Watson, Jeopardy and me, the obsolete know-it-all

Trivia whiz Ken Jennings has made a career as a keeper of facts; he holds the longest winning streak in history on the U.S. game show Jeopardy. But in 2011, he played a challenge match against supercomputer Watson -- and lost. With humor and humility, Jennings tells us how it felt to have a computer literally beat him at his own game, and also makes the case for good old-fashioned human knowledge. (Filmed at TEDxSeattleU.)

In two weeks time, that's the ninth anniversary

of the day I first stepped out onto that hallowed "Jeopardy" set.

I mean, nine years is a long time.

And given "Jeopardy's" average demographics,

I think what that means

is most of the people who saw me on that show are now dead.

(Laughter)

But not all, a few are still alive.

Occasionally I still get recognized at the mall or whatever.

And when I do, it's as a bit of a know-it-all.

I think that ship has sailed, it's too late for me.

For better or for worse, that's what I'm going to be known as,

as the guy who knew a lot of weird stuff.

And I can't complain about this.

I feel like that was always sort of my destiny,

although I had for many years been pretty deeply in the trivia closet.

If nothing else, you realize very quickly as a teenager,

it is not a hit with girls to know Captain Kirk's middle name.

(Laughter)

And as a result, I was sort of the deeply closeted kind of know-it-all for many years.

But if you go further back, if you look at it, it's all there.

I was the kind of kid who was always bugging Mom and Dad

with whatever great fact I had just read about --

Haley's comet or giant squids

or the size of the world's biggest pumpkin pie or whatever it was.

I now have a 10-year-old of my own who's exactly the same.

And I know how deeply annoying it is, so karma does work.

(Laughter)

And I loved game shows, fascinated with game shows.

I remember crying on my first day of kindergarten back in 1979

because it had just hit me, as badly as I wanted to go to school,

that I was also going to miss "Hollywood Squares" and "Family Feud."

I was going to miss my game shows.

And later, in the mid-'80s,

when "Jeopardy" came back on the air,

I remember running home from school every day to watch the show.

It was my favorite show, even before it paid for my house.

And we lived overseas, we lived in South Korea where my dad was working,

where there was only one English language TV channel.

There was Armed Forces TV,

and if you didn't speak Korean, that's what you were watching.

So me and all my friends would run home every day and watch "Jeopardy."

I was always that kind of obsessed trivia kid.

I remember being able to play Trivial Pursuit against my parents back in the '80s

and holding my own, back when that was a fad.

There's a weird sense of mastery you get

when you know some bit of boomer trivia that Mom and Dad don't know.

You know some Beatles factoid that Dad didn't know.

And you think, ah hah, knowledge really is power --

the right fact deployed at exactly the right place.

I never had a guidance counselor

who thought this was a legitimate career path,

that thought you could major in trivia

or be a professional ex-game show contestant.

And so I sold out way too young.

I didn't try to figure out what one does with that.

I studied computers because I heard that was the thing,

and I became a computer programmer --

not an especially good one,

not an especially happy one at the time when I was first on "Jeopardy" in 2004.

But that's what I was doing.

And it made it doubly ironic -- my computer background --

a few years later, I think 2009 or so,

when I got another phone call from "Jeopardy" saying,

"It's early days yet, but IBM tells us

they want to build a supercomputer to beat you at 'Jeopardy.'

Are you up for this?"

This was the first I'd heard of it.

And of course I said yes, for several reasons.

One, because playing "Jeopardy" is a great time.

It's fun. It's the most fun you can have with your pants on.

(Laughter)

And I would do it for nothing.

I don't think they know that, luckily,

but I would go back and play for Arby's coupons.

I just love "Jeopardy," and I always have.

And second of all, because I'm a nerdy guy and this seemed like the future.

People playing computers on game shows

was the kind of thing I always imagined would happen in the future,

and now I could be on the stage with it.

I was not going to say no.

The third reason I said yes

is because I was pretty confident that I was going to win.

