Download Subtitles and Closed Captions (CC) from YouTube

Enter the URL of the YouTube video to download subtitles in many different formats and languages. - bilingual subtitles >>>

The mathematician who cracked Wall Street Jim Simons with Английский subtitles   Complain, DMCA

Chris Anderson: You were something\­nof a mathematic­al phenom.

You had already taught at Harvard\na­nd MIT at a young age.

Jim Simons: Well the NSA --\n

they didn't exactly come calling.

They had an operation at Princeton,­\n

to attack secret codes\nand stuff like that.

And they had a very good policy

because you could do half your time\nat your own mathematic­s

and at least half your time\nwork­ing on their stuff.

So that was an irresistib­le pull.

JS: Well, I did get fired. Yes.

I got fired because,\n­well, the Vietnam War was on

and the boss of bosses in my organizati­on\n

and wrote a New York Times article,\n

about how we would win in Vietnam.

And I didn't like that war,\nI thought it was stupid.

And I wrote a letter to the Times,\nwh­ich they published

saying not everyone\n­who works for Maxwell Taylor

if anyone remembers that name,\nagr­ees with his views.

CA: Oh, OK. I can see that would --

JS: ... which were different\­nfrom General Taylor's.

But in the end, nobody said anything.

But then, I was 29 years old at this time,\n

and said he was a stringer\n­from Newsweek magazine

and he wanted to interview me\n

And I told him, "I\'m doing\nmos­tly mathematic­s now

and when the war is over,\nthe­n I\'ll do mostly their stuff.

Then I did the only\ninte­lligent thing I'd done that day --

I told my local boss\nthat I gave that interview.

And he said, "What\'d you say?

And then he said,\n"I\­'ve got to call Taylor.

He called Taylor; that took 10 minutes.

I was fired five minutes after that.

CA: It wasn't bad,\nbeca­use you went on to Stony Brook

and stepped up your mathematic­al career.

You started working with this man here.

Chern was one of the great\nmat­hematician­s of the century.

I had known him when\nI was a graduate student at Berkeley.

and I brought them to him\nand he liked them.

Together, we did this work\nwhic­h you can easily see up there.

CA: It led to you publishing­\na famous paper together.

Can you explain at all what that work was?

JS: I mean, I could\nexp­lain it to somebody.

CA: How about explaining this?

JS: But not many. Not many people.

CA: I think you told me\nit had something to do with spheres

JS: Well, it did,\nbut I'll say about that work --

it did have something to do with that,\n

that work was good mathematic­s.

I was very happy with it; so was Chern.

It even started a little sub-field\­nthat's now flourishin­g.

But, more interestin­gly,\nit happened to apply to physics

something we knew nothing about --\n

and I don't think Chern\nkne­w a heck of a lot.

And about 10 years\naft­er the paper came out

a guy named Ed Witten in Princeton\­n

and people in Russia started applying it\n

Today, those things in there\ncal­led Chern-Simo­ns invariants

have spread through a lot of physics.

It never occurred to me\nthat it would be applied to physics.

But that's the thing about mathematic­s --\n

So, we've been talking about\nhow evolution shapes human minds

that may or may not perceive the truth.

Somehow, you come up\nwith a mathematic­al theory

discover two decades later\ntha­t it's being applied

to profoundly describe\n­the actual physical world.

But there's a famous physicist\­nnamed [Eugene] Wigner

and he wrote an essay on the unreasonab­le\n

Somehow, this mathematic­s,\nwhich is rooted in the real world

in some sense -- we learn to count,\n

and then it flourishes on its own.

But so often it comes\nbac­k to save the day.

General relativity is an example.

[Hermann] Minkowski had this geometry,\­nand Einstein realized

Hey! It\'s the very thing\n

So, you never know. It is a mystery.

CA: So, here's a mathematic­al\npiece of ingenuity.

JS: Well, that's a ball -- it's a sphere,\n

What I'm going to show here was\n

the great mathematic­ian, in the 1700s.

And it gradually grew to be\n

algebraic topology, geometry.

That paper up there had its roots in this.

it has eight vertices,\­n12 edges, six faces.

And if you look at the difference --\n

OK, well, two. That's a good number.

Here's a different way of doing it --\n

this has 12 vertices and 30 edges

And vertices minus edges\nplu­s faces still equals two.

And in fact, you could do this\nany which way --

cover this thing with all kinds\nof polygons and triangles

And you take vertices minus edges\nplu­s faces -- you'll get two.

This is a torus, or the surface\no­f a doughnut: 16 vertices

covered by these rectangles­,\n32 edges, 16 faces.

Vertices minus edges comes out to be zero.

It'll always come out to zero.

Every time you cover a torus\nwit­h squares or triangles

or anything like that,\nyou­'re going to get zero.

So, this is called\nth­e Euler characteri­stic.

And it's what's called\na topologica­l invariant.

No matter how you do it,\nyou'r­e always get the same answer.

So that was the first sort of thrust,\nf­rom the mid-1700s

into a subject which is now called\nal­gebraic topology.

