[Script Info]
Title: 
[Events]
Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
Dialogue: 0,0:00:01.44,0:00:03.30,Default,,0000,0000,0000,,In the movie "Interstellar,"
Dialogue: 0,0:00:03.32,0:00:06.65,Default,,0000,0000,0000,,we get an up-close look\Nat a supermassive black hole.
Dialogue: 0,0:00:06.67,0:00:08.81,Default,,0000,0000,0000,,Set against a backdrop of bright gas,
Dialogue: 0,0:00:08.84,0:00:10.96,Default,,0000,0000,0000,,the black hole's massive\Ngravitational pull
Dialogue: 0,0:00:10.98,0:00:12.42,Default,,0000,0000,0000,,bends light into a ring.
Dialogue: 0,0:00:12.44,0:00:14.55,Default,,0000,0000,0000,,However, this isn't a real photograph,
Dialogue: 0,0:00:14.57,0:00:16.36,Default,,0000,0000,0000,,but a computer graphic rendering --
Dialogue: 0,0:00:16.38,0:00:19.77,Default,,0000,0000,0000,,an artistic interpretation\Nof what a black hole might look like.
Dialogue: 0,0:00:20.40,0:00:21.57,Default,,0000,0000,0000,,A hundred years ago,
Dialogue: 0,0:00:21.59,0:00:25.19,Default,,0000,0000,0000,,Albert Einstein first published\Nhis theory of general relativity.
Dialogue: 0,0:00:25.22,0:00:26.66,Default,,0000,0000,0000,,In the years since then,
Dialogue: 0,0:00:26.68,0:00:29.65,Default,,0000,0000,0000,,scientists have provided\Na lot of evidence in support of it.
Dialogue: 0,0:00:29.68,0:00:32.76,Default,,0000,0000,0000,,But one thing predicted\Nfrom this theory, black holes,
Dialogue: 0,0:00:32.78,0:00:35.13,Default,,0000,0000,0000,,still have not been directly observed.
Dialogue: 0,0:00:35.16,0:00:38.36,Default,,0000,0000,0000,,Although we have some idea\Nas to what a black hole might look like,
Dialogue: 0,0:00:38.39,0:00:41.17,Default,,0000,0000,0000,,we've never actually taken\Na picture of one before.
Dialogue: 0,0:00:41.19,0:00:45.47,Default,,0000,0000,0000,,However, you might be surprised to know\Nthat that may soon change.
Dialogue: 0,0:00:45.49,0:00:49.66,Default,,0000,0000,0000,,We may be seeing our first picture\Nof a black hole in the next couple years.
Dialogue: 0,0:00:49.68,0:00:53.64,Default,,0000,0000,0000,,Getting this first picture will come down\Nto an international team of scientists,
Dialogue: 0,0:00:53.66,0:00:55.23,Default,,0000,0000,0000,,an Earth-sized telescope
Dialogue: 0,0:00:55.26,0:00:58.09,Default,,0000,0000,0000,,and an algorithm that puts together\Nthe final picture.
Dialogue: 0,0:00:58.11,0:01:01.64,Default,,0000,0000,0000,,Although I won't be able to show you\Na real picture of a black hole today,
Dialogue: 0,0:01:01.66,0:01:04.57,Default,,0000,0000,0000,,I'd like to give you a brief glimpse\Ninto the effort involved
Dialogue: 0,0:01:04.60,0:01:06.21,Default,,0000,0000,0000,,in getting that first picture.
Dialogue: 0,0:01:07.48,0:01:08.91,Default,,0000,0000,0000,,My name is Katie Bouman,
Dialogue: 0,0:01:08.94,0:01:11.50,Default,,0000,0000,0000,,and I'm a PhD student at MIT.
Dialogue: 0,0:01:11.53,0:01:13.56,Default,,0000,0000,0000,,I do research in a computer science lab
Dialogue: 0,0:01:13.58,0:01:16.88,Default,,0000,0000,0000,,that works on making computers\Nsee through images and video.
Dialogue: 0,0:01:16.90,0:01:19.06,Default,,0000,0000,0000,,But although I'm not an astronomer,
Dialogue: 0,0:01:19.09,0:01:20.37,Default,,0000,0000,0000,,today I'd like to show you
Dialogue: 0,0:01:20.40,0:01:23.30,Default,,0000,0000,0000,,how I've been able to contribute\Nto this exciting project.
Dialogue: 0,0:01:23.32,0:01:26.15,Default,,0000,0000,0000,,If you go out past\Nthe bright city lights tonight,
Dialogue: 0,0:01:26.18,0:01:28.61,Default,,0000,0000,0000,,you may just be lucky enough\Nto see a stunning view
Dialogue: 0,0:01:28.64,0:01:30.13,Default,,0000,0000,0000,,of the Milky Way Galaxy.
