Secrets of the Universe (2022–…): Season 1, Episode 7 - Chasing Black Holes - full transcript
Secrets of the Universe -
Chasing Black Holes
You had to admire people that would devote years,
and years, and years to trying to do it
with the likelihood being that they would fail.
More than two decades ago,
two teams decided to try what seemed impossible,
to prove the existence of black holes.
Both of these kind of converged at almost the same time,
but coming about it in very different ways.
We're crazy optimists in this business.
You have to be.
If you look at all the things that could go wrong,
then why would you dedicate your career to it?
One team called LIGO
plan to use giant instruments
called laser interferometers
to detect tiny ripples in space-time
called gravitational waves
caused by the merger of two black holes.
Albert Einstein thought that it will be impossible
to ever detect them.
We're looking for motions
that are about 1,000th of the size
of a proton inside the nucleus of an atom.
The other team
called the Event Horizon Telescope
would try to capture an actual image of a black hole,
a task comparable to seeing a tennis ball on the moon.
And so you need a telescope the size of the entire earth.
But obviously,
we couldn't build a telescope the size of the the earth.
But in radio astronomy,
you can play this near magical trick
where you take two telescopes separated by some distance,
bring the data that they receive together
to form a telescope
as though you had one as large as the distance
between these telescopes.
Finding black holes,
either by taking an image of one
or by detecting them with gravitational waves
would be among the most difficult challenges
in the history of science.
I was sort of trembling in front of the challenge.
It was hopeless
because the technology would never be there.
We staged experiments that went nowhere.
We tried and we failed.
And if they found what they were looking for,
could they keep it secret until they were sure?
Many people had a eureka moment.
I was completely blown away
by how clear cut this signal was.
And I had a moment of panic.
We all kept the results secret.
Because I knew if that was right,
we would be writing history.
All I could do is figure out how we had gone wrong,
how are we fooling ourselves?
At some point,
you have to stop and just listen to the universe.
Black holes were born of theory.
They might really exist,
but there was no direct proof.
Many felt that the idea itself must be fatally flawed.
An extremely massive object
will curve the space-time around it so much
that anything that passes by
will actually fall into that object
and will not be able to escape.
Even light wouldn't be able to escape,
and that's basically a black hole.
But is this real?
Clearly, something like this would not occur in nature.
No one felt that nature would be that crazy.
These cosmic objects are more extraordinary
than we could ever have dreamt up.
At the center of a black hole,
the laws of physics as we know today,
they break down.
There's no matter there.
The structure that we see is a vacuum structure.
It's like a tornado in space-time.
In their purest form,
the concept of black holes starts with Albert Einstein.
For physicists,
the great year for Einstein was 1905
when he solved several of the biggest problems in physics
within a matter of months.
And then he spent 10 years developing his theory of gravity,
which is general relativity.
What we see coming out of general relativity
is that matter and energy has the ability
to affect the space time around it.
Space time is a mathematical concept
that unites time with three dimensions of physical space
so that they are intimately woven together
into what is called the fabric of space time.
The actual presence of matter distorts this fabric.
If you send a light beam past a, say, a star or a planet,
the reason it bends is
because it's trying to follow the straightest path possible
in a geometry that is now curved.
And thus, that is what gravity is.
It's basically the geometry of space time.
Within weeks of Einstein
publishing his revolutionary theory of gravity,
people began trying to figure out what it all meant.
One of the first
was a German physicist called Karl Schwarzschild.
Despite being in his forties,
Karl decided to volunteer to fight in the First World War
and he worked as an artillery man.
He was able to predict how space
and time would look like around a point mass.
And he realizes if you compact matter
into a small enough volume,
that there is this event horizon
at what we can now call the Schwarzschild radius
where the speed of escape is larger than the speed of light,
where even light cannot make it out.
A one way boundary
where information and light can only go in,
but never can come out again.
That's the moment when black holes really were born.
Of course, completely crazy idea.
This wouldn't exist.
Neither man realized that it was anything more
than a mathematical construct
that they didn't really have to worry about
because no one felt that nature would allow it to happen.
Poor Karl Schwarzschild,
he died very quickly after finding the solutions
due to an illness that he got in the trenches there.
Throughout the 20th century,
the theoretical study of black holes became vibrant
and popular among scientists.
They had theories for how they might form.
And astronomers had even started to speculate
about where they might lurk in the cosmos.
So there are two varieties of black holes,
the stellar variety,
which are born during the death of stars.
And they weigh a few or 10 times what our sun does.
Stellar mass black holes
are thought to have a Schwarzschild radius
of around the size of a small modern city.
Even though they have a mass of 10 or 20 suns,
they are tiny in comparison to the stars
from which they were born.
Super massive black holes, however,
would contain the combined mass of millions
or even billions of suns.
They could have a Schwarzschild radius
the size of an entire solar system.
Astronomers wondered if stellar black holes
were hiding in all galaxies
and that super massive black holes
might be found at the heart of many.
They even named the super massive black hole
that they suspected might lie at the heart of our galaxy
Sagittarius A star.
Are they out there?
The theory predicted it, but are they really real?
To try and answer that question,
the team called LIGO
would try to find the smaller black holes
with gravitational waves,
while the Event Horizon team
would hunt the super massive black holes
using a type of astronomy that uses radio waves.
If you have optical light,
the wavelengths are tiny, tiny, tiny.
But if you study radio emission,
the wavelengths are actually big.
And so you have a gigantic dish
that can reflect radio waves.
It was intense radio emissions
coming from the center of galaxies
that first made astronomers suspect
that super massive black holes must be the culprit.
So we had to think a little bit
about these emission mechanisms,
like what would cause is glow in radio waves?
And so what can power something like that
is gravitational accretion onto a super massive black hole.
That's the only way we know of to power something like this.
In the 20th century,
the idea of capturing an image of a black hole
with these radio emissions
was far beyond the realms of possibility,
but some people did dare to dream.
Many times in science,
you wind up in a situation
where somebody makes an initial discovery
and it goes under appreciated.
And this is the case with Jean-Pierre Luminet.
I mean, this was a real visionary.
In 1979, he came with the first full simulation
of what a black hole would look like if you were there,
like if you could really see in infinite detail.
To capture an image of a real black hole
in the way that John-Pierre Luminet imagined
would require a radio telescope
with a dish as big as the planet earth
and technologies that simply didn't exist yet.
However, for the gravitational wave team,
they had an even more fundamental problem to solve.
Nobody even knew if gravitational waves were real at all.
The theory behind them originated with Albert Einstein
from around the same time
that Karl Schwarzschild was formulating his ideas
of black holes.
In 1916,
Einstein wrote a little paper, five pages long maybe.
He looked at the equations of general relativity
that he had developed,
and he noticed it had a great similarity
if you put it in a certain way
to the equations of electricity and magnetism.
And since electricity and magnetism have waves,
he conjectured that gravity must also have waves.
By the 1930s,
Einstein had written a further two papers
on the subject of gravitational waves.
According to Einstein,
this phenomenon of gravitational waves
is very interesting from a theoretical point of view,
but any real effect on earth will be so small
that it will be very likely
always impossible to ever detect them.
They were never fully accepted
during this whole period and beyond,
even until his death.
The great Albert Einstein died in 1955.
Black holes and gravitational waves,
ideas that were rooted in his theories,
were far from being considered a reality.
But in 1958,
50 of the world's greatest experts
in general relativity had a meeting.
So at this now renowned conference,
the question of whether gravitational waves really do exist
and produce something that could be measured was raised.
Richard Feynman was there.
He said, if gravitational waves really exist,
they have to be able to do something.
They can't just exist.
They have to be able to transfer energy.
And so he made a,
what we call like a experiment,
just a thought experiment.
And that was if you have a bar
and you put a couple of rings on it,
and then a gravitational wave goes through the bar,
it'll take the bar,
and it'll expand it, and contract it,
and expand it, and contract it
at the frequency of the gravitational wave.
As it expands, it, of course, pushes on the rings
and they would move.
But what's happening
is that they're transferring energy friction to the ring.
This clever thought experiment
inspired one of the people at the meeting
to actually build such a device.
His name was Joseph Weber
and the machines he created are called Weber bars,
the first ever gravitational wave detectors.
It's accepted today that his bar detectors
were not sensitive enough to be able to make any detections,
but he played a very important role
in sparking the rest of the world's interest
into what a different type of detector might look like
that could ultimately be more sensitive.
So gravitational waves, if they exist,
how do they manifest themselves?
If I take just a place
and a gravitational wave comes through,
it distorts the space and time
such that it stretches it in one direction
and squashes it in the other direction.
The easiest way to think about it
is what happens when you go to the amusement park
and you see these mirrors.
You look in one and you get tall and thin.
