What do we do with all this big data?
-
0:01 - 0:04Technology has brought us so much:
-
0:04 - 0:07the moon landing, the Internet,
-
0:07 - 0:09the ability to sequence the human genome.
-
0:09 - 0:13But it also taps into a lot of our deepest fears,
-
0:13 - 0:15and about 30 years ago,
-
0:15 - 0:17the culture critic Neil Postman wrote a book
-
0:17 - 0:19called "Amusing Ourselves to Death,"
-
0:19 - 0:22which lays this out really brilliantly.
-
0:22 - 0:24And here's what he said,
-
0:24 - 0:26comparing the dystopian visions
-
0:26 - 0:30of George Orwell and Aldous Huxley.
-
0:30 - 0:33He said, Orwell feared we would become
-
0:33 - 0:35a captive culture.
-
0:35 - 0:39Huxley feared we would become a trivial culture.
-
0:39 - 0:41Orwell feared the truth would be
-
0:41 - 0:43concealed from us,
-
0:43 - 0:45and Huxley feared we would be drowned
-
0:45 - 0:48in a sea of irrelevance.
-
0:48 - 0:50In a nutshell, it's a choice between
-
0:50 - 0:52Big Brother watching you
-
0:52 - 0:55and you watching Big Brother.
-
0:55 - 0:57(Laughter)
-
0:57 - 0:59But it doesn't have to be this way.
-
0:59 - 1:02We are not passive consumers
of data and technology. -
1:02 - 1:04We shape the role it plays in our lives
-
1:04 - 1:07and the way we make meaning from it,
-
1:07 - 1:08but to do that,
-
1:08 - 1:12we have to pay as much attention to how we think
-
1:12 - 1:14as how we code.
-
1:14 - 1:17We have to ask questions, and hard questions,
-
1:17 - 1:19to move past counting things
-
1:19 - 1:21to understanding them.
-
1:21 - 1:24We're constantly bombarded with stories
-
1:24 - 1:26about how much data there is in the world,
-
1:26 - 1:28but when it comes to big data
-
1:28 - 1:30and the challenges of interpreting it,
-
1:30 - 1:32size isn't everything.
-
1:32 - 1:35There's also the speed at which it moves,
-
1:35 - 1:37and the many varieties of data types,
-
1:37 - 1:40and here are just a few examples:
-
1:40 - 1:42images,
-
1:42 - 1:46text,
-
1:46 - 1:48video,
-
1:48 - 1:50audio.
-
1:50 - 1:53And what unites this disparate types of data
-
1:53 - 1:55is that they're created by people
-
1:55 - 1:58and they require context.
-
1:58 - 2:00Now, there's a group of data scientists
-
2:00 - 2:02out of the University of Illinois-Chicago,
-
2:02 - 2:05and they're called the Health Media Collaboratory,
-
2:05 - 2:08and they've been working with
the Centers for Disease Control -
2:08 - 2:09to better understand
-
2:09 - 2:12how people talk about quitting smoking,
-
2:12 - 2:15how they talk about electronic cigarettes,
-
2:15 - 2:17and what they can do collectively
-
2:17 - 2:19to help them quit.
-
2:19 - 2:21The interesting thing is, if you want to understand
-
2:21 - 2:23how people talk about smoking,
-
2:23 - 2:25first you have to understand
-
2:25 - 2:27what they mean when they say "smoking."
-
2:27 - 2:31And on Twitter, there are four main categories:
-
2:31 - 2:34number one, smoking cigarettes;
-
2:34 - 2:37number two, smoking marijuana;
-
2:37 - 2:40number three, smoking ribs;
-
2:40 - 2:43and number four, smoking hot women.
-
2:43 - 2:46(Laughter)
-
2:46 - 2:49So then you have to think about, well,
-
2:49 - 2:51how do people talk about electronic cigarettes?
-
2:51 - 2:53And there are so many different ways
-
2:53 - 2:55that people do this, and you can see from the slide
-
2:55 - 2:58it's a complex kind of a query.