Dialogue: 0,0:01:30.16,0:01:32.62,Default,,0000,0000,0000,,And if you could zoom past\Nmillions of stars,
Dialogue: 0,0:01:32.64,0:01:36.40,Default,,0000,0000,0000,,26,000 light-years toward the heart\Nof the spiraling Milky Way,
Dialogue: 0,0:01:36.42,0:01:39.94,Default,,0000,0000,0000,,we'd eventually reach\Na cluster of stars right at the center.
Dialogue: 0,0:01:39.96,0:01:43.17,Default,,0000,0000,0000,,Peering past all the galactic dust\Nwith infrared telescopes,
Dialogue: 0,0:01:43.20,0:01:47.06,Default,,0000,0000,0000,,astronomers have watched these stars\Nfor over 16 years.
Dialogue: 0,0:01:47.09,0:01:50.68,Default,,0000,0000,0000,,But it's what they don't see\Nthat is the most spectacular.
Dialogue: 0,0:01:50.70,0:01:53.76,Default,,0000,0000,0000,,These stars seem to orbit\Nan invisible object.
Dialogue: 0,0:01:53.79,0:01:56.11,Default,,0000,0000,0000,,By tracking the paths of these stars,
Dialogue: 0,0:01:56.14,0:01:57.43,Default,,0000,0000,0000,,astronomers have concluded
Dialogue: 0,0:01:57.45,0:02:00.58,Default,,0000,0000,0000,,that the only thing small and heavy\Nenough to cause this motion
Dialogue: 0,0:02:00.61,0:02:02.58,Default,,0000,0000,0000,,is a supermassive black hole --
Dialogue: 0,0:02:02.60,0:02:06.78,Default,,0000,0000,0000,,an object so dense that it sucks up\Nanything that ventures too close --
Dialogue: 0,0:02:06.80,0:02:08.30,Default,,0000,0000,0000,,even light.
Dialogue: 0,0:02:08.32,0:02:11.38,Default,,0000,0000,0000,,But what happens if we were\Nto zoom in even further?
Dialogue: 0,0:02:11.40,0:02:16.14,Default,,0000,0000,0000,,Is it possible to see something\Nthat, by definition, is impossible to see?
Dialogue: 0,0:02:16.72,0:02:19.96,Default,,0000,0000,0000,,Well, it turns out that if we were\Nto zoom in at radio wavelengths,
Dialogue: 0,0:02:19.99,0:02:21.67,Default,,0000,0000,0000,,we'd expect to see a ring of light
Dialogue: 0,0:02:21.69,0:02:24.10,Default,,0000,0000,0000,,caused by the gravitational\Nlensing of hot plasma
Dialogue: 0,0:02:24.13,0:02:25.96,Default,,0000,0000,0000,,zipping around the black hole.
Dialogue: 0,0:02:25.98,0:02:27.14,Default,,0000,0000,0000,,In other words,
Dialogue: 0,0:02:27.16,0:02:30.34,Default,,0000,0000,0000,,the black hole casts a shadow\Non this backdrop of bright material,
Dialogue: 0,0:02:30.36,0:02:32.20,Default,,0000,0000,0000,,carving out a sphere of darkness.
Dialogue: 0,0:02:32.23,0:02:35.56,Default,,0000,0000,0000,,This bright ring reveals\Nthe black hole's event horizon,
Dialogue: 0,0:02:35.59,0:02:37.99,Default,,0000,0000,0000,,where the gravitational pull\Nbecomes so great
Dialogue: 0,0:02:38.01,0:02:39.64,Default,,0000,0000,0000,,that not even light can escape.
Dialogue: 0,0:02:39.66,0:02:42.52,Default,,0000,0000,0000,,Einstein's equations predict\Nthe size and shape of this ring,
Dialogue: 0,0:02:42.55,0:02:45.75,Default,,0000,0000,0000,,so taking a picture of it\Nwouldn't only be really cool,
Dialogue: 0,0:02:45.78,0:02:48.40,Default,,0000,0000,0000,,it would also help to verify\Nthat these equations hold
Dialogue: 0,0:02:48.42,0:02:50.89,Default,,0000,0000,0000,,in the extreme conditions\Naround the black hole.
Dialogue: 0,0:02:50.91,0:02:53.47,Default,,0000,0000,0000,,However, this black hole\Nis so far away from us,
Dialogue: 0,0:02:53.49,0:02:56.59,Default,,0000,0000,0000,,that from Earth, this ring appears\Nincredibly small --
Dialogue: 0,0:02:56.61,0:03:00.20,Default,,0000,0000,0000,,the same size to us as an orange\Non the surface of the moon.
Dialogue: 0,0:03:00.76,0:03:03.58,Default,,0000,0000,0000,,That makes taking a picture of it\Nextremely difficult.
Dialogue: 0,0:03:04.64,0:03:05.95,Default,,0000,0000,0000,,Why is that?
Dialogue: 0,0:03:06.51,0:03:09.70,Default,,0000,0000,0000,,Well, it all comes down\Nto a simple equation.