I had taken some artificial intelligence classes.

I knew there were no computers that could do what you need to do to win on "Jeopardy."

People don't realize how tough it is to write that kind of program

that can read a "Jeopardy" clue in a natural language like English

and understand all the double meanings, the puns, the red herrings,

unpack the meaning of the clue.

The kind of thing that a three- or four-year-old human, little kid could do,

very hard for a computer.

And I thought, well this is going to be child's play.

Yes, I will come destroy the computer and defend my species.

(Laughter)

But as the years went on,

as IBM started throwing money and manpower and processor speed at this,

I started to get occasional updates from them,

and I started to get a little more worried.

I remember a journal article about this new question answering software that had a graph.

It was a scatter chart showing performance on "Jeopardy,"

tens of thousands of dots representing "Jeopardy" champions up at the top

with their performance plotted on number of --

I was going to say questions answered, but answers questioned, I guess,

clues responded to --

versus the accuracy of those answers.

So there's a certain performance level that the computer would need to get to.

And at first, it was very low.

There was no software that could compete at this kind of arena.

But then you see the line start to go up.

And it's getting very close to what they call the winner's cloud.

And I noticed in the upper right of the scatter chart

some darker dots, some black dots, that were a different color.

And thought, what are these?

"The black dots in the upper right represent 74-time 'Jeopardy' champion Ken Jennings."

And I saw this line coming for me.

And I realized, this is it.

This is what it looks like when the future comes for you.

(Laughter)

It's not the Terminator's gun sight;

it's a little line coming closer and closer to the thing you can do,

the only thing that makes you special, the thing you're best at.

And when the game eventually happened about a year later,

it was very different than the "Jeopardy" games I'd been used to.

We were not playing in L.A. on the regular "Jeopardy" set.

Watson does not travel.

Watson's actually huge.

It's thousands of processors, a terabyte of memory,

trillions of bytes of memory.

We got to walk through his climate-controlled server room.

The only other "Jeopardy" contestant to this day I've ever been inside.

And so Watson does not travel.

You must come to it; you must make the pilgrimage.

So me and the other human player

wound up at this secret IBM research lab

in the middle of these snowy woods in Westchester County

to play the computer.

And we realized right away

that the computer had a big home court advantage.

There was a big Watson logo in the middle of the stage.

Like you're going to play the Chicago Bulls,

and there's the thing in the middle of their court.

And the crowd was full of IBM V.P.s and programmers

cheering on their little darling,

having poured millions of dollars into this

hoping against hope that the humans screw up,

and holding up "Go Watson" signs

and just applauding like pageant moms every time their little darling got one right.

I think guys had "W-A-T-S-O-N" written on their bellies in grease paint.

If you can imagine computer programmers with the letters "W-A-T-S-O-N" written on their gut,

it's an unpleasant sight.

But they were right. They were exactly right.

I don't want to spoil it, if you still have this sitting on your DVR,

but Watson won handily.

And I remember standing there behind the podium

as I could hear that little insectoid thumb clicking.

It had a robot thumb that was clicking on the buzzer.

And you could hear that little tick, tick, tick, tick.

And I remember thinking, this is it.

I felt obsolete.

I felt like a Detroit factory worker of the '80s

seeing a robot that could now do his job on the assembly line.

I felt like quiz show contestant was now the first job that had become obsolete

under this new regime of thinking computers.

And it hasn't been the last.

If you watch the news, you'll see occasionally --

and I see this all the time --

that pharmacists now, there's a machine that can fill prescriptions automatically

without actually needing a human pharmacist.

And a lot of law firms are getting rid of paralegals

because there's software that can sum up case laws and legal briefs and decisions.

You don't need human assistants for that anymore.

I read the other day about a program where you feed it a box score

from a baseball or football game

and it spits out a news article as if a human had watched the game

and was commenting on it.

And obviously these new technologies can't do as clever or creative a job

as the humans they're replacing,

but they're faster, and crucially, they're much, much cheaper.