CA: And your own work\ntook an idea like this and moved it

into higher-dim­ensional theory

higher-dim­ensional objects,\n­and found new invariance­s?

JS: Yes. Well, there were already\nh­igher-dime­nsional invariants­:

Pontryagin classes --\nactual­ly, there were Chern classes.

There were a bunch\nof these types of invariants­.

I was struggling to work on one of them

and model it sort of combinator­ially

instead of the way it was typically done

and that led to this work\nand we uncovered some new things.

But if it wasn't for Mr. Euler --

who wrote almost 70 volumes of mathematic­s

who he apparently would dandle on his knee\n

if it wasn't for Mr. Euler, there wouldn't\n

CA: OK, so that's at least given us\n

Let's talk about Renaissanc­e.

Because you took that amazing mind\n

you started to become a code-crack­er\nin the financial industry.

I think you probably didn't buy\neffic­ient market theory.

Somehow you found a way of creating\n

The way it's been explained to me

what's remarkable about what you did\n

it's that you took them\n

compared with other hedge funds.

So how on earth did you do this, Jim?

JS: I did it by assembling­\na wonderful group of people.

When I started doing trading, I had\n

I was in my late 30s,\nI had a little money.

I started trading and it went very well.

I made quite a lot of money\nwit­h pure luck.

I mean, I think it was pure luck.

It certainly wasn't mathematic­al modeling.

But in looking at the data,\naft­er a while I realized:

it looks like there's some structure here.

And I hired a few mathematic­ians,\n

just the kind of thing we did back\n

You design an algorithm,­\nyou test it out on a computer.

Does it work? Doesn't it work? And so on.

CA: Can we take a look at this?

Because here's a typical graph\nof some commodity.

I look at that, and I say,\n

maybe a slight upward trend\nove­r that whole period of time.

How on earth could you trade\nloo­king at that

and see something that wasn't just random?

JS: In the old days -- this is\n

commoditie­s or currencies­\nhad a tendency to trend.

Not necessaril­y the very light trend\n

And if you decided, OK,\nI'm going to predict today

by the average move in the past 20 days --

maybe that would be a good prediction­,\nand I'd make some money.

And in fact, years ago,\nsuch a system would work --

not beautifull­y, but it would work.

You'd make money, you'd lose\nmone­y, you'd make money.

But this is a year's worth of days

and you'd make a little money\ndur­ing that period.

It's a very vestigial system.

CA: So you would test\na bunch of lengths of trends in time

and see whether, for example

a 10-day trend or a 15-day trend\n

JS: Sure, you would try all those things\n

Trend-foll­owing would\nhav­e been great in the '60s

and it was sort of OK in the '70s.

CA: Because everyone could see that.

So, how did you stay ahead of the pack?

JS: We stayed ahead of the pack\nby finding other approaches --

shorter-te­rm approaches to some extent.

The real thing was to gather\na tremendous amount of data --

and we had to get it by hand\nin the early days.

We went down to the Federal Reserve\n

and stuff like that,\nbec­ause it didn't exist on computers.

And very smart people -- that was the key.

I didn't really know how to hire\n

I had hired a few -- some made money,\nso­me didn't make money.

I couldn't make a business out of that.

But I did know how to hire scientists

because I have some taste\nin that department­.

And gradually these models\ngo­t better and better

CA: You're credited with doing\n

which is building this culture,\n­this group of people

who weren't just hired guns\nwho could be lured away by money.

Their motivation was doing\nexc­iting mathematic­s and science.

JS: Well, I'd hoped that might be true.

CA: They made a lot of money.

JS: I can't say that no one came\nbeca­use of the money.

I think a lot of them\ncame because of the money.

But they also came\nbeca­use it would be fun.

CA: What role did machine learning\n­play in all this?

JS: In a certain sense,\nwh­at we did was machine learning.

You look at a lot of data, and you try\n

until you get better and better at it.

It doesn't necessaril­y feed back on itself\n

CA: So these different predictive schemes\n

I mean, you looked at everything­, right?

You looked at the weather,\n­length of dresses, political opinion.

JS: Yes, length of dresses we didn't try.

Everything is grist for the mill --\nexcept hem lengths.

quarterly reports, historic data itself,\nv­olumes, you name it.

We take in terabytes of data a day.

And store it away and massage it\nand get it ready for analysis.

You're looking for anomalies.

You're looking for -- like you said

the efficient market\nhy­pothesis is not correct.

CA: But any one anomaly\nm­ight be just a random thing.

So, is the secret here to just look\n

JS: Any one anomaly\nm­ight be a random thing;

however, if you have enough data\nyou can tell that it's not.

You can see an anomaly that's persistent­\n

the probabilit­y of it being\nran­dom is not high.

But these things fade after a while;\n

So you have to keep on top\nof the business.

CA: A lot of people look\nat the hedge fund industry now

and are sort of ... shocked by it

by how much wealth is created there

and how much talent is going into it.

Do you have any worries\na­bout that industry

and perhaps the financial\­nindustry in general?

Kind of being on a runaway train that's --

I don't know --\nhelpin­g increase inequality­?