You go to the next one, you get short and fat.
If you imagine that you're now a detector,
it's basically going back and forth
between those two mirrors.
You're getting taller and shorter
and fatter at the frequency of the gravitational waves.
So maybe the gravitational wave is 60 Hertz,
so 60 times a second, you're going.
And that's what we have to measure.
So how are we gonna measure that?
So we had to find out a way
to be able to detect here on earth
some very, very small perturbations.
And it turned out that the most sensitive instrument
that we could make and design
is what is called a laser interferometer.
So what does an interferometer do?
I take a beam of light
and I get to a place where there's a mirror
that sends half the light in one direction
and half the light in the perpendicular direction.
If somewhere down the way, I put a mirror,
it'll bounce back.
And if I've calibrated it so that they're the same length,
they'll come back at exactly the same time.
If I invert the signal from one to the other,
they'll have exactly canceled.
I have a little photo detector.
It sees nothing, okay?
But now imagine the gravitational wave
went through this same thing.
And 60 times a second,
one arm's gonna get a little longer than the other.
The light will come back at a slightly different time
and the light will go 60 times a second,
depending on how strong the gravitational wave was.
And that's all we have.
We have a detector
that's measuring the length of these two arms.
That's called interferometry.
Two of the people
who believe that interferometry might be the solution
for finding gravitational waves,
and therefore, for finding black holes
were Rainer Weiss from MIT and Kip Thorne from Caltech.
I started on the faculty at Caltech
as a particle physicist at the same time
Kip Thorne started as a general relativist,
and we were close friends.
By the 1980s,
several groups from around the world
would start to consider building interferometers
to search for gravitational waves.
The British and Germans formed a collaboration.
The Italians and French
started a collaboration called Virgo.
And in the U.S.,
Rainer Weiss, Ron Drever, and Kip Thorne formed LIGO.
But getting funding to build such a venture
was not going to be easy.
What was pointed out in the U.S. Congress
about whether to appropriate funds
was just how extraordinarily sensitive
these detectors would have to be.
You would need to be able to measure these mirrors
moving at distances
that were some 10,000 times smaller
than the diameter of an atom.
And when thinking about that in an astronomical terms,
that's maybe, say, the same as being able
to measure the distance to the nearest star
and being able to say definitively
whether that distance has changed
by the width of a human hair.
And so these mind-boggling ideas
were used as a way to cast doubt
on the plausibility of such an experiment.
In 1990,
the National Science Foundation
approved the construction of two detectors,
Livingston, Louisiana, and Hanford, Washington State.
But LIGO was going to need a lot of money
and firm leadership
if it was to stand a chance of finding stellar black holes
with gravitational waves.
The Event Horizon Telescope team
that would try to take an image
of a super massive black hole
would not officially form for many more years,
though the ideas upon which it would be based
were beginning to emerge.
When I was in graduate school,
I had the great fortune to work with Dr. Alan Rogers,
who was one of the pioneers of radio astronomy.
And he got me hooked on using interferometry
to make the most detailed images of the sky
that we could make at the time.
And we started doing this with Sagittarius A star,
which we thought was a super massive black hole
in the center of our Milky Way galaxy.
So there is a technique out there
that helps us with increasing the resolution
that we can get from a telescope,
which is not to use a single one,
but to use pairs of telescopes
and look at the distant object at the same time
with these pairs of telescopes,
record the incoming light.
And because we time tag exactly when each signal arrived,
we can after the fact combine these signals
using super computers,
just match them up
and pretend we had the resolution of a telescope
that's as big as the separation between the two.
We call it very long baseline interferometry
and the key thing here is that we interfere the signals
that we receive on different parts of the earth.
To use very long baseline interferometry
or VLBI to actually capture an image
of a super massive black hole
was going to require a radical leap forward
in our understanding of how matter
spins around a black hole that is feeding,
the so-called accretion disk.
I think any revolution requires a group of people
sorting out the ideas that they have,
figuring out what's the best way to move forward.
I met Dimitrios when we were both at Harvard.
In fact, our first paper together is on Sagittarius A star,
the emission from it,
and whether the accretion flow around the black hole
would allow us to see down to the horizon.
It was the time
where we had just started thinking differently
about how do black holes accumulate matter
from their surroundings.
For about 20 to 30 years, there was a paradigm,
but then it became obvious sometime in the '90s
that that paradigm was not right.
So it was that small group
that started building a new paradigm.
We were doing the theoretically, you know?
We were thinking about,
where does this radio emission come from in black holes?
Some people proposed that this would come
from the matter of that shoots out from the plasma jets,
and other colleagues were calculating
that the radio emission would come from the matter
falling into the black hole.
We had group meetings that lasted eight hours.
We just had coffee,
we just had the white board and we were trying out ideas.
And most of them were wrong,
most of them failed,
but a few of them turned out to be really important.
One of those questions that we wanted to ask is,
what would the black hole look like
if I were to take a picture?
Will it be a big fuzzy, fluff of cloud,
or will it be something really small,
as small as the Event Horizon of the black hole?
So we spent about 10 years learning how to tweak the theory
and make predictions of black holes
that were not the same
as the ones that Albert Einstein predicted.
And lo and behold,
we always saw this dark shadow in the very center.
Irrespectively of how the matter was rotating,
whether it was falling in,
whether it was flowing out,
whether the black hole was rotating,
as long as the mass was known as a black hole,
the shadow had to be there.
So in year 2000,
I realized that at certain wavelengths,
the entire accretion flow is actually transparent
and allows us to see all the way down
to the horizon of the black hole,
which is what we're after, that shadow.
How would that look like?
Could we actually see that?
And it turns out that the answer depended on
how did I take the picture,
what wavelengths of light did I use?
This new multinational community
of radio astronomers and theorists realized
that the key was to push very long baseline interferometry
to see higher frequencies of light.
The higher frequency you observe,
the finer your angular resolution.
Your pictures get much crisper
and you can see the details that you want to see.
But the second thing is
that as you move higher in frequency,
you can see more deeply into all the hot gas.
So you really want to see all the way to the event horizon.
You want to get sharper images,
and that pushed us to higher frequencies.
The target that everyone was focusing on
was the suspected super massive black hole
at the center of our galaxy, Sagittarius A star.
I was convinced we would be able to see
the black hole at the center of our Milky Way
with a global telescope array,
but that required a lot of money,
new telescopes, new receivers,
and it required a big community to work together.
There's a transition
that many areas of science go through
in terms of moving from things in the laboratory,
taking those ideas and concepts and scaling them up
to turn those into a large scale project,
it's a skill in itself.
Barry Barish
had been working in particle physics
at the superconducting supercollider in Texas.
The U.S. Congress canceled the supercollider
at a time when LIGO needed somebody
that was capable of taking it
to get both the funding that was needed
and be able to put it together
in a way that would make it work.
In 1994, the National Science Foundation
made Barish the laboratory director of LIGO.
About six months later,
went to the National Science Board
that oversees the National Science Foundation,
and convinced them, and they funded us.
I think the real hero is not me, it's the NSF.
For me, the big issue was to get a strong enough team,
as good as these two institutions are,
Caltech and MIT.
For our problem as hard as gravitational waves,
we needed to tap the best people in the world.
The underlying technologies
that are relevant for gravitational wave detection
were being developed in different places.
In 1997,
Barish established the LIGO Scientific Collaboration,
which merged several international groups into LIGO.
This meant there were now two major multinational groups
vying to build detectors,
LIGO and Virgo.
The group that would become the Event Horizon Telescope
were also making progress.
So we all started to work together
at three millimeter wavelengths,
and then we realized that to push it even further,
we had to go to one millimeter wavelength.
And that's when we began this little race,
a competitive but also a collegial race.
Could we push this technique to its real limits?
And that was great time.
And then we ran into this roadblock.
For many years, we were stymied.
We were just stuck
because we didn't have the sensitivity we needed
to make these observations at high frequencies.
As soon as you go to high frequencies,
everything becomes harder.
The atmosphere reduces the signal coming from black holes.
The superconducting cameras
that we mount on each of our telescopes
to receive the radio waves from the black hole,
they become more noisy.
So everything is working against you.
The one thing that we had,
our secret weapon was that we could increase the bandwidth.
Starting at around 2000 and going til about 2006,
there was this explosion in capability
that happened because we started building our instruments
out of commercial electronics.
Imagine that.
Up until that point,
we had been developing specialized instrumentation
that took a decade to design,
and manufacturer, and get into the field
because it was so exquisitely specialized.
Graphics processing units or GPUs
are specialized computer chips
that are used primarily for video graphics.
We weren't pushing the development of GPUs.
In fact, the gaming industry was,
but we thought, hey,
we can solve Einstein's equations on these things
and we can do it much faster than with traditional CPUs.