Dialogue: 0,0:03:09.72,0:03:12.14,Default,,0000,0000,0000,,Due to a phenomenon called diffraction,
Dialogue: 0,0:03:12.16,0:03:13.52,Default,,0000,0000,0000,,there are fundamental limits
Dialogue: 0,0:03:13.54,0:03:16.21,Default,,0000,0000,0000,,to the smallest objects\Nthat we can possibly see.
Dialogue: 0,0:03:16.79,0:03:20.46,Default,,0000,0000,0000,,This governing equation says\Nthat in order to see smaller and smaller,
Dialogue: 0,0:03:20.48,0:03:23.07,Default,,0000,0000,0000,,we need to make our telescope\Nbigger and bigger.
Dialogue: 0,0:03:23.10,0:03:26.16,Default,,0000,0000,0000,,But even with the most powerful\Noptical telescopes here on Earth,
Dialogue: 0,0:03:26.19,0:03:28.61,Default,,0000,0000,0000,,we can't even get close\Nto the resolution necessary
Dialogue: 0,0:03:28.63,0:03:30.83,Default,,0000,0000,0000,,to image on the surface of the moon.
Dialogue: 0,0:03:30.85,0:03:34.47,Default,,0000,0000,0000,,In fact, here I show one of the highest\Nresolution images ever taken
Dialogue: 0,0:03:34.50,0:03:35.89,Default,,0000,0000,0000,,of the moon from Earth.
Dialogue: 0,0:03:35.92,0:03:38.47,Default,,0000,0000,0000,,It contains roughly 13,000 pixels,
Dialogue: 0,0:03:38.50,0:03:42.55,Default,,0000,0000,0000,,and yet each pixel would contain\Nover 1.5 million oranges.
Dialogue: 0,0:03:43.40,0:03:45.37,Default,,0000,0000,0000,,So how big of a telescope do we need
Dialogue: 0,0:03:45.39,0:03:48.16,Default,,0000,0000,0000,,in order to see an orange\Non the surface of the moon
Dialogue: 0,0:03:48.18,0:03:50.40,Default,,0000,0000,0000,,and, by extension, our black hole?
Dialogue: 0,0:03:50.42,0:03:52.76,Default,,0000,0000,0000,,Well, it turns out\Nthat by crunching the numbers,
Dialogue: 0,0:03:52.78,0:03:55.39,Default,,0000,0000,0000,,you can easily calculate\Nthat we would need a telescope
Dialogue: 0,0:03:55.42,0:03:56.81,Default,,0000,0000,0000,,the size of the entire Earth.
Dialogue: 0,0:03:56.83,0:03:57.86,Default,,0000,0000,0000,,(Laughter)
Dialogue: 0,0:03:57.88,0:04:00.00,Default,,0000,0000,0000,,If we could build\Nthis Earth-sized telescope,
Dialogue: 0,0:04:00.02,0:04:02.95,Default,,0000,0000,0000,,we could just start to make out\Nthat distinctive ring of light
Dialogue: 0,0:04:02.97,0:04:05.16,Default,,0000,0000,0000,,indicative of the black\Nhole's event horizon.
Dialogue: 0,0:04:05.18,0:04:08.10,Default,,0000,0000,0000,,Although this picture wouldn't contain\Nall the detail we see
Dialogue: 0,0:04:08.12,0:04:09.63,Default,,0000,0000,0000,,in computer graphic renderings,
Dialogue: 0,0:04:09.65,0:04:11.95,Default,,0000,0000,0000,,it would allow us to safely get\Nour first glimpse
Dialogue: 0,0:04:11.98,0:04:14.46,Default,,0000,0000,0000,,of the immediate environment\Naround a black hole.
Dialogue: 0,0:04:14.49,0:04:16.10,Default,,0000,0000,0000,,However, as you can imagine,
Dialogue: 0,0:04:16.12,0:04:19.75,Default,,0000,0000,0000,,building a single-dish telescope\Nthe size of the Earth is impossible.
Dialogue: 0,0:04:19.77,0:04:21.66,Default,,0000,0000,0000,,But in the famous words of Mick Jagger,
Dialogue: 0,0:04:21.68,0:04:23.47,Default,,0000,0000,0000,,"You can't always get what you want,
Dialogue: 0,0:04:23.50,0:04:25.68,Default,,0000,0000,0000,,but if you try sometimes,\Nyou just might find
Dialogue: 0,0:04:25.71,0:04:26.92,Default,,0000,0000,0000,,you get what you need."