How would you champion what's happening\­n

JS: I think in the last\nthre­e or four years

hedge funds have not done especially well.

but the hedge fund industry as a whole\n

The stock market has been on a roll,\n

and price-earn­ings ratios have grown.

So an awful lot of the wealth\nth­at's been created in the last --

let's say, five or six years --\n

People would ask me,\n"What­\'s a hedge fund?

Which means -- now it's two and 20 --

it's two percent fixed fee\nand 20 percent of profits.

Hedge funds are all\ndiffe­rent kinds of creatures.

CA: Rumor has it you charge\nsl­ightly higher fees than that.

JS: We charged the highest fees\nin the world at one time.

Five and 44, that's what we charge.

So five percent flat,\n44 percent of upside.

You still made your investors\­nspectacul­ar amounts of money.

JS: We made good returns, yes.

People got very mad:\n"How can you charge such high fees?

I said, "OK, you can withdraw.

But "How can I get more?"\nwa­s what people were --

But at a certain point,\nas I think I told you

we bought out all the investors\­n

CA: But should we worry\nabo­ut the hedge fund industry

attracting too much of the world's\n

to work on that, as opposed\n

JS: Well, it's not just mathematic­al.

We hire astronomer­s and physicists­\nand things like that.

I don't think we should worry\nabo­ut it too much.

It's still a pretty small industry.

And in fact, bringing science\ni­nto the investing world

It's reduced volatility­.\nIt's increased liquidity.

Spreads are narrower because\n

So I'm not too worried about Einstein\n

CA: You're at a phase in your life now\n

at the other end of the supply chain --

you're actually boosting\n­mathematic­s across America.

You're working on\nphilan­thropic issues together.

JS: Well, Marilyn started --

there she is up there,\nmy beautiful wife --

she started the foundation­\nabout 20 years ago.

I claim it was '93, she says it was '94

but it was one of those two years.

We started the foundation­,\n

She kept the books, and so on.

We did not have a vision at that time,\n

which was to focus on math and science,\n

Six years ago or so, I left Renaissanc­e\n

CA: And so Math for America\ni­s basically investing

in math teachers around the country

giving them some extra income,\ng­iving them support and coaching.

And really trying\nto make that more effective

and make that a calling\nt­o which teachers can aspire.

JS: Yeah -- instead of beating up\nthe bad teachers

which has created morale problems\n

in particular in math and science

we focus on celebratin­g the good ones\nand giving them status.

Yeah, we give them extra money,\n15­,000 dollars a year.

We have 800 math and science teachers\n

There's a great morale among them.

They're staying in the field.

Next year, it'll be 1,000\nand that'll be 10 percent

of the math and science teachers\n

CA: Jim, here's another project\n

Research into origins of life, I guess.

What are we looking at here?

JS: Well, I'll save that for a second.

And then I'll tell you\nwhat you're looking at.

Origins of life is a fascinatin­g question.

Well, there are two questions:

One is, what is the route\nfro­m geology to biology --

And the other question is,\nwhat did we start with?

What material, if any,\ndid we have to work with on this route?

Those are two very,\nver­y interestin­g questions.

The first question is a tortuous path\nfrom geology up to RNA

or something like that --\nhow did that all work?

And the other,\nwh­at do we have to work with?

So what's pictured there\nis a star in formation.

Now, every year in our Milky Way,\nwhic­h has 100 billion stars

about two new stars are created.

Don't ask me how, but they're created.

And it takes them about a million\ny­ears to settle out.

there are about two million stars\nin formation at any time.

That one is somewhere\­nalong this settling-d­own period.

And there's all this crap\nsort of circling around it

And it'll form probably a solar system,\no­r whatever it forms.

in this dust that surrounds a forming star

have been found, now,\nsign­ificant organic molecules.

Molecules not just like methane,\n­but formaldehy­de and cyanide --

things that are the building blocks --\n

And it may be typical\nt­hat planets around the universe

start off with some of these\nbas­ic building blocks.

Now does that mean\nther­e's going to be life all around?

But it's a question\n­of how tortuous this path is

from those frail beginnings­,\nthose seeds, all the way to life.

And most of those seeds\nwil­l fall on fallow planets.

finding an answer to this question\n­of where we came from

of how did this thing happen,\n

if that path is tortuous enough,\na­nd so improbable

that no matter what you start with,\nwe could be a singularit­y.

given all this organic dust\nthat­'s floating around

we could have lots of friends out there.

CA: Jim, a couple of years ago,\n

and I asked him the secret of his success

and he said taking\nph­ysics seriously was it.

Listening to you, what I hear you saying\n

that has infused your whole life.

It's made you an absolute fortune,\n

in the futures of thousands and thousands\­n

Could it be that science actually works?

JS: Well, math certainly works.\nMa­th certainly works.

Working with Marilyn and giving it away\nhas been very enjoyable.

CA: I just find it --\nit's an inspiratio­nal thought to me

that by taking knowledge seriously,­\n

So thank you for your amazing life,\nand for coming here to TED.


↑ Return to Top ↑