-
2:58 - 3:01And what it reminds us is that
-
3:01 - 3:04language is created by people,
-
3:04 - 3:06and people are messy and we're complex
-
3:06 - 3:09and we use metaphors and slang and jargon
-
3:09 - 3:12and we do this 24/7 in many, many languages,
-
3:12 - 3:15and then as soon as we figure it out, we change it up.
-
3:15 - 3:20So did these ads that the CDC put on,
-
3:20 - 3:23these television ads that featured a woman
-
3:23 - 3:25with a hole in her throat and that were very graphic
-
3:25 - 3:27and very disturbing,
-
3:27 - 3:29did they actually have an impact
-
3:29 - 3:31on whether people quit?
-
3:31 - 3:35And the Health Media Collaboratory
respected the limits of their data, -
3:35 - 3:37but they were able to conclude
-
3:37 - 3:40that those advertisements —
and you may have seen them — -
3:40 - 3:42that they had the effect of jolting people
-
3:42 - 3:44into a thought process
-
3:44 - 3:48that may have an impact on future behavior.
-
3:48 - 3:52And what I admire and
appreciate about this project, -
3:52 - 3:53aside from the fact, including the fact
-
3:53 - 3:57that it's based on real human need,
-
3:57 - 4:00is that it's a fantastic example of courage
-
4:00 - 4:05in the face of a sea of irrelevance.
-
4:05 - 4:08And so it's not just big data that causes
-
4:08 - 4:11challenges of interpretation, because let's face it,
-
4:11 - 4:13we human beings have a very rich history
-
4:13 - 4:16of taking any amount of data, no matter how small,
-
4:16 - 4:17and screwing it up.
-
4:17 - 4:21So many years ago, you may remember
-
4:21 - 4:24that former President Ronald Reagan
-
4:24 - 4:25was very criticized for making a statement
-
4:25 - 4:29that facts are stupid things.
-
4:29 - 4:31And it was a slip of the tongue, let's be fair.
-
4:31 - 4:34He actually meant to quote John Adams' defense
-
4:34 - 4:36of British soldiers in the Boston Massacre trials
-
4:36 - 4:40that facts are stubborn things.
-
4:40 - 4:42But I actually think there's
-
4:42 - 4:46a bit of accidental wisdom in what he said,
-
4:46 - 4:48because facts are stubborn things,
-
4:48 - 4:51but sometimes they're stupid, too.
-
4:51 - 4:53I want to tell you a personal story
-
4:53 - 4:57about why this matters a lot to me.
-
4:57 - 4:59I need to take a breath.
-
4:59 - 5:02My son Isaac, when he was two,
-
5:02 - 5:04was diagnosed with autism,
-
5:04 - 5:07and he was this happy, hilarious,
-
5:07 - 5:09loving, affectionate little guy,
-
5:09 - 5:12but the metrics on his developmental evaluations,
-
5:12 - 5:14which looked at things like
the number of words — -
5:14 - 5:17at that point, none —
-
5:17 - 5:21communicative gestures and minimal eye contact,
-
5:21 - 5:23put his developmental level
-
5:23 - 5:27at that of a nine-month-old baby.
-
5:27 - 5:30And the diagnosis was factually correct,
-
5:30 - 5:33but it didn't tell the whole story.
-
5:33 - 5:35And about a year and a half later,
-
5:35 - 5:37when he was almost four,
-
5:37 - 5:39I found him in front of the computer one day
-
5:39 - 5:45running a Google image search on women,
-
5:45 - 5:48spelled "w-i-m-e-n."
-
5:48 - 5:51And I did what any obsessed parent would do,
-
5:51 - 5:53which is immediately started
hitting the "back" button -
5:53 - 5:56to see what else he'd been searching for.
-
5:56 - 5:58And they were, in order: men,
-
5:58 - 6:06school, bus and computer.
-
6:06 - 6:08And I was stunned,
-
6:08 - 6:10because we didn't know that he could spell,
-
6:10 - 6:12much less read, and so I asked him,
-
6:12 - 6:14"Isaac, how did you do this?"
-
6:14 - 6:16And he looked at me very seriously and said,
-
6:16 - 6:20"Typed in the box."