Dialogue: 0,0:04:26.95,0:04:29.41,Default,,0000,0000,0000,,And by connecting telescopes\Nfrom around the world,
Dialogue: 0,0:04:29.44,0:04:32.97,Default,,0000,0000,0000,,an international collaboration\Ncalled the Event Horizon Telescope
Dialogue: 0,0:04:32.100,0:04:36.11,Default,,0000,0000,0000,,is creating a computational telescope\Nthe size of the Earth,
Dialogue: 0,0:04:36.13,0:04:37.67,Default,,0000,0000,0000,,capable of resolving structure
Dialogue: 0,0:04:37.69,0:04:39.89,Default,,0000,0000,0000,,on the scale of a black\Nhole's event horizon.
Dialogue: 0,0:04:39.92,0:04:43.30,Default,,0000,0000,0000,,This network of telescopes is scheduled\Nto take its very first picture
Dialogue: 0,0:04:43.33,0:04:45.14,Default,,0000,0000,0000,,of a black hole next year.
Dialogue: 0,0:04:45.16,0:04:48.50,Default,,0000,0000,0000,,Each telescope in the worldwide\Nnetwork works together.
Dialogue: 0,0:04:48.53,0:04:51.24,Default,,0000,0000,0000,,Linked through the precise timing\Nof atomic clocks,
Dialogue: 0,0:04:51.26,0:04:53.92,Default,,0000,0000,0000,,teams of researchers at each\Nof the sites freeze light
Dialogue: 0,0:04:53.94,0:04:56.91,Default,,0000,0000,0000,,by collecting thousands\Nof terabytes of data.
Dialogue: 0,0:04:56.93,0:05:01.95,Default,,0000,0000,0000,,This data is then processed in a lab\Nright here in Massachusetts.
Dialogue: 0,0:05:01.97,0:05:03.76,Default,,0000,0000,0000,,So how does this even work?
Dialogue: 0,0:05:03.79,0:05:07.19,Default,,0000,0000,0000,,Remember if we want to see the black hole\Nin the center of our galaxy,
Dialogue: 0,0:05:07.22,0:05:10.20,Default,,0000,0000,0000,,we need to build this impossibly large\NEarth-sized telescope?
Dialogue: 0,0:05:10.22,0:05:12.45,Default,,0000,0000,0000,,For just a second,\Nlet's pretend we could build
Dialogue: 0,0:05:12.48,0:05:14.32,Default,,0000,0000,0000,,a telescope the size of the Earth.
Dialogue: 0,0:05:14.34,0:05:16.80,Default,,0000,0000,0000,,This would be a little bit\Nlike turning the Earth
Dialogue: 0,0:05:16.82,0:05:18.57,Default,,0000,0000,0000,,into a giant spinning disco ball.
Dialogue: 0,0:05:18.59,0:05:20.79,Default,,0000,0000,0000,,Each individual mirror would collect light
Dialogue: 0,0:05:20.82,0:05:23.42,Default,,0000,0000,0000,,that we could then combine\Ntogether to make a picture.
Dialogue: 0,0:05:23.44,0:05:26.10,Default,,0000,0000,0000,,However, now let's say\Nwe remove most of those mirrors
Dialogue: 0,0:05:26.12,0:05:28.10,Default,,0000,0000,0000,,so only a few remained.
Dialogue: 0,0:05:28.12,0:05:30.100,Default,,0000,0000,0000,,We could still try to combine\Nthis information together,
Dialogue: 0,0:05:31.02,0:05:33.01,Default,,0000,0000,0000,,but now there are a lot of holes.
Dialogue: 0,0:05:33.04,0:05:37.41,Default,,0000,0000,0000,,These remaining mirrors represent\Nthe locations where we have telescopes.
Dialogue: 0,0:05:37.44,0:05:41.51,Default,,0000,0000,0000,,This is an incredibly small number\Nof measurements to make a picture from.
Dialogue: 0,0:05:41.54,0:05:45.38,Default,,0000,0000,0000,,But although we only collect light\Nat a few telescope locations,
Dialogue: 0,0:05:45.40,0:05:48.82,Default,,0000,0000,0000,,as the Earth rotates, we get to see\Nother new measurements.
Dialogue: 0,0:05:48.85,0:05:52.67,Default,,0000,0000,0000,,In other words, as the disco ball spins,\Nthose mirrors change locations
Dialogue: 0,0:05:52.69,0:05:55.59,Default,,0000,0000,0000,,and we get to observe\Ndifferent parts of the image.
Dialogue: 0,0:05:55.61,0:05:59.63,Default,,0000,0000,0000,,The imaging algorithms we develop\Nfill in the missing gaps of the disco ball
Dialogue: 0,0:05:59.66,0:06:02.69,Default,,0000,0000,0000,,in order to reconstruct\Nthe underlying black hole image.
Dialogue: 0,0:06:02.71,0:06:05.35,Default,,0000,0000,0000,,If we had telescopes located\Neverywhere on the globe --
Dialogue: 0,0:06:05.37,0:06:07.31,Default,,0000,0000,0000,,in other words, the entire disco ball --
Dialogue: 0,0:06:07.34,0:06:08.62,Default,,0000,0000,0000,,this would be trivial.