So it makes me wonder what the economic effects of this might be.

I've read economists saying that, as a result of these new technologies,

we'll enter a new golden age of leisure

when we'll all have time for the things we really love

because all these onerous tasks will be taken over by Watson and his digital brethren.

I've heard other people say quite the opposite,

that this is yet another tier of the middle class

that's having the thing they can do taken away from them by a new technology

and that this is actually something ominous,

something that we should worry about.

I'm not an economist myself.

All I know is how it felt to be the guy put out of work.

And it was friggin' demoralizing. It was terrible.

Here's the one thing that I was ever good at,

and all it took was IBM pouring tens of millions of dollars and its smartest people

and thousands of processors working in parallel

and they could do the same thing.

They could do it a little bit faster and a little better on national TV,

and "I'm sorry, Ken. We don't need you anymore."

And it made me think, what does this mean,

if we're going to be able to start outsourcing,

not just lower unimportant brain functions.

I'm sure many of you remember a distant time

when we had to know phone numbers, when we knew our friends' phone numbers.

And suddenly there was a machine that did that,

and now we don't need to remember that anymore.

I have read that there's now actually evidence

that the hippocampus, the part of our brain that handles spacial relationships,

physically shrinks and atrophies

in people who use tools like GPS,

because we're not exercising our sense of direction anymore.

We're just obeying a little talking voice on our dashboard.

And as a result, a part of our brain that's supposed to do that kind of stuff

gets smaller and dumber.

And it made me think, what happens when computers are now better

at knowing and remembering stuff than we are?

Is all of our brain going to start to shrink and atrophy like that?

Are we as a culture going to start to value knowledge less?

As somebody who has always believed in the importance of the stuff that we know,

this was a terrifying idea to me.

The more I thought about it, I realized, no, it's still important.

The things we know are still important.

I came to believe there were two advantages

that those of us who have these things in our head have

over somebody who says, "Oh, yeah. I can Google that. Hold on a second."

There's an advantage of volume, and there's an advantage of time.

The advantage of volume, first,

just has to do with the complexity of the world nowadays.

There's so much information out there.

Being a Renaissance man or woman,

that's something that was only possible in the Renaissance.

Now it's really not possible

to be reasonably educated on every field of human endeavor.

There's just too much.

They say that the scope of human information

is now doubling every 18 months or so,

the sum total of human information.

That means between now and late 2014,

we will generate as much information, in terms of gigabytes,

as all of humanity has in all the previous millenia put together.

It's doubling every 18 months now.

This is terrifying because a lot of the big decisions we make

require the mastery of lots of different kinds of facts.

A decision like where do I go to school? What should I major in?

Who do I vote for?

Do I take this job or that one?

These are the decisions that require correct judgments

about many different kinds of facts.

If we have those facts at our mental fingertips,

we're going to be able to make informed decisions.

If, on the other hand, we need to look them all up,

we may be in trouble.

According to a National Geographic survey I just saw,

somewhere along the lines of 80 percent

of the people who vote in a U.S. presidential election about issues like foreign policy

cannot find Iraq or Afghanistan on a map.

If you can't do that first step,

are you really going to look up the other thousand facts you're going to need to know

to master your knowledge of U.S. foreign policy?

Quite probably not.

At some point you're just going to be like,

"You know what? There's too much to know. Screw it."

And you'll make a less informed decision.

The other issue is the advantage of time that you have

if you have all these things at your fingertips.

I always think of the story of a little girl named Tilly Smith.

She was a 10-year-old girl from Surrey, England

on vacation with her parents a few years ago in Phuket, Thailand.

She runs up to them on the beach one morning

and says, "Mom, Dad, we've got to get off the beach."

And they say, "What do you mean? We just got here."

And she said, "In Mr. Kearney's geography class last month,

he told us that when the tide goes out abruptly out to sea

and you see the waves churning way out there,

that's the sign of a tsunami, and you need to clear the beach."

What would you do if your 10-year-old daughter came up to you with this?

