WEBVTT 00:00.000 --> 00:04.160 Thank you to Novium, the team behind  the hoverpen, for supporting PBS. 00:04.160 --> 00:08.240 We’ve been looking for messages from the stars  ever since Frank Drake pointed the Green Bank   00:08.240 --> 00:14.560 radio telescope at Tau Ceti and Epsilon Eridany  65 years ago. He saw nothing that couldn’t be   00:14.560 --> 00:19.760 explained by natural causes. Nor have the much  more extensive SETI surveys conducted since. So,   00:19.760 --> 00:25.200 maybe there are no alien signals to see. Or  maybe we need to update how we search for   00:25.200 --> 00:31.840 them. And maybe when we do that we’ll find that  the aliens have been talking to us all along,   00:31.840 --> 00:36.880 and the messages are already sitting  in our astronomical data archives,   00:36.880 --> 00:48.800 or will be glaringly obvious is the new generation  of giant surveys that are just now starting. 00:36.880 --> 00:43.280 it lights up after sun exposure—even in  the depths of space. Now on to the episode. 00:36.880 --> 00:43.280 it lights up after sun exposure—even in  the depths of space. Now on to the episode. 00:42.238 --> 00:47.758 Although it’s not exclusively the case, most major  SETI programs have been based on ideas of what WE,   00:47.758 --> 00:52.878 humanity could transmit—if not right now,  then in the relatively near future. Well,   00:52.878 --> 00:57.838 we’re now IN the relatively near future  compared to the first SETI search of 1960.   00:57.838 --> 01:04.078 So how do we shift our thinking based on  what we can now do, and what we now know? 01:04.078 --> 01:08.958 As a launching point, let’s look at a new  paper by Ben Zuckerman. Dr. Zuckerman has   01:08.958 --> 01:14.478 been contributing to and commenting on SETI  research for many decades. The new paper pulls   01:14.478 --> 01:20.398 together his thoughts on how that search needs to  evolve given some updated thinking. Zuckerman’s   01:20.398 --> 01:25.758 proposed strategy updates are centered  around making better guesses about what a   01:25.758 --> 01:31.758 technological civilization might do, or at least  rejecting some previous unfounded assumptions. 01:31.758 --> 01:37.518 There are three main, interconnected points to  consider: transmission technology, how to target   01:37.518 --> 01:44.638 those transmissions; and tying it all together,  the energy available for those transmissions. 01:44.638 --> 01:49.278 Starting with transmission technology  itself. In the 60s we were just getting   01:49.278 --> 01:55.278 really good at building giant radio antennae.  And there were advantages compared to, say,   01:55.278 --> 02:01.598 visible or infrared light transmission. The tech  is simpler—just currents running along chunky   02:01.598 --> 02:07.758 wires—versus advanced materials and micro-scale  engineering needed for visible and infrared light   02:07.758 --> 02:13.038 detectors and transmitters. Radio frequencies also  travel relatively unimpeded through the dustier   02:13.038 --> 02:19.198 regions of interstellar space. So yeah, obviously  aliens would be doing interstellar ham radio. 02:19.198 --> 02:26.558 But the big challenge with radio is that it’s  hard to send a tight beam. For an EM wave,   02:26.558 --> 02:33.038 the tightest possible beam has a spread that’s  equal to the wavelength times distance traveled   02:33.038 --> 02:38.798 divided by the size of the transmitting aperture.  An alien 100 light years away who wanted to flood   02:38.798 --> 02:45.598 earth’s entire orbit with a radio signal would  need a radio array with a 1000 km baseline. 02:45.598 --> 02:52.398 We can and have built the 1000km-scale  baseline radio arrays by now. But even   02:52.398 --> 02:56.718 with this level of collimation, the power  of the signal is still enormously spread   02:56.718 --> 03:02.078 out—over an area 140 billion times the  surface area of the Earth. It would take   03:02.078 --> 03:06.238 an enormous amount of power to make such a  diffuse signal stand out above the galactic   03:06.238 --> 03:12.798 radio background. Early SETI folk guessed  that an energy-limited civilization might   03:12.798 --> 03:18.078 solve this issue by channeling their limited  power into a very narrow radio frequency band.   03:18.078 --> 03:22.478 That spike in the radio spectrum would then  stand out above the galactic background. 03:22.478 --> 03:27.278 Frank Drake even guessed what frequency  aliens might use—the so-called water hole.   03:27.278 --