Dialogue: 0,0:06:08.64,0:06:11.97,Default,,0000,0000,0000,,However, we only see a few samples,\Nand for that reason,
Dialogue: 0,0:06:11.99,0:06:14.38,Default,,0000,0000,0000,,there are an infinite number\Nof possible images
Dialogue: 0,0:06:14.40,0:06:17.37,Default,,0000,0000,0000,,that are perfectly consistent\Nwith our telescope measurements.
Dialogue: 0,0:06:17.39,0:06:20.41,Default,,0000,0000,0000,,However, not all images are created equal.
Dialogue: 0,0:06:20.85,0:06:25.31,Default,,0000,0000,0000,,Some of those images look more like\Nwhat we think of as images than others.
Dialogue: 0,0:06:25.33,0:06:28.55,Default,,0000,0000,0000,,And so, my role in helping to take\Nthe first image of a black hole
Dialogue: 0,0:06:28.58,0:06:31.51,Default,,0000,0000,0000,,is to design algorithms that find\Nthe most reasonable image
Dialogue: 0,0:06:31.53,0:06:33.76,Default,,0000,0000,0000,,that also fits the telescope measurements.
Dialogue: 0,0:06:34.73,0:06:38.67,Default,,0000,0000,0000,,Just as a forensic sketch artist\Nuses limited descriptions
Dialogue: 0,0:06:38.69,0:06:42.21,Default,,0000,0000,0000,,to piece together a picture using\Ntheir knowledge of face structure,
Dialogue: 0,0:06:42.23,0:06:45.55,Default,,0000,0000,0000,,the imaging algorithms I develop\Nuse our limited telescope data
Dialogue: 0,0:06:45.57,0:06:49.89,Default,,0000,0000,0000,,to guide us to a picture that also\Nlooks like stuff in our universe.
Dialogue: 0,0:06:49.92,0:06:53.57,Default,,0000,0000,0000,,Using these algorithms,\Nwe're able to piece together pictures
Dialogue: 0,0:06:53.59,0:06:55.77,Default,,0000,0000,0000,,from this sparse, noisy data.
Dialogue: 0,0:06:55.80,0:07:00.32,Default,,0000,0000,0000,,So here I show a sample reconstruction\Ndone using simulated data,
Dialogue: 0,0:07:00.35,0:07:02.28,Default,,0000,0000,0000,,when we pretend to point our telescopes
Dialogue: 0,0:07:02.30,0:07:04.89,Default,,0000,0000,0000,,to the black hole\Nin the center of our galaxy.
Dialogue: 0,0:07:04.91,0:07:09.37,Default,,0000,0000,0000,,Although this is just a simulation,\Nreconstruction such as this give us hope
Dialogue: 0,0:07:09.39,0:07:12.85,Default,,0000,0000,0000,,that we'll soon be able to reliably take\Nthe first image of a black hole
Dialogue: 0,0:07:12.87,0:07:15.46,Default,,0000,0000,0000,,and from it, determine\Nthe size of its ring.
Dialogue: 0,0:07:16.12,0:07:19.32,Default,,0000,0000,0000,,Although I'd love to go on\Nabout all the details of this algorithm,
Dialogue: 0,0:07:19.34,0:07:21.52,Default,,0000,0000,0000,,luckily for you, I don't have the time.
Dialogue: 0,0:07:21.54,0:07:23.54,Default,,0000,0000,0000,,But I'd still like\Nto give you a brief idea
Dialogue: 0,0:07:23.56,0:07:25.87,Default,,0000,0000,0000,,of how we define\Nwhat our universe looks like,
Dialogue: 0,0:07:25.89,0:07:30.36,Default,,0000,0000,0000,,and how we use this to reconstruct\Nand verify our results.
Dialogue: 0,0:07:30.38,0:07:32.88,Default,,0000,0000,0000,,Since there are an infinite number\Nof possible images
Dialogue: 0,0:07:32.90,0:07:35.26,Default,,0000,0000,0000,,that perfectly explain\Nour telescope measurements,
Dialogue: 0,0:07:35.29,0:07:37.89,Default,,0000,0000,0000,,we have to choose\Nbetween them in some way.
Dialogue: 0,0:07:37.92,0:07:39.76,Default,,0000,0000,0000,,We do this by ranking the images
Dialogue: 0,0:07:39.78,0:07:42.61,Default,,0000,0000,0000,,based upon how likely they are\Nto be the black hole image,
Dialogue: 0,0:07:42.64,0:07:45.12,Default,,0000,0000,0000,,and then choosing the one\Nthat's most likely.
Dialogue: 0,0:07:45.14,0:07:47.34,Default,,0000,0000,0000,,So what do I mean by this exactly?
Dialogue: 0,0:07:47.86,0:07:49.84,Default,,0000,0000,0000,,Let's say we were trying to make a model
Dialogue: 0,0:07:49.86,0:07:53.05,Default,,0000,0000,0000,,that told us how likely an image\Nwere to appear on Facebook.