The same was true for data storage.
Consumer hard drives were becoming ever faster
and greater in capacity.
All of a sudden, we could go to the store,
buy components, hook them together,
and we could make something that was 10 times more capable,
10 times lower cost,
and we could design it 10 times faster.
It was nothing short of a miracle.
The LIGO detectors had to be
the most sensitive scientific instruments in the world.
And it's all because our planet is such a noisy place.
Much of the good work happens at night,
and that's simply because the environment around you
is quieter at night.
There are fewer cars driving on the road,
hitting bumps and causing the ground to shake,
people, falling trees that are falling down in the forest.
The operation of heavy machinery nearby,
it's also lower at night.
There's also natural effects.
Of course, earthquakes are an obvious consideration
that might jump into many people's minds.
Believe it or not,
even if you're in the center of a continent,
there is a peak of motion that occurs at very low frequency
due to waves beating on the ocean shore.
The reason that all of these sorts of things matter
is because ultimately,
these mirrors that we're trying to detect
very small motions of are connected to the ground.
The earth ground is shaking all the time
by about a millionth of a meter,
which is a million of a million times more
than what we're trying to measure.
So this is one reason that our system
can not just sit on the ground,
but our mirrors are the most quiet place on earth.
The beam splitter itself is a mirror,
a big piece of glass suspended on wires.
That has to be isolated from external disturbances,
so it's housed inside a vacuum tank.
That vacuum tank is inside a big building.
Out from that building go two arms.
Inside that arch, there's a vacuum pipe.
Inside that travel the laser beams
that travel the length of these four kilometer arms.
The ends of the arms inside those buildings,
there are vacuum tanks.
Inside those tanks are some isolation systems
that isolate from ground vibrations,
mirrors that are suspended in the initial incarnation
of LIGO on metal wires.
But it's not just vibrations and noise
that LIGO has to fight against,
but barely imaginable quantum effects.
To measure precisely how the two waves get together,
you have to measure how many photons hit your detector.
But since the number of photons
is what we call a quantum variable,
there's an intrinsic uncertainty there.
We cannot do better
than what the Heisenberg uncertainty principle dictate.
That's one reason we need very powerful lasers.
As we increase the laser power,
the force exerted by the light on the mirrors
also increases proportionally.
So that becomes one of the big challenges
of increasing the laser power.
Every time a photon bounces off our suspended mirror,
it transfers some momentum to the mirror.
It gives it a kick,
If it were to strike it off center,
it's going to actually create a torque.
The light causes the mirrors to twist.
In an ideal situation,
you would have the light striking the center of the mirror.
So to be able to understand more about black holes
and big stuff out there in the universe,
we need to understand very well
the physics of the quantum mechanics
and the thermal motion of the atom,
which is very small scale physics.
In 2006,
we fielded these new electronic systems for the first time,
ad we took them to two sites,
one in Arizona, Mount Graham,
and one to Mauna Kea in Hawaii.
We set up this experiment,
but we were really flying a bit by the seat of our pants.
So back at the core later,
we played these data streams from Hawaii
and Arizona back again, and again, and again.
And we searched for months.
And it was a heartbreaker.
After months, we threw in the towel.
We realized that we were just not gonna be able
to detect Sagittarius A star.
Later, we found out why.
A small little chip of metal
had fallen into the heart of the superconducting junction
in the receiver of the Caltech Submillimeter Observatory.
We were able to go back the next year,
adding a new site in California.
And this time, we were successful.
We got the detections.
And that showed us immediately
that we were seeing horizon scale structure.
That's the moment we knew
that we could make an image of a black hole.
The LIGO team was sure they had detectors
that were working well,
but if black holes were spiraling into one another
somewhere in the universe,
they were not hearing it.
You don't know that you don't see something
until you look for a while.
And so it would run for six months or a year,
and wouldn't see any events.
We'd turn off, lick our wounds,
put in some improvements, and then run again.
But I think that is also maybe part of the excitement.
So there was years, and years, and years
of continually improving the detectors,
collecting data, improving them,
collecting data on no detection, no detection, no detection.
There were times where myself or the community thought
that maybe we got ourself into something too big.
The initial version of LIGO wasn't good enough
to see gravitational waves,
it turns out.
In 2010,
the detectives were scheduled to shut down
in preparation for a big upgrade called advanced LIGO.
The team trying to make an image of a black hole realized
that the friendly competition
between several European and U.S. institutions
wasn't going to be enough.
We would need many more dishes.
We would need the European dishes,
we need the American dishes.
So it was clear at some point,
those different efforts had to come together
and merge into a global collaboration.
And so there was a meeting in 2009 in California,
and I was sitting there at a coffee break together
with Shep Doeleman and was saying,
if we want to get this funded later,
we can't just keep talking about
submillimeter VLBI array, blah, blah, blah.
Nobody will understand what it is.
We have to give it a flashy name.
What about if we call it Event Horizon array
or something like this?
And in the end, we came down to Event Horizon Telescope,
and that's how it started.
The Event Horizon Telescope was finally
and officially born.
Meanwhile, LIGO was in the middle of its upgrade
from initial LIGO to advanced.
We wanted to build a very sensitive instrument.
And that meant that we had to take some risks, if you want.
Much of the effort
that went into the advanced LIGO upgrade
involved the mirrors
and how they were isolated from the outside world.
So instead of suspending these mirrors on metal wires,
we were actually suspending them on glass fibers,
fused silica fibers.
Those fibers, they're very strong if you pull them,
but if a grain of dust hit them, they shatter.
Perhaps the greatest secret weapon
that the LIGO team employed
to isolate their detectives from unwanted vibrations
was the same technology found
in noise cancellation headphones.
You put on these earphones on an airplane
and the roar of the engines goes away,
and you still hear the stewardess ask you,
do you want coffee?
So what it does is measure the ambient noise of the engines
and cancel it.
But the stewardess talking to you,
it's not ambient, it's a signal,
and so you hear that fine.
So the idea was to bury inside of these shock absorbers
little seismic sensors
that measured any residual motion of the earth.
And then we just pushed back against it,
make little actuators that push
to cancel the residual motion that's there
after the shock absorbers.
And that gained us a factor of 10 in sensitivity.
The Event Horizon Telescope group
had begun to consider trying to image
another super massive black hole,
as well as Sagittarius S star.
There was another black hole out there,
which was 1,000 times further way,
but also 1,000 times more massive.
M87 is a galaxy
about 53 million light years from earth.
And just like the Milky Way,
we suspect it has a super massive black hole at its center,
only one that is much bigger.
M87 is so massive
that it doesn't change during the course of an evening,
whereas Sagittarius A star is speedier.
And during the course of a night,
it changes its appearance.
The advanced LIGO detectors
were now successfully upgraded
and showing much higher sensitivity
than anything achieved before.
When we start a data run each year when we do this,
there's a little period of time
when people who are expert on particular things
can still decide to make some changes of settings,
so we call it an engineering run.
It happened during that period.
You can imagine I can remember very well that day.
So it was a Monday.
And I remember that
because that was the day after I ran my first marathon,
so I was thinking that there was enough excitement
for a while, right?
I remember that quite vividly in my office in Glasgow
when our colleague suddenly said,
You do know the signal had just arrived?"
And we were like, "No?"
My colleague came to tell me
that it seems there's been a detection.
I just brushed this comment off.
To me, it seemed,
like we're not even in the observation run yet.
It's too early for this to actually happen.
When that event happened,
it happened seven milliseconds earlier in Louisiana
than in Hanford.
And my first thought was, this can't be real.
This too good to be real.
And you have to imagine
for people who have spent decades measuring noise,
and being very good at mentioning noise,
to have a signal arrive and it be large,
it really did take people by surprise.
The detection made
on the 14th of September, 2015
was calculated to have been created
by the merging of two stellar mass black holes
1.3 billion light years from earth.
We were expecting to have to fight our way
through a lot of justification to convince people
that we really had to think of additional ways,
but that was exactly what you would expect
out of a textbook.
So the first thing I thought is, this is not real.
Somebody did this,
because at the time,
we had a program which was called blind injections
to test if you were actually able
to detect gravitational waves.
Some people, without telling anybody,
would add on purpose fake signals.
So all I could think is that, great,
this is an artificial signal.
But we got word from whoever was in charge
of this blind injection saying, no,
we didn't have time to set up our systems.
Sorry, we're late.
So this is not a blind injection.
But that was not the end of the story.
That was the beginning of maybe six months of very hard work
to try and prove that we didn't do anything wrong.
Because you don't want to cry wolf
the first time that you detect a gravitational wave
'cause that had happened in the history in the past.
So we don't want to do that
because then you lose credibility, of course.
I think for many people, justifiably, after many years,
they had a eureka moment.