Her parents thought about it,

and they finally, to their credit, decided to believe her.

They told the lifeguard, they went back to the hotel,

and the lifeguard cleared over 100 people off the beach, luckily,

because that was the day of the Boxing Day tsunami,

the day after Christmas, 2004,

that killed thousands of people in Southeast Asia and around the Indian Ocean.

But not on that beach, not on Mai Khao Beach,

because this little girl had remembered one fact from her geography teacher a month before.

Now when facts come in handy like that --

I love that story because it shows you the power of one fact,

one remembered fact in exactly the right place at the right time --

normally something that's easier to see on game shows than in real life.

But in this case it happened in real life.

And it happens in real life all the time.

It's not always a tsunami, often it's a social situation.

It's a meeting or job interview or first date

or some relationship that gets lubricated

because two people realize they share some common piece of knowledge.

You say where you're from, and I say, "Oh, yeah."

Or your alma mater or your job,

and I know just a little something about it,

enough to get the ball rolling.

People love that shared connection that gets created

when somebody knows something about you.

It's like they took the time to get to know you before you even met.

That's often the advantage of time.

And it's not effective if you say, "Well, hold on.

You're from Fargo, North Dakota. Let me see what comes up.

Oh, yeah. Roger Maris was from Fargo."

That doesn't work. That's just annoying.

(Laughter)

The great 18th-century British theologian and thinker, friend of Dr. Johnson,

Samuel Parr once said, "It's always better to know a thing than not to know it."

And if I have lived my life by any kind of creed, it's probably that.

I have always believed that the things we know -- that knowledge is an absolute good,

that the things we have learned and carry with us in our heads

are what make us who we are,

as individuals and as a species.

I don't know if I want to live in a world where knowledge is obsolete.

I don't want to live in a world where cultural literacy has been replaced

by these little bubbles of specialty,

so that none of us know about the common associations

that used to bind our civilization together.

I don't want to be the last trivia know-it-all

sitting on a mountain somewhere,

reciting to himself the state capitals and the names of "Simpsons" episodes

and the lyrics of Abba songs.

I feel like our civilization works when this is a vast cultural heritage that we all share

and that we know without having to outsource it to our devices,

to our search engines and our smartphones.

In the movies, when computers like Watson start to think,

things don't always end well.

Those movies are never about beautiful utopias.

It's always a terminator or a matrix or an astronaut getting sucked out an airlock in "2001."

Things always go terribly wrong.

And I feel like we're sort of at the point now

where we need to make that choice of what kind of future we want to be living in.

This is a question of leadership,

because it becomes a question of who leads the future.

On the one hand, we can choose between a new golden age

where information is more universally available

than it's ever been in human history,

where we all have the answers to our questions at our fingertips.

And on the other hand,

we have the potential to be living in some gloomy dystopia

where the machines have taken over

and we've all decided it's not important what we know anymore,

that knowledge isn't valuable because it's all out there in the cloud,

and why would we ever bother learning anything new.

Those are the two choices we have. I know which future I would rather be living in.

And we can all make that choice.

We make that choice by being curious, inquisitive people who like to learn,

who don't just say, "Well, as soon as the bell has rung and the class is over,

I don't have to learn anymore,"

or "Thank goodness I have my diploma. I'm done learning for a lifetime.

I don't have to learn new things anymore."

No, every day we should be striving to learn something new.

We should have this unquenchable curiosity for the world around us.

That's where the people you see on "Jeopardy" come from.

These know-it-alls, they're not Rainman-style savants

sitting at home memorizing the phone book.

I've met a lot of them.

For the most part, they are just normal folks

who are universally interested in the world around them, curious about everything,

thirsty for this knowledge about whatever subject.

We can live in one of these two worlds.

We can live in a world where our brains, the things that we know,

continue to be the thing that makes us special,

or a world in which we've outsourced all of that to evil supercomputers from the future like Watson.

Ladies and gentlemen, the choice is yours.

Thank you very much.