Dialogue: 0,0:07:53.07,0:07:54.77,Default,,0000,0000,0000,,We'd probably want the model to say
Dialogue: 0,0:07:54.80,0:07:58.35,Default,,0000,0000,0000,,it's pretty unlikely that someone\Nwould post this noise image on the left,
Dialogue: 0,0:07:58.38,0:08:00.80,Default,,0000,0000,0000,,and pretty likely that someone\Nwould post a selfie
Dialogue: 0,0:08:00.82,0:08:02.15,Default,,0000,0000,0000,,like this one on the right.
Dialogue: 0,0:08:02.18,0:08:03.82,Default,,0000,0000,0000,,The image in the middle is blurry,
Dialogue: 0,0:08:03.84,0:08:06.48,Default,,0000,0000,0000,,so even though it's more likely\Nwe'd see it on Facebook
Dialogue: 0,0:08:06.50,0:08:07.86,Default,,0000,0000,0000,,compared to the noise image,
Dialogue: 0,0:08:07.89,0:08:10.85,Default,,0000,0000,0000,,it's probably less likely we'd see it\Ncompared to the selfie.
Dialogue: 0,0:08:10.87,0:08:13.16,Default,,0000,0000,0000,,But when it comes to images\Nfrom the black hole,
Dialogue: 0,0:08:13.19,0:08:16.69,Default,,0000,0000,0000,,we're posed with a real conundrum:\Nwe've never seen a black hole before.
Dialogue: 0,0:08:16.71,0:08:19.00,Default,,0000,0000,0000,,In that case, what is a likely\Nblack hole image,
Dialogue: 0,0:08:19.03,0:08:21.96,Default,,0000,0000,0000,,and what should we assume\Nabout the structure of black holes?
Dialogue: 0,0:08:21.99,0:08:24.62,Default,,0000,0000,0000,,We could try to use images\Nfrom simulations we've done,
Dialogue: 0,0:08:24.64,0:08:27.18,Default,,0000,0000,0000,,like the image of the black hole\Nfrom "Interstellar,"
Dialogue: 0,0:08:27.20,0:08:30.14,Default,,0000,0000,0000,,but if we did this,\Nit could cause some serious problems.
Dialogue: 0,0:08:30.16,0:08:33.54,Default,,0000,0000,0000,,What would happen\Nif Einstein's theories didn't hold?
Dialogue: 0,0:08:33.56,0:08:37.53,Default,,0000,0000,0000,,We'd still want to reconstruct\Nan accurate picture of what was going on.
Dialogue: 0,0:08:37.55,0:08:40.92,Default,,0000,0000,0000,,If we bake Einstein's equations\Ntoo much into our algorithms,
Dialogue: 0,0:08:40.94,0:08:43.70,Default,,0000,0000,0000,,we'll just end up seeing\Nwhat we expect to see.
Dialogue: 0,0:08:43.72,0:08:46.00,Default,,0000,0000,0000,,In other words,\Nwe want to leave the option open
Dialogue: 0,0:08:46.02,0:08:48.95,Default,,0000,0000,0000,,for there being a giant elephant\Nat the center of our galaxy.
Dialogue: 0,0:08:48.97,0:08:50.03,Default,,0000,0000,0000,,(Laughter)
Dialogue: 0,0:08:50.05,0:08:53.04,Default,,0000,0000,0000,,Different types of images have\Nvery distinct features.
Dialogue: 0,0:08:53.06,0:08:56.61,Default,,0000,0000,0000,,We can easily tell the difference\Nbetween black hole simulation images
Dialogue: 0,0:08:56.64,0:08:58.91,Default,,0000,0000,0000,,and images we take\Nevery day here on Earth.
Dialogue: 0,0:08:58.94,0:09:02.04,Default,,0000,0000,0000,,We need a way to tell our algorithms\Nwhat images look like
Dialogue: 0,0:09:02.06,0:09:05.31,Default,,0000,0000,0000,,without imposing one type\Nof image's features too much.
Dialogue: 0,0:09:05.86,0:09:07.76,Default,,0000,0000,0000,,One way we can try to get around this
Dialogue: 0,0:09:07.78,0:09:10.84,Default,,0000,0000,0000,,is by imposing the features\Nof different kinds of images
Dialogue: 0,0:09:10.87,0:09:14.100,Default,,0000,0000,0000,,and seeing how the type of image we assume\Naffects our reconstructions.
Dialogue: 0,0:09:15.71,0:09:19.20,Default,,0000,0000,0000,,If all images' types produce\Na very similar-looking image,
Dialogue: 0,0:09:19.23,0:09:21.28,Default,,0000,0000,0000,,then we can start to become more confident
Dialogue: 0,0:09:21.31,0:09:25.48,Default,,0000,0000,0000,,that the image assumptions we're making\Nare not biasing this picture that much.