And I had a moment of panic.
All I could do is figure out how we had gone wrong.
There then followed a frantic
and intensely busy period of forensically analyzing
whether they believed their own data.
After a month, we met and we decided it was real
and I thought it was real.
That moment of looking at the data
really kind of made me jump.
We kept it quiet even though we have 1,000 people.
We had one office with several students in it
and all of them apart from one
were in the LIGO collaboration.
So this poor one student for about six months,
every time he walked into the office,
everybody stopped talking.
And he would just must've been wondering
what he had done wrong.
Was it him, was it me?
Barry and his team agreed
that they would publish their results in the journal,
Physical Review Letters, before Christmas 2015.
We had our final meeting to decide
that it was ready to go,
but it hung up in that meeting.
We couldn't agree to publish it.
So, why?
Basically, we argued over adjectives.
Is the title discovery of gravitational waves?
Is a title evidence for?
Anyway, we hung up on this.
And so it took another, I don't know, week.
And then we call Physical Review Letters and they said,
oops, it's too late.
It's too close to Christmas.
So we said, okay then,
we're not giving it to you until after Christmas vacation.
I tell this story because on December 26th, Boxing Day,
we saw our second event.
Only three and a half months
after the first one,
another black hole merger has been detected.
And even though I had gone through all this intense fall
and was absolutely, I thought, convinced,
seeing the second event was a sigh of relief in me.
I didn't anticipate it,
but there's something about confirmation,
no matter how much you look at something
and believe what you've done.
In February, 2016,
LIGO, in collaboration with Virgo, told the world.
Ladies and gentlemen,
we have detected gravitational waves.
We did it.
The discovery of gravitational waves,
probably the brightest discovery
since we first learned that the universe was expanding.
Not only do we now have to believe in black holes,
we have to believe that they collide too.
Seeing the first detection of gravitational waves
while we were still in the thick
of doing the Event Horizon Telescope observations
was awesome, was remarkable.
There's always been a little bit of friendly competition
with the LIGO team,
although I would say that I think
people didn't appreciate how close we came.
We accreted new partners in Europe,
new partners even from Asia,
and we all began to work together.
And finally in 2017, we were ready.
But a telescope that covers the whole planet
has some uniquely frustrating problems.
With normal astronomy,
you need good weather at just the telescope
that you're observing with.
But for very long baseline interferometry,
you need weather to be good all over the globe.
And I was going up to the 30 meter dish at IRAM,
but it was perfect weather.
And it was perfect weather all around the world
in the other places.
And I said, this can't be.
We have good weather here.
And I was just nervously looking at the hard drives
and the equipment,
but all the equipment was working.
At each of these telescopes,
we have what's called a hydrogen maser
that time tags the data.
It's digitized, stored on a hard disk drive.
And they're sent by trucks,
by planes to a central facility where they're played back.
The first data was coming back to Haystack
and they were doing the first correlations
and the first success messages came.
These two telescopes work together,
these two telescopes work together.
And after a few weeks and months,
we knew it probably worked.
We didn't know what it was showing us.
That moment of joy and epiphany came
when we saw the data making this ringing pattern.
And it was the curve of data going down,
going up again.
Just looking at it, we thought we have a ring.
If that's true, we're really lucky.
My God, this worked.
Once all the data
from all the radio telescopes are collected,
the process of creating an image can begin.
And so what they do is they take the light
that has been frozen at every one of these telescope sites
and they play them back together,
and they average it and average it down so much
because there's only a tiny little signal
riding on a huge amount of noise.
You get a little nugget of information.
It's passed onto the imaging algorithms.
And that little nugget is what we use to make the picture.
But the problem is,
we're not collecting light from everywhere.
There's a lot of holes now.
And so the big challenge was taking that sparse
and noisy data,
and using it to recover an image.
It's very similar to,
if you're listening to a song being played on a piano
that has a lot of broken keys.
And even though a lot of the keys are broken,
a lot of times,
you can still try to make out what the song is.
The problem is what we call.
There are many different kinds of images
that correspond to the same data that we measure.
And so we wanted to put all that aside and just say,
ignoring all of our ideas of general relativity
and all the ideas that we believe right now
as far as what we think a black hole should look like,
and just see, what is the data telling us?
Does the data say that there should be a rig?
How about a disk?
How about, you know, an elephant?
It would be very easy within a collaboration of this size
to have three different imaging techniques, for example,
and to decide that one of them was the way to go,
but that would have created a problem
because proponents of the other two methods,
which might be also very, very good,
would have felt left out.
So we adopted this method
of using all of these different techniques
as checks and balances.
We allowed everyone to proceed
with their particular algorithm,
their particular imaging method.
And we worked in isolation with our teams for seven weeks,
trying to make what we thought was the best image
from the data.
And after those seven weeks,
we actually all gathered together
in Cambridge, Massachusetts,
and we revealed those images to each other.
But no matter what,
even though the different teams
resulted in slightly different images,
the same underlying structure was there.
This ring of roughly 40 micro arcseconds
that was brighter on the bottom than the top.
But even at that point, we weren't sure.
And so what we did is then we spent the next couple months
essentially trying to break our images.
So we took the data and we trained our methods.
We figured out, what were the hyper parameters?
What were the knob settings of our methods
to recover things like disks instead?
So we'd generate a disk on the sky
as if the Event Horizon Telescope were seeing a disk
with no hole in the center.
And then we'd transfer those exact parameters
onto the real M87 black hole data.
And although we had tried our hardest
to find parameters
that would recover a disk with no hole in the center,
we saw that when we used those parameters
on the real black hole data,
the data forced us to put a hole there.
Although each individual image looks slightly different,
that ring structure is consistent across all of them.
That's when we knew
that we could then come with a real scientific publication
and a real press release to announce
that we'd seen the first image of a black hole.
The image had to be released all around the world
at the right moment.
It was all timed to the second.
And for that, Professor Heino Falcke is here.
We all knew in July of 2018 what we had.
Dr. Shep Doeleman, EHT's director.
And from that time until April of the following year,
nobody breathed a word of it.
We all kept the results secret.
Now, I have to fill the time actually
until we are allowed to actually, you know,
start the unveilings.
Despite all the struggle,
all the exhaustion,
we'd kept it under the lid
until we were able to really release it.
And even up to the day we made the announcement,
everyone thought that we were going to say something
about Sagittarius A star,
when in fact we had imaged M87.
In April of 2017,
all the dishes in the Event Horizon Telescope swiveled,
turned, and stared at a galaxy 55 million light years away.
It's called Messier 87 or M87.
And there's a super massive black hole at its core.
And we are delighted to be able to report to you today
that we have seen what we thought was unseeable.
We have seen and taken a picture of a black hole.
Here it is.
This is the first ever image of a black hole.
We all expected a lot of good reaction from the public,
but I think we're all extremely surprised
by how quickly it became a classic.
It became the iconic image of the black hole.
And this is the strongest evidence
that we have to date for the existence of black holes.
When I saw that picture,
I just thought that that was the most beautiful photo
I've ever seen.
I picked up the New York Times, as I do every morning,
and there was this beautiful picture of a black hole.
But I just didn't see it coming.
And then suddenly, there were these pictures,
and I was like, wow.
Nobody had ever seen a black hole.
At some point, seeing is believing.
Some people think of black holes as being frightening
and confusing objects,
but I don't.
I'm in love with them.
I have no problem cozying up to black holes
but for the simple reason that they are the one way
we'll be able to understand how gravity
and quantum mechanics ultimately have to join forces.
Physics is a terribly embarrassing field.
We've got two wonderful theories
and never the twain shall meet.
One works very well,
predicts almost anything
that you can do with short distances.
And the other one predicts almost anything we see
in relativistic astronomy.
Most of the time,
gravity is completely separate from quantum mechanics.
Gravity is so much weaker
than the quantum forces on the nuclear level.
The black hole is the one place where gravity
plays with all the other forces on an equal footing.
You're getting close to the singularity,
maybe you actually never get to the point
where you have an infinite density
because in quantum mechanics,
you can not have something which is more
because it will always have space-time
kind of fuzzy around that size.
But maybe what's happening there
have some effect at the level of the event horizon.
And that's where we can probe with gravitational waves
because when two black holes collide,
the two event horizons, they merge,
and they are very excited and are shedding a lot of energy.
The shape of the gravitational waves that come out
might carry a little bit of information
of what's going on inside.
Might.
We don't know, but maybe we will find out soon.
Einstein didn't think
we could detect gravitational waves,
and he didn't believe black holes existed at all,
even though they were born from his own theories.
Now, we know for sure that they do exist
and they may reveal the inner workings of the universe.
And that's why black holes draws
because they represent what we don't know
and they represent what we could know.