-
6:20 - 6:23He was teaching himself to communicate,
-
6:23 - 6:26but we were looking in the wrong place,
-
6:26 - 6:29and this is what happens when assessments
-
6:29 - 6:31and analytics overvalue one metric —
-
6:31 - 6:34in this case, verbal communication —
-
6:34 - 6:39and undervalue others, such
as creative problem-solving. -
6:39 - 6:42Communication was hard for Isaac,
-
6:42 - 6:44and so he found a workaround
-
6:44 - 6:47to find out what he needed to know.
-
6:47 - 6:48And when you think about it, it makes a lot of sense,
-
6:48 - 6:51because forming a question
-
6:51 - 6:53is a really complex process,
-
6:53 - 6:56but he could get himself a lot of the way there
-
6:56 - 7:00by putting a word in a search box.
-
7:00 - 7:03And so this little moment
-
7:03 - 7:05had a really profound impact on me
-
7:05 - 7:07and our family
-
7:07 - 7:10because it helped us change our frame of reference
-
7:10 - 7:12for what was going on with him,
-
7:12 - 7:15and worry a little bit less and appreciate
-
7:15 - 7:17his resourcefulness more.
-
7:17 - 7:20Facts are stupid things.
-
7:20 - 7:23And they're vulnerable to misuse,
-
7:23 - 7:24willful or otherwise.
-
7:24 - 7:27I have a friend, Emily Willingham, who's a scientist,
-
7:27 - 7:30and she wrote a piece for Forbes not long ago
-
7:30 - 7:32entitled "The 10 Weirdest Things
-
7:32 - 7:34Ever Linked to Autism."
-
7:34 - 7:37It's quite a list.
-
7:37 - 7:40The Internet, blamed for everything, right?
-
7:40 - 7:44And of course mothers, because.
-
7:44 - 7:46And actually, wait, there's more,
-
7:46 - 7:49there's a whole bunch in
the "mother" category here. -
7:49 - 7:54And you can see it's a pretty
rich and interesting list. -
7:54 - 7:56I'm a big fan of
-
7:56 - 8:00being pregnant near freeways, personally.
-
8:00 - 8:01The final one is interesting,
-
8:01 - 8:04because the term "refrigerator mother"
-
8:04 - 8:07was actually the original hypothesis
-
8:07 - 8:08for the cause of autism,
-
8:08 - 8:11and that meant somebody
who was cold and unloving. -
8:11 - 8:13And at this point, you might be thinking,
-
8:13 - 8:14"Okay, Susan, we get it,
-
8:14 - 8:16you can take data, you can
make it mean anything." -
8:16 - 8:21And this is true, it's absolutely true,
-
8:21 - 8:26but the challenge is that
-
8:26 - 8:29we have this opportunity
-
8:29 - 8:31to try to make meaning out of it ourselves,
-
8:31 - 8:37because frankly, data doesn't
create meaning. We do. -
8:37 - 8:40So as businesspeople, as consumers,
-
8:40 - 8:42as patients, as citizens,
-
8:42 - 8:45we have a responsibility, I think,
-
8:45 - 8:47to spend more time
-
8:47 - 8:50focusing on our critical thinking skills.
-
8:50 - 8:51Why?
-
8:51 - 8:54Because at this point in our history, as we've heard
-
8:54 - 8:56many times over,
-
8:56 - 8:58we can process exabytes of data
-
8:58 - 9:00at lightning speed,
-
9:00 - 9:03and we have the potential to make bad decisions
-
9:03 - 9:05far more quickly, efficiently,
-
9:05 - 9:10and with far greater impact than we did in the past.
-
9:10 - 9:12Great, right?
-
9:12 - 9:15And so what we need to do instead
-
9:15 - 9:17is spend a little bit more time
-
9:17 - 9:20on things like the humanities
-
9:20 - 9:23and sociology, and the social sciences,
-
9:23 - 9:26rhetoric, philosophy, ethics,
-
9:26 - 9:28because they give us context that is so important
-
9:28 - 9:31for big data, and because
-
9:31 - 9:33they help us become better critical thinkers.
-
9:33 - 9:38Because after all, if I can spot
-
9:38 - 9:40a problem in an argument, it doesn't much matter
-
9:40 - 9:43whether it's expressed in words or in numbers.
