66:09
is active fighting malaria in a given play
66:13
and yet if we all knew that together we could save
66:17
so much money and we can take away the time that NGOs and
66:22
others spend just trying to figure that out
66:26
and redeploy that money and I think it is a like a couple dollars
66:31
yeah know it’s I think okay the great point and I think some people in this
66:34
room may actually be working on a preliminary data problem to that
66:38
to that hasn’t yet there is not a unique identifiers for each charity or
66:42
non-profit or social contests that I’m aware oven
66:45
that starts as a inhibiting factor because then
66:49
everybody’s gotta go on and say I’m this charity and I’m doing this and I gotta
66:52
keep it up today and I run into all the data problems that everybody hates
66:55
around any kind of reporting
66:56
an exception interns their update that all the time it gets old and stale no
67:00
news
67:01
so there’s actually a great effort out there now i i to you come
67:04
create one consolidated almost like a tax number
67:08
used by corporations I for all
67:11
NGOs are charities around the world that’s not going to release all that
67:14
entire problem but I think that’s an exciting day to problem
67:17
getting one step closer and I get to say here all the NGOs operating and maybe
67:22
the next step is here all the ones working in these topic areas and here’s
67:25
where they’re located because that ought to go in on
67:27
tax forms and registration forms so I’m very
67:30
a bullish about that that movement and people should watch the
67:34
the guide stars and and the IRS people like that at least in the US that ur
67:39
working on things like that so there’s hope but I agree which
67:43
we should see that and so said just before we transition
67:47
we have 1 minute left on camera we’re gonna go to where you can be questions
67:50
and second but
67:51
on at big this word a day to drive me crazy so
67:56
over time that rebranding up at my friends and I are always says that it’s
68:01
not about that date it’s actually about granular data
68:04
we’re gonna the data the data points are smaller and we have more of them
68:08
packed with exciting so granular data it’s not gonna sell
68:12
but if you have to rebranding data with the word
68:15
if swell something data what will it be
68:19
well it could just be data but there’s a lot more other I think
68:23
few if you if you realize people are all around us
68:27
doing all these transactions are producing a date on its coming from
68:29
mobile devices
68:30
the date is passing through our bodies right now gigabytes of data is passing
68:35
through your body as you sit here
68:36
ambient data ambient air have to take a left
68:39
well I’m a big fan on up the slow movement
68:42
some looking for slow day to it personally
68:46
a carefully crafted to data every
68:49
routes 11 traveling with me if mister I F
68:53
sherri I’ll up to varsity so data diversity
68:57
Ted I think that you know it’s about just more
69:00
and more sources that’s only gonna get worse up
69:04
it’s gonna be or sources to tap into just like I said
69:07
with the sizes a pic i right
69:11
cell I’m now the cameras down we can square freely
69:15
so a.m. at let’s go to the back for a question hopefully laced with profanity
69:20
yeah