If I were to meet Einstein today,
I would tell him, "Hey, Albert, I'm sorry,
we proved you wrong."
Chasing Black Holes
You had to admire people that would devote years,
and years, and years to trying to do it
with the likelihood being that they would fail.
More than two decades ago,
two teams decided to try what seemed impossible,
to prove the existence of black holes.
Both of these kind of converged at almost the same time,
but coming about it in very different ways.
We're crazy optimists in this business.
You have to be.
If you look at all the things that could go wrong,
then why would you dedicate your career to it?
One team called LIGO
plan to use giant instruments
called laser interferometers
to detect tiny ripples in space-time
called gravitational waves
caused by the merger of two black holes.
Albert Einstein thought that it will be impossible
to ever detect them.
We're looking for motions
that are about 1,000th of the size
of a proton inside the nucleus of an atom.
The other team
called the Event Horizon Telescope
would try to capture an actual image of a black hole,
a task comparable to seeing a tennis ball on the moon.
And so you need a telescope the size of the entire earth.
But obviously,
we couldn't build a telescope the size of the the earth.
But in radio astronomy,
you can play this near magical trick
where you take two telescopes separated by some distance,
bring the data that they receive together
to form a telescope
as though you had one as large as the distance
between these telescopes.
Finding black holes,
either by taking an image of one
or by detecting them with gravitational waves
would be among the most difficult challenges
in the history of science.
I was sort of trembling in front of the challenge.
It was hopeless
because the technology would never be there.
We staged experiments that went nowhere.
We tried and we failed.
And if they found what they were looking for,
could they keep it secret until they were sure?
Many people had a eureka moment.
I was completely blown away
by how clear cut this signal was.
And I had a moment of panic.
We all kept the results secret.
Because I knew if that was right,
we would be writing history.
All I could do is figure out how we had gone wrong,
how are we fooling ourselves?
At some point,
you have to stop and just listen to the universe.
Black holes were born of theory.
They might really exist,
but there was no direct proof.
Many felt that the idea itself must be fatally flawed.
An extremely massive object
will curve the space-time around it so much
that anything that passes by
will actually fall into that object
and will not be able to escape.
Even light wouldn't be able to escape,
and that's basically a black hole.
But is this real?
Clearly, something like this would not occur in nature.
No one felt that nature would be that crazy.
These cosmic objects are more extraordinary
than we could ever have dreamt up.
At the center of a black hole,
the laws of physics as we know today,
they break down.
There's no matter there.
The structure that we see is a vacuum structure.
It's like a tornado in space-time.
In their purest form,
the concept of black holes starts with Albert Einstein.
For physicists,
the great year for Einstein was 1905
when he solved several of the biggest problems in physics
within a matter of months.
And then he spent 10 years developing his theory of gravity,
which is general relativity.
What we see coming out of general relativity
is that matter and energy has the ability
to affect the space time around it.
Space time is a mathematical concept
that unites time with three dimensions of physical space
so that they are intimately woven together
into what is called the fabric of space time.
The actual presence of matter distorts this fabric.
If you send a light beam past a, say, a star or a planet,
the reason it bends is
because it's trying to follow the straightest path possible
in a geometry that is now curved.
And thus, that is what gravity is.
It's basically the geometry of space time.
Within weeks of Einstein
publishing his revolutionary theory of gravity,
people began trying to figure out what it all meant.
One of the first
was a German physicist called Karl Schwarzschild.
Despite being in his forties,
Karl decided to volunteer to fight in the First World War
and he worked as an artillery man.
He was able to predict how space
and time would look like around a point mass.
And he realizes if you compact matter
into a small enough volume,
that there is this event horizon
at what we can now call the Schwarzschild radius
where the speed of escape is larger than the speed of light,
where even light cannot make it out.
A one way boundary
where information and light can only go in,
but never can come out again.
That's the moment when black holes really were born.
Of course, completely crazy idea.
This wouldn't exist.
Neither man realized that it was anything more
than a mathematical construct
that they didn't really have to worry about
because no one felt that nature would allow it to happen.
Poor Karl Schwarzschild,
he died very quickly after finding the solutions
due to an illness that he got in the trenches there.
Throughout the 20th century,
the theoretical study of black holes became vibrant
and popular among scientists.
They had theories for how they might form.
And astronomers had even started to speculate
about where they might lurk in the cosmos.
So there are two varieties of black holes,
the stellar variety,
which are born during the death of stars.
And they weigh a few or 10 times what our sun does.
Stellar mass black holes
are thought to have a Schwarzschild radius
of around the size of a small modern city.
Even though they have a mass of 10 or 20 suns,
they are tiny in comparison to the stars
from which they were born.
Super massive black holes, however,
would contain the combined mass of millions
or even billions of suns.
They could have a Schwarzschild radius
the size of an entire solar system.
Astronomers wondered if stellar black holes
were hiding in all galaxies
and that super massive black holes
might be found at the heart of many.
They even named the super massive black hole
that they suspected might lie at the heart of our galaxy
Sagittarius A star.
Are they out there?
The theory predicted it, but are they really real?
To try and answer that question,
the team called LIGO
would try to find the smaller black holes
with gravitational waves,
while the Event Horizon team
would hunt the super massive black holes
using a type of astronomy that uses radio waves.
If you have optical light,
the wavelengths are tiny, tiny, tiny.
But if you study radio emission,
the wavelengths are actually big.
And so you have a gigantic dish
that can reflect radio waves.
It was intense radio emissions
coming from the center of galaxies
that first made astronomers suspect
that super massive black holes must be the culprit.
So we had to think a little bit
about these emission mechanisms,
like what would cause is glow in radio waves?
And so what can power something like that
is gravitational accretion onto a super massive black hole.
That's the only way we know of to power something like this.
In the 20th century,
the idea of capturing an image of a black hole
with these radio emissions
was far beyond the realms of possibility,
but some people did dare to dream.
Many times in science,
you wind up in a situation
where somebody makes an initial discovery
and it goes under appreciated.
And this is the case with Jean-Pierre Luminet.
I mean, this was a real visionary.
In 1979, he came with the first full simulation
of what a black hole would look like if you were there,
like if you could really see in infinite detail.
To capture an image of a real black hole
in the way that John-Pierre Luminet imagined
would require a radio telescope
with a dish as big as the planet earth
and technologies that simply didn't exist yet.
However, for the gravitational wave team,
they had an even more fundamental problem to solve.
Nobody even knew if gravitational waves were real at all.
The theory behind them originated with Albert Einstein
from around the same time
that Karl Schwarzschild was formulating his ideas
of black holes.
In 1916,
Einstein wrote a little paper, five pages long maybe.
He looked at the equations of general relativity
that he had developed,
and he noticed it had a great similarity
if you put it in a certain way
to the equations of electricity and magnetism.
And since electricity and magnetism have waves,
he conjectured that gravity must also have waves.
By the 1930s,
Einstein had written a further two papers
on the subject of gravitational waves.
According to Einstein,
this phenomenon of gravitational waves
is very interesting from a theoretical point of view,
but any real effect on earth will be so small
that it will be very likely
always impossible to ever detect them.
They were never fully accepted
during this whole period and beyond,
even until his death.
The great Albert Einstein died in 1955.
Black holes and gravitational waves,
ideas that were rooted in his theories,
were far from being considered a reality.
But in 1958,
50 of the world's greatest experts
in general relativity had a meeting.
So at this now renowned conference,
the question of whether gravitational waves really do exist
and produce something that could be measured was raised.
Richard Feynman was there.
He said, if gravitational waves really exist,
they have to be able to do something.
They can't just exist.
They have to be able to transfer energy.
And so he made a,
what we call like a experiment,
just a thought experiment.
And that was if you have a bar
and you put a couple of rings on it,
and then a gravitational wave goes through the bar,
it'll take the bar,
and it'll expand it, and contract it,
and expand it, and contract it
at the frequency of the gravitational wave.
As it expands, it, of course, pushes on the rings
and they would move.
But what's happening
is that they're transferring energy friction to the ring.
This clever thought experiment
inspired one of the people at the meeting
to actually build such a device.
His name was Joseph Weber
and the machines he created are called Weber bars,
the first ever gravitational wave detectors.
It's accepted today that his bar detectors
were not sensitive enough to be able to make any detections,
but he played a very important role
in sparking the rest of the world's interest
into what a different type of detector might look like
that could ultimately be more sensitive.
So gravitational waves, if they exist,
how do they manifest themselves?
If I take just a place
and a gravitational wave comes through,
it distorts the space and time
such that it stretches it in one direction
and squashes it in the other direction.
The easiest way to think about it
is what happens when you go to the amusement park
and you see these mirrors.
You look in one and you get tall and thin.
You go to the next one, you get short and fat.
If you imagine that you're now a detector,
it's basically going back and forth
between those two mirrors.