Dialogue: 0,0:09:25.50,0:09:28.50,Default,,0000,0000,0000,,This is a little bit like\Ngiving the same description
Dialogue: 0,0:09:28.52,0:09:31.52,Default,,0000,0000,0000,,to three different sketch artists\Nfrom all around the world.
Dialogue: 0,0:09:31.54,0:09:34.40,Default,,0000,0000,0000,,If they all produce\Na very similar-looking face,
Dialogue: 0,0:09:34.42,0:09:36.22,Default,,0000,0000,0000,,then we can start to become confident
Dialogue: 0,0:09:36.24,0:09:39.86,Default,,0000,0000,0000,,that they're not imposing their own\Ncultural biases on the drawings.
Dialogue: 0,0:09:39.88,0:09:43.20,Default,,0000,0000,0000,,One way we can try to impose\Ndifferent image features
Dialogue: 0,0:09:43.22,0:09:45.66,Default,,0000,0000,0000,,is by using pieces of existing images.
Dialogue: 0,0:09:46.21,0:09:48.37,Default,,0000,0000,0000,,So we take a large collection of images,
Dialogue: 0,0:09:48.40,0:09:51.12,Default,,0000,0000,0000,,and we break them down\Ninto their little image patches.
Dialogue: 0,0:09:51.14,0:09:55.42,Default,,0000,0000,0000,,We then can treat each image patch\Na little bit like pieces of a puzzle.
Dialogue: 0,0:09:55.45,0:09:59.73,Default,,0000,0000,0000,,And we use commonly seen puzzle pieces\Nto piece together an image
Dialogue: 0,0:09:59.75,0:10:02.20,Default,,0000,0000,0000,,that also fits our telescope measurements.
Dialogue: 0,0:10:03.04,0:10:06.78,Default,,0000,0000,0000,,Different types of images have\Nvery distinctive sets of puzzle pieces.
Dialogue: 0,0:10:06.81,0:10:09.61,Default,,0000,0000,0000,,So what happens when we take the same data
Dialogue: 0,0:10:09.64,0:10:13.77,Default,,0000,0000,0000,,but we use different sets of puzzle pieces\Nto reconstruct the image?
Dialogue: 0,0:10:13.79,0:10:18.56,Default,,0000,0000,0000,,Let's first start with black hole\Nimage simulation puzzle pieces.
Dialogue: 0,0:10:18.58,0:10:20.17,Default,,0000,0000,0000,,OK, this looks reasonable.
Dialogue: 0,0:10:20.20,0:10:22.89,Default,,0000,0000,0000,,This looks like what we expect\Na black hole to look like.
Dialogue: 0,0:10:22.91,0:10:24.11,Default,,0000,0000,0000,,But did we just get it
Dialogue: 0,0:10:24.13,0:10:27.44,Default,,0000,0000,0000,,because we just fed it little pieces\Nof black hole simulation images?
Dialogue: 0,0:10:27.47,0:10:29.35,Default,,0000,0000,0000,,Let's try another set of puzzle pieces
Dialogue: 0,0:10:29.37,0:10:31.88,Default,,0000,0000,0000,,from astronomical, non-black hole objects.
Dialogue: 0,0:10:32.91,0:10:35.04,Default,,0000,0000,0000,,OK, we get a similar-looking image.
Dialogue: 0,0:10:35.06,0:10:37.30,Default,,0000,0000,0000,,And then how about pieces\Nfrom everyday images,
Dialogue: 0,0:10:37.32,0:10:40.11,Default,,0000,0000,0000,,like the images you take\Nwith your own personal camera?
Dialogue: 0,0:10:41.31,0:10:43.43,Default,,0000,0000,0000,,Great, we see the same image.
Dialogue: 0,0:10:43.45,0:10:46.82,Default,,0000,0000,0000,,When we get the same image\Nfrom all different sets of puzzle pieces,
Dialogue: 0,0:10:46.84,0:10:48.89,Default,,0000,0000,0000,,then we can start to become more confident
Dialogue: 0,0:10:48.91,0:10:50.88,Default,,0000,0000,0000,,that the image assumptions we're making
Dialogue: 0,0:10:50.90,0:10:53.82,Default,,0000,0000,0000,,aren't biasing the final\Nimage we get too much.
Dialogue: 0,0:10:53.85,0:10:57.10,Default,,0000,0000,0000,,Another thing we can do is take\Nthe same set of puzzle pieces,
Dialogue: 0,0:10:57.12,0:10:59.61,Default,,0000,0000,0000,,such as the ones derived\Nfrom everyday images,
Dialogue: 0,0:10:59.64,0:11:03.24,Default,,0000,0000,0000,,and use them to reconstruct\Nmany different kinds of source images.