-
9:43 - 9:46And this means
-
9:46 - 9:50teaching ourselves to find
those confirmation biases -
9:50 - 9:52and false correlations
-
9:52 - 9:54and being able to spot a naked emotional appeal
-
9:54 - 9:56from 30 yards,
-
9:56 - 9:58because something that happens after something
-
9:58 - 10:01doesn't mean it happened
because of it, necessarily, -
10:01 - 10:03and if you'll let me geek out on you for a second,
-
10:03 - 10:08the Romans called this
"post hoc ergo propter hoc," -
10:08 - 10:11after which therefore because of which.
-
10:11 - 10:15And it means questioning
disciplines like demographics. -
10:15 - 10:17Why? Because they're based on assumptions
-
10:17 - 10:20about who we all are based on our gender
-
10:20 - 10:21and our age and where we live
-
10:21 - 10:24as opposed to data on what
we actually think and do. -
10:24 - 10:26And since we have this data,
-
10:26 - 10:29we need to treat it with appropriate privacy controls
-
10:29 - 10:33and consumer opt-in,
-
10:33 - 10:36and beyond that, we need to be clear
-
10:36 - 10:38about our hypotheses,
-
10:38 - 10:41the methodologies that we use,
-
10:41 - 10:43and our confidence in the result.
-
10:43 - 10:46As my high school algebra teacher used to say,
-
10:46 - 10:47show your math,
-
10:47 - 10:51because if I don't know what steps you took,
-
10:51 - 10:53I don't know what steps you didn't take,
-
10:53 - 10:55and if I don't know what questions you asked,
-
10:55 - 10:58I don't know what questions you didn't ask.
-
10:58 - 11:00And it means asking ourselves, really,
-
11:00 - 11:01the hardest question of all:
-
11:01 - 11:05Did the data really show us this,
-
11:05 - 11:07or does the result make us feel
-
11:07 - 11:11more successful and more comfortable?
-
11:11 - 11:14So the Health Media Collaboratory,
-
11:14 - 11:15at the end of their project, they were able
-
11:15 - 11:19to find that 87 percent of tweets
-
11:19 - 11:21about those very graphic and disturbing
-
11:21 - 11:25anti-smoking ads expressed fear,
-
11:25 - 11:27but did they conclude
-
11:27 - 11:30that they actually made people stop smoking?
-
11:30 - 11:33No. It's science, not magic.
-
11:33 - 11:36So if we are to unlock
-
11:36 - 11:39the power of data,
-
11:39 - 11:42we don't have to go blindly into
-
11:42 - 11:45Orwell's vision of a totalitarian future,
-
11:45 - 11:49or Huxley's vision of a trivial one,
-
11:49 - 11:52or some horrible cocktail of both.
-
11:52 - 11:54What we have to do
-
11:54 - 11:57is treat critical thinking with respect
-
11:57 - 11:59and be inspired by examples
-
11:59 - 12:01like the Health Media Collaboratory,
-
12:01 - 12:04and as they say in the superhero movies,
-
12:04 - 12:05let's use our powers for good.
-
12:05 - 12:08Thank you.
-
12:08 - 12:10(Applause)
- Title:
- What do we do with all this big data?
- Speaker:
- Susan Etlinger
- Description:
-
Does a set of data make you feel more comfortable? More successful? Then your interpretation of it is likely wrong. In a surprisingly moving talk, Susan Etlinger explains why, as we receive more and more data, we need to deepen our critical thinking skills. Because it's hard to move beyond counting things to really understanding them.
- Video Language:
- English
- Team:
- closed TED
- Project:
- TEDTalks
- Duration:
- 12:23
Morton Bast edited English subtitles for What do we do with all this big data? | ||
Morton Bast approved English subtitles for What do we do with all this big data? | ||
Morton Bast edited English subtitles for What do we do with all this big data? | ||
Morton Bast edited English subtitles for What do we do with all this big data? | ||
Morton Bast edited English subtitles for What do we do with all this big data? | ||
Madeleine Aronson accepted English subtitles for What do we do with all this big data? | ||
Madeleine Aronson edited English subtitles for What do we do with all this big data? | ||
Madeleine Aronson edited English subtitles for What do we do with all this big data? |