You're getting taller and shorter
and fatter at the frequency of the gravitational waves.
So maybe the gravitational wave is 60 Hertz,
so 60 times a second, you're going.
And that's what we have to measure.
So how are we gonna measure that?
So we had to find out a way
to be able to detect here on earth
some very, very small perturbations.
And it turned out that the most sensitive instrument
that we could make and design
is what is called a laser interferometer.
So what does an interferometer do?
I take a beam of light
and I get to a place where there's a mirror
that sends half the light in one direction
and half the light in the perpendicular direction.
If somewhere down the way, I put a mirror,
it'll bounce back.
And if I've calibrated it so that they're the same length,
they'll come back at exactly the same time.
If I invert the signal from one to the other,
they'll have exactly canceled.
I have a little photo detector.
It sees nothing, okay?
But now imagine the gravitational wave
went through this same thing.
And 60 times a second,
one arm's gonna get a little longer than the other.
The light will come back at a slightly different time
and the light will go 60 times a second,
depending on how strong the gravitational wave was.
And that's all we have.
We have a detector
that's measuring the length of these two arms.
That's called interferometry.
Two of the people
who believe that interferometry might be the solution
for finding gravitational waves,
and therefore, for finding black holes
were Rainer Weiss from MIT and Kip Thorne from Caltech.
I started on the faculty at Caltech
as a particle physicist at the same time
Kip Thorne started as a general relativist,
and we were close friends.
By the 1980s,
several groups from around the world
would start to consider building interferometers
to search for gravitational waves.
The British and Germans formed a collaboration.
The Italians and French
started a collaboration called Virgo.
And in the U.S.,
Rainer Weiss, Ron Drever, and Kip Thorne formed LIGO.
But getting funding to build such a venture
was not going to be easy.
What was pointed out in the U.S. Congress
about whether to appropriate funds
was just how extraordinarily sensitive
these detectors would have to be.
You would need to be able to measure these mirrors
moving at distances
that were some 10,000 times smaller
than the diameter of an atom.
And when thinking about that in an astronomical terms,
that's maybe, say, the same as being able
to measure the distance to the nearest star
and being able to say definitively
whether that distance has changed
by the width of a human hair.
And so these mind-boggling ideas
were used as a way to cast doubt
on the plausibility of such an experiment.
In 1990,
the National Science Foundation
approved the construction of two detectors,
Livingston, Louisiana, and Hanford, Washington State.
But LIGO was going to need a lot of money
and firm leadership
if it was to stand a chance of finding stellar black holes
with gravitational waves.
The Event Horizon Telescope team
that would try to take an image
of a super massive black hole
would not officially form for many more years,
though the ideas upon which it would be based
were beginning to emerge.
When I was in graduate school,
I had the great fortune to work with Dr. Alan Rogers,
who was one of the pioneers of radio astronomy.
And he got me hooked on using interferometry
to make the most detailed images of the sky
that we could make at the time.
And we started doing this with Sagittarius A star,
which we thought was a super massive black hole
in the center of our Milky Way galaxy.
So there is a technique out there
that helps us with increasing the resolution
that we can get from a telescope,
which is not to use a single one,
but to use pairs of telescopes
and look at the distant object at the same time
with these pairs of telescopes,
record the incoming light.
And because we time tag exactly when each signal arrived,
we can after the fact combine these signals
using super computers,
just match them up
and pretend we had the resolution of a telescope
that's as big as the separation between the two.
We call it very long baseline interferometry
and the key thing here is that we interfere the signals
that we receive on different parts of the earth.
To use very long baseline interferometry
or VLBI to actually capture an image
of a super massive black hole
was going to require a radical leap forward
in our understanding of how matter
spins around a black hole that is feeding,
the so-called accretion disk.
I think any revolution requires a group of people
sorting out the ideas that they have,
figuring out what's the best way to move forward.
I met Dimitrios when we were both at Harvard.
In fact, our first paper together is on Sagittarius A star,
the emission from it,
and whether the accretion flow around the black hole
would allow us to see down to the horizon.
It was the time
where we had just started thinking differently
about how do black holes accumulate matter
from their surroundings.
For about 20 to 30 years, there was a paradigm,
but then it became obvious sometime in the '90s
that that paradigm was not right.
So it was that small group
that started building a new paradigm.
We were doing the theoretically, you know?
We were thinking about,
where does this radio emission come from in black holes?
Some people proposed that this would come
from the matter of that shoots out from the plasma jets,
and other colleagues were calculating
that the radio emission would come from the matter
falling into the black hole.
We had group meetings that lasted eight hours.
We just had coffee,
we just had the white board and we were trying out ideas.
And most of them were wrong,
most of them failed,
but a few of them turned out to be really important.
One of those questions that we wanted to ask is,
what would the black hole look like
if I were to take a picture?
Will it be a big fuzzy, fluff of cloud,
or will it be something really small,
as small as the Event Horizon of the black hole?
So we spent about 10 years learning how to tweak the theory
and make predictions of black holes
that were not the same
as the ones that Albert Einstein predicted.
And lo and behold,
we always saw this dark shadow in the very center.
Irrespectively of how the matter was rotating,
whether it was falling in,
whether it was flowing out,
whether the black hole was rotating,
as long as the mass was known as a black hole,
the shadow had to be there.
So in year 2000,
I realized that at certain wavelengths,
the entire accretion flow is actually transparent
and allows us to see all the way down
to the horizon of the black hole,
which is what we're after, that shadow.
How would that look like?
Could we actually see that?
And it turns out that the answer depended on
how did I take the picture,
what wavelengths of light did I use?
This new multinational community
of radio astronomers and theorists realized
that the key was to push very long baseline interferometry
to see higher frequencies of light.
The higher frequency you observe,
the finer your angular resolution.
Your pictures get much crisper
and you can see the details that you want to see.
But the second thing is
that as you move higher in frequency,
you can see more deeply into all the hot gas.
So you really want to see all the way to the event horizon.
You want to get sharper images,
and that pushed us to higher frequencies.
The target that everyone was focusing on
was the suspected super massive black hole
at the center of our galaxy, Sagittarius A star.
I was convinced we would be able to see
the black hole at the center of our Milky Way
with a global telescope array,
but that required a lot of money,
new telescopes, new receivers,
and it required a big community to work together.
There's a transition
that many areas of science go through
in terms of moving from things in the laboratory,
taking those ideas and concepts and scaling them up
to turn those into a large scale project,
it's a skill in itself.
Barry Barish
had been working in particle physics
at the superconducting supercollider in Texas.
The U.S. Congress canceled the supercollider
at a time when LIGO needed somebody
that was capable of taking it
to get both the funding that was needed
and be able to put it together
in a way that would make it work.
In 1994, the National Science Foundation
made Barish the laboratory director of LIGO.
About six months later,
went to the National Science Board
that oversees the National Science Foundation,
and convinced them, and they funded us.
I think the real hero is not me, it's the NSF.
For me, the big issue was to get a strong enough team,
as good as these two institutions are,
Caltech and MIT.
For our problem as hard as gravitational waves,
we needed to tap the best people in the world.
The underlying technologies
that are relevant for gravitational wave detection
were being developed in different places.
In 1997,
Barish established the LIGO Scientific Collaboration,
which merged several international groups into LIGO.
This meant there were now two major multinational groups
vying to build detectors,
LIGO and Virgo.
The group that would become the Event Horizon Telescope
were also making progress.
So we all started to work together
at three millimeter wavelengths,
and then we realized that to push it even further,
we had to go to one millimeter wavelength.
And that's when we began this little race,
a competitive but also a collegial race.
Could we push this technique to its real limits?
And that was great time.
And then we ran into this roadblock.
For many years, we were stymied.
We were just stuck
because we didn't have the sensitivity we needed
to make these observations at high frequencies.
As soon as you go to high frequencies,
everything becomes harder.
The atmosphere reduces the signal coming from black holes.
The superconducting cameras
that we mount on each of our telescopes
to receive the radio waves from the black hole,
they become more noisy.
So everything is working against you.
The one thing that we had,
our secret weapon was that we could increase the bandwidth.
Starting at around 2000 and going til about 2006,
there was this explosion in capability
that happened because we started building our instruments
out of commercial electronics.
Imagine that.
Up until that point,
we had been developing specialized instrumentation
that took a decade to design,
and manufacturer, and get into the field
because it was so exquisitely specialized.
Graphics processing units or GPUs
are specialized computer chips
that are used primarily for video graphics.
We weren't pushing the development of GPUs.
In fact, the gaming industry was,
but we thought, hey,
we can solve Einstein's equations on these things
and we can do it much faster than with traditional CPUs.
The same was true for data storage.
Consumer hard drives were becoming ever faster
and greater in capacity.
All of a sudden, we could go to the store,
buy components, hook them together,
and we could make something that was 10 times more capable,
10 times lower cost,
and we could design it 10 times faster.
It was nothing short of a miracle.
The LIGO detectors had to be
the most sensitive scientific instruments in the world.
And it's all because our planet is such a noisy place.
Much of the good work happens at night,
and that's simply because the environment around you
is quieter at night.
There are fewer cars driving on the road,
hitting bumps and causing the ground to shake,
people, falling trees that are falling down in the forest.
The operation of heavy machinery nearby,
it's also lower at night.
There's also natural effects.
Of course, earthquakes are an obvious consideration
that might jump into many people's minds.
Believe it or not,
even if you're in the center of a continent,
there is a peak of motion that occurs at very low frequency
due to waves beating on the ocean shore.
The reason that all of these sorts of things matter
is because ultimately,
these mirrors that we're trying to detect
very small motions of are connected to the ground.
The earth ground is shaking all the time
by about a millionth of a meter,
which is a million of a million times more
than what we're trying to measure.
So this is one reason that our system
can not just sit on the ground,
but our mirrors are the most quiet place on earth.
The beam splitter itself is a mirror,
a big piece of glass suspended on wires.
That has to be isolated from external disturbances,
so it's housed inside a vacuum tank.
That vacuum tank is inside a big building.
Out from that building go two arms.
Inside that arch, there's a vacuum pipe.
Inside that travel the laser beams
that travel the length of these four kilometer arms.
The ends of the arms inside those buildings,
there are vacuum tanks.
Inside those tanks are some isolation systems
that isolate from ground vibrations,
mirrors that are suspended in the initial incarnation
of LIGO on metal wires.
But it's not just vibrations and noise
that LIGO has to fight against,
but barely imaginable quantum effects.
To measure precisely how the two waves get together,
you have to measure how many photons hit your detector.
But since the number of photons
is what we call a quantum variable,
there's an intrinsic uncertainty there.
We cannot do better
than what the Heisenberg uncertainty principle dictate.
That's one reason we need very powerful lasers.
As we increase the laser power,
the force exerted by the light on the mirrors
also increases proportionally.
So that becomes one of the big challenges
of increasing the laser power.
Every time a photon bounces off our suspended mirror,
it transfers some momentum to the mirror.
It gives it a kick,
If it were to strike it off center,
it's going to actually create a torque.
The light causes the mirrors to twist.
In an ideal situation,
you would have the light striking the center of the mirror.
So to be able to understand more about black holes
and big stuff out there in the universe,
we need to understand very well
the physics of the quantum mechanics
and the thermal motion of the atom,
which is very small scale physics.
In 2006,
we fielded these new electronic systems for the first time,
ad we took them to two sites,
one in Arizona, Mount Graham,
and one to Mauna Kea in Hawaii.
We set up this experiment,
but we were really flying a bit by the seat of our pants.
So back at the core later,
we played these data streams from Hawaii
and Arizona back again, and again, and again.
And we searched for months.
And it was a heartbreaker.
After months, we threw in the towel.
We realized that we were just not gonna be able
to detect Sagittarius A star.
Later, we found out why.
A small little chip of metal
had fallen into the heart of the superconducting junction
in the receiver of the Caltech Submillimeter Observatory.
We were able to go back the next year,
adding a new site in California.
And this time, we were successful.
We got the detections.
And that showed us immediately
that we were seeing horizon scale structure.
That's the moment we knew
that we could make an image of a black hole.
The LIGO team was sure they had detectors
that were working well,
but if black holes were spiraling into one another
somewhere in the universe,
they were not hearing it.
You don't know that you don't see something
until you look for a while.
And so it would run for six months or a year,
and wouldn't see any events.
We'd turn off, lick our wounds,
put in some improvements, and then run again.
But I think that is also maybe part of the excitement.
So there was years, and years, and years
of continually improving the detectors,
collecting data, improving them,
collecting data on no detection, no detection, no detection.
There were times where myself or the community thought
that maybe we got ourself into something too big.
The initial version of LIGO wasn't good enough
to see gravitational waves,
it turns out.
In 2010,
the detectives were scheduled to shut down
in preparation for a big upgrade called advanced LIGO.
The team trying to make an image of a black hole realized
that the friendly competition
between several European and U.S. institutions
wasn't going to be enough.
We would need many more dishes.
We would need the European dishes,
we need the American dishes.
So it was clear at some point,
those different efforts had to come together
and merge into a global collaboration.
And so there was a meeting in 2009 in California,
and I was sitting there at a coffee break together
with Shep Doeleman and was saying,
if we want to get this funded later,
we can't just keep talking about
submillimeter VLBI array, blah, blah, blah.
Nobody will understand what it is.
We have to give it a flashy name.
What about if we call it Event Horizon array
or something like this?
And in the end, we came down to Event Horizon Telescope,
and that's how it started.
The Event Horizon Telescope was finally
and officially born.
Meanwhile, LIGO was in the middle of its upgrade
from initial LIGO to advanced.
We wanted to build a very sensitive instrument.
And that meant that we had to take some risks, if you want.
Much of the effort
that went into the advanced LIGO upgrade
involved the mirrors
and how they were isolated from the outside world.
So instead of suspending these mirrors on metal wires,
we were actually suspending them on glass fibers,
fused silica fibers.
Those fibers, they're very strong if you pull them,
but if a grain of dust hit them, they shatter.
Perhaps the greatest secret weapon
that the LIGO team employed
to isolate their detectives from unwanted vibrations
was the same technology found
in noise cancellation headphones.
You put on these earphones on an airplane
and the roar of the engines goes away,
and you still hear the stewardess ask you,
do you want coffee?
So what it does is measure the ambient noise of the engines
and cancel it.
But the stewardess talking to you,
it's not ambient, it's a signal,
and so you hear that fine.
So the idea was to bury inside of these shock absorbers
little seismic sensors
that measured any residual motion of the earth.
And then we just pushed back against it,
make little actuators that push
to cancel the residual motion that's there
after the shock absorbers.
And that gained us a factor of 10 in sensitivity.
The Event Horizon Telescope group
had begun to consider trying to image
another super massive black hole,
as well as Sagittarius S star.
There was another black hole out there,
which was 1,000 times further way,
but also 1,000 times more massive.
M87 is a galaxy
about 53 million light years from earth.
And just like the Milky Way,
we suspect it has a super massive black hole at its center,
only one that is much bigger.
M87 is so massive
that it doesn't change during the course of an evening,
whereas Sagittarius A star is speedier.
And during the course of a night,
it changes its appearance.
The advanced LIGO detectors
were now successfully upgraded
and showing much higher sensitivity
than anything achieved before.
When we start a data run each year when we do this,
there's a little period of time
when people who are expert on particular things
can still decide to make some changes of settings,
so we call it an engineering run.
It happened during that period.
You can imagine I can remember very well that day.
So it was a Monday.
And I remember that
because that was the day after I ran my first marathon,
so I was thinking that there was enough excitement
for a while, right?
I remember that quite vividly in my office in Glasgow
when our colleague suddenly said,
You do know the signal had just arrived?"
And we were like, "No?"
My colleague came to tell me
that it seems there's been a detection.
I just brushed this comment off.
To me, it seemed,
like we're not even in the observation run yet.
It's too early for this to actually happen.
When that event happened,
it happened seven milliseconds earlier in Louisiana
than in Hanford.
And my first thought was, this can't be real.
This too good to be real.
And you have to imagine
for people who have spent decades measuring noise,
and being very good at mentioning noise,
to have a signal arrive and it be large,
it really did take people by surprise.
The detection made
on the 14th of September, 2015
was calculated to have been created
by the merging of two stellar mass black holes
1.3 billion light years from earth.
We were expecting to have to fight our way
through a lot of justification to convince people
that we really had to think of additional ways,
but that was exactly what you would expect
out of a textbook.
So the first thing I thought is, this is not real.
Somebody did this,
because at the time,
we had a program which was called blind injections
to test if you were actually able
to detect gravitational waves.
Some people, without telling anybody,
would add on purpose fake signals.
So all I could think is that, great,
this is an artificial signal.
But we got word from whoever was in charge
of this blind injection saying, no,
we didn't have time to set up our systems.
Sorry, we're late.
So this is not a blind injection.
But that was not the end of the story.
That was the beginning of maybe six months of very hard work
to try and prove that we didn't do anything wrong.
Because you don't want to cry wolf
the first time that you detect a gravitational wave
'cause that had happened in the history in the past.
So we don't want to do that
because then you lose credibility, of course.
I think for many people, justifiably, after many years,
they had a eureka moment.
And I had a moment of panic.
All I could do is figure out how we had gone wrong.
There then followed a frantic
and intensely busy period of forensically analyzing
whether they believed their own data.
After a month, we met and we decided it was real
and I thought it was real.
That moment of looking at the data
really kind of made me jump.
We kept it quiet even though we have 1,000 people.
We had one office with several students in it
and all of them apart from one
were in the LIGO collaboration.
So this poor one student for about six months,
every time he walked into the office,
everybody stopped talking.
And he would just must've been wondering
what he had done wrong.
Was it him, was it me?
Barry and his team agreed
that they would publish their results in the journal,
Physical Review Letters, before Christmas 2015.
We had our final meeting to decide
that it was ready to go,
but it hung up in that meeting.
We couldn't agree to publish it.
So, why?
Basically, we argued over adjectives.
Is the title discovery of gravitational waves?
Is a title evidence for?
Anyway, we hung up on this.
And so it took another, I don't know, week.
And then we call Physical Review Letters and they said,
oops, it's too late.
It's too close to Christmas.
So we said, okay then,
we're not giving it to you until after Christmas vacation.
I tell this story because on December 26th, Boxing Day,
we saw our second event.
Only three and a half months
after the first one,
another black hole merger has been detected.
And even though I had gone through all this intense fall
and was absolutely, I thought, convinced,
seeing the second event was a sigh of relief in me.
I didn't anticipate it,
but there's something about confirmation,
no matter how much you look at something
and believe what you've done.
In February, 2016,
LIGO, in collaboration with Virgo, told the world.
Ladies and gentlemen,
we have detected gravitational waves.
We did it.
The discovery of gravitational waves,
probably the brightest discovery
since we first learned that the universe was expanding.
Not only do we now have to believe in black holes,
we have to believe that they collide too.
Seeing the first detection of gravitational waves
while we were still in the thick
of doing the Event Horizon Telescope observations
was awesome, was remarkable.
There's always been a little bit of friendly competition
with the LIGO team,
although I would say that I think
people didn't appreciate how close we came.
We accreted new partners in Europe,
new partners even from Asia,
and we all began to work together.
And finally in 2017, we were ready.
But a telescope that covers the whole planet
has some uniquely frustrating problems.
With normal astronomy,
you need good weather at just the telescope
that you're observing with.
But for very long baseline interferometry,
you need weather to be good all over the globe.
And I was going up to the 30 meter dish at IRAM,
but it was perfect weather.
And it was perfect weather all around the world
in the other places.
And I said, this can't be.
We have good weather here.
And I was just nervously looking at the hard drives
and the equipment,
but all the equipment was working.
At each of these telescopes,
we have what's called a hydrogen maser
that time tags the data.
It's digitized, stored on a hard disk drive.
And they're sent by trucks,
by planes to a central facility where they're played back.
The first data was coming back to Haystack
and they were doing the first correlations
and the first success messages came.
These two telescopes work together,
these two telescopes work together.
And after a few weeks and months,
we knew it probably worked.
We didn't know what it was showing us.
That moment of joy and epiphany came
when we saw the data making this ringing pattern.
And it was the curve of data going down,
going up again.
Just looking at it, we thought we have a ring.
If that's true, we're really lucky.
My God, this worked.
Once all the data
from all the radio telescopes are collected,
the process of creating an image can begin.
And so what they do is they take the light
that has been frozen at every one of these telescope sites
and they play them back together,
and they average it and average it down so much
because there's only a tiny little signal
riding on a huge amount of noise.
You get a little nugget of information.
It's passed onto the imaging algorithms.
And that little nugget is what we use to make the picture.
But the problem is,
we're not collecting light from everywhere.
There's a lot of holes now.
And so the big challenge was taking that sparse
and noisy data,
and using it to recover an image.
It's very similar to,
if you're listening to a song being played on a piano
that has a lot of broken keys.
And even though a lot of the keys are broken,
a lot of times,
you can still try to make out what the song is.
The problem is what we call.
There are many different kinds of images
that correspond to the same data that we measure.
And so we wanted to put all that aside and just say,
ignoring all of our ideas of general relativity
and all the ideas that we believe right now
as far as what we think a black hole should look like,
and just see, what is the data telling us?
Does the data say that there should be a rig?
How about a disk?
How about, you know, an elephant?
It would be very easy within a collaboration of this size
to have three different imaging techniques, for example,
and to decide that one of them was the way to go,
but that would have created a problem
because proponents of the other two methods,
which might be also very, very good,
would have felt left out.
So we adopted this method
of using all of these different techniques
as checks and balances.
We allowed everyone to proceed
with their particular algorithm,
their particular imaging method.
And we worked in isolation with our teams for seven weeks,
trying to make what we thought was the best image
from the data.
And after those seven weeks,
we actually all gathered together
in Cambridge, Massachusetts,
and we revealed those images to each other.
But no matter what,
even though the different teams
resulted in slightly different images,
the same underlying structure was there.
This ring of roughly 40 micro arcseconds
that was brighter on the bottom than the top.
But even at that point, we weren't sure.
And so what we did is then we spent the next couple months
essentially trying to break our images.
So we took the data and we trained our methods.
We figured out, what were the hyper parameters?
What were the knob settings of our methods
to recover things like disks instead?
So we'd generate a disk on the sky
as if the Event Horizon Telescope were seeing a disk
with no hole in the center.
And then we'd transfer those exact parameters
onto the real M87 black hole data.
And although we had tried our hardest
to find parameters
that would recover a disk with no hole in the center,
we saw that when we used those parameters
on the real black hole data,
the data forced us to put a hole there.
Although each individual image looks slightly different,
that ring structure is consistent across all of them.
That's when we knew
that we could then come with a real scientific publication
and a real press release to announce
that we'd seen the first image of a black hole.
The image had to be released all around the world
at the right moment.
It was all timed to the second.
And for that, Professor Heino Falcke is here.
We all knew in July of 2018 what we had.
Dr. Shep Doeleman, EHT's director.
And from that time until April of the following year,
nobody breathed a word of it.
We all kept the results secret.
Now, I have to fill the time actually
until we are allowed to actually, you know,
start the unveilings.
Despite all the struggle,
all the exhaustion,
we'd kept it under the lid
until we were able to really release it.
And even up to the day we made the announcement,
everyone thought that we were going to say something
about Sagittarius A star,
when in fact we had imaged M87.
In April of 2017,
all the dishes in the Event Horizon Telescope swiveled,
turned, and stared at a galaxy 55 million light years away.
It's called Messier 87 or M87.
And there's a super massive black hole at its core.
And we are delighted to be able to report to you today
that we have seen what we thought was unseeable.
We have seen and taken a picture of a black hole.
Here it is.
This is the first ever image of a black hole.
We all expected a lot of good reaction from the public,
but I think we're all extremely surprised
by how quickly it became a classic.
It became the iconic image of the black hole.
And this is the strongest evidence
that we have to date for the existence of black holes.
When I saw that picture,
I just thought that that was the most beautiful photo
I've ever seen.
I picked up the New York Times, as I do every morning,
and there was this beautiful picture of a black hole.
But I just didn't see it coming.
And then suddenly, there were these pictures,
and I was like, wow.
Nobody had ever seen a black hole.
At some point, seeing is believing.
Some people think of black holes as being frightening
and confusing objects,
but I don't.
I'm in love with them.
I have no problem cozying up to black holes
but for the simple reason that they are the one way
we'll be able to understand how gravity
and quantum mechanics ultimately have to join forces.
Physics is a terribly embarrassing field.
We've got two wonderful theories
and never the twain shall meet.
One works very well,
predicts almost anything
that you can do with short distances.
And the other one predicts almost anything we see
in relativistic astronomy.
Most of the time,
gravity is completely separate from quantum mechanics.
Gravity is so much weaker
than the quantum forces on the nuclear level.
The black hole is the one place where gravity
plays with all the other forces on an equal footing.
You're getting close to the singularity,
maybe you actually never get to the point
where you have an infinite density
because in quantum mechanics,
you can not have something which is more
because it will always have space-time
kind of fuzzy around that size.
But maybe what's happening there
have some effect at the level of the event horizon.
And that's where we can probe with gravitational waves
because when two black holes collide,
the two event horizons, they merge,
and they are very excited and are shedding a lot of energy.
The shape of the gravitational waves that come out
might carry a little bit of information
of what's going on inside.
Might.
We don't know, but maybe we will find out soon.
Einstein didn't think
we could detect gravitational waves,
and he didn't believe black holes existed at all,
even though they were born from his own theories.
Now, we know for sure that they do exist
and they may reveal the inner workings of the universe.
And that's why black holes draws
because they represent what we don't know
and they represent what we could know.
If I were to meet Einstein today,
I would tell him, "Hey, Albert, I'm sorry,
we proved you wrong."