Dialogue: 0,0:11:03.26,0:11:04.53,Default,,0000,0000,0000,,So in our simulations,
Dialogue: 0,0:11:04.56,0:11:08.33,Default,,0000,0000,0000,,we pretend a black hole looks like\Nastronomical non-black hole objects,
Dialogue: 0,0:11:08.35,0:11:12.20,Default,,0000,0000,0000,,as well as everyday images like\Nthe elephant in the center of our galaxy.
Dialogue: 0,0:11:12.23,0:11:15.40,Default,,0000,0000,0000,,When the results of our algorithms\Non the bottom look very similar
Dialogue: 0,0:11:15.42,0:11:17.52,Default,,0000,0000,0000,,to the simulation's truth image on top,
Dialogue: 0,0:11:17.54,0:11:20.88,Default,,0000,0000,0000,,then we can start to become\Nmore confident in our algorithms.
Dialogue: 0,0:11:20.91,0:11:22.78,Default,,0000,0000,0000,,And I really want to emphasize here
Dialogue: 0,0:11:22.80,0:11:24.73,Default,,0000,0000,0000,,that all of these pictures were created
Dialogue: 0,0:11:24.76,0:11:27.69,Default,,0000,0000,0000,,by piecing together little pieces\Nof everyday photographs,
Dialogue: 0,0:11:27.72,0:11:29.93,Default,,0000,0000,0000,,like you'd take with your own\Npersonal camera.
Dialogue: 0,0:11:29.96,0:11:33.23,Default,,0000,0000,0000,,So an image of a black hole\Nwe've never seen before
Dialogue: 0,0:11:33.26,0:11:37.20,Default,,0000,0000,0000,,may eventually be created by piecing\Ntogether pictures we see all the time
Dialogue: 0,0:11:37.22,0:11:39.97,Default,,0000,0000,0000,,of people, buildings,\Ntrees, cats and dogs.
Dialogue: 0,0:11:39.99,0:11:42.64,Default,,0000,0000,0000,,Imaging ideas like this\Nwill make it possible for us
Dialogue: 0,0:11:42.66,0:11:45.28,Default,,0000,0000,0000,,to take our very first pictures\Nof a black hole,
Dialogue: 0,0:11:45.30,0:11:47.75,Default,,0000,0000,0000,,and hopefully, verify\Nthose famous theories
Dialogue: 0,0:11:47.78,0:11:50.20,Default,,0000,0000,0000,,on which scientists rely on a daily basis.
Dialogue: 0,0:11:50.22,0:11:52.83,Default,,0000,0000,0000,,But of course, getting\Nimaging ideas like this working
Dialogue: 0,0:11:52.85,0:11:56.18,Default,,0000,0000,0000,,would never have been possible\Nwithout the amazing team of researchers
Dialogue: 0,0:11:56.20,0:11:58.09,Default,,0000,0000,0000,,that I have the privilege to work with.
Dialogue: 0,0:11:58.11,0:11:59.27,Default,,0000,0000,0000,,It still amazes me
Dialogue: 0,0:11:59.30,0:12:02.65,Default,,0000,0000,0000,,that although I began this project\Nwith no background in astrophysics,
Dialogue: 0,0:12:02.67,0:12:05.29,Default,,0000,0000,0000,,what we have achieved\Nthrough this unique collaboration
Dialogue: 0,0:12:05.32,0:12:08.07,Default,,0000,0000,0000,,could result in the very first\Nimages of a black hole.
Dialogue: 0,0:12:08.10,0:12:10.80,Default,,0000,0000,0000,,But big projects like\Nthe Event Horizon Telescope
Dialogue: 0,0:12:10.82,0:12:13.63,Default,,0000,0000,0000,,are successful due to all\Nthe interdisciplinary expertise
Dialogue: 0,0:12:13.66,0:12:15.45,Default,,0000,0000,0000,,different people bring to the table.
Dialogue: 0,0:12:15.47,0:12:17.18,Default,,0000,0000,0000,,We're a melting pot of astronomers,
Dialogue: 0,0:12:17.20,0:12:19.43,Default,,0000,0000,0000,,physicists, mathematicians and engineers.
Dialogue: 0,0:12:19.46,0:12:22.01,Default,,0000,0000,0000,,This is what will make it soon possible
Dialogue: 0,0:12:22.04,0:12:24.89,Default,,0000,0000,0000,,to achieve something\Nonce thought impossible.
Dialogue: 0,0:12:24.91,0:12:27.17,Default,,0000,0000,0000,,I'd like to encourage all of you to go out
Dialogue: 0,0:12:27.19,0:12:29.29,Default,,0000,0000,0000,,and help push the boundaries of science,
Dialogue: 0,0:12:29.31,0:12:33.21,Default,,0000,0000,0000,,even if it may at first seem\Nas mysterious to you as a black hole.
Dialogue: 0,0:12:33.24,0:12:34.41,Default,,0000,0000,0000,,Thank you.
Dialogue: 0,0:12:34.44,0:12:36.83,Default,,0000,0000,0000,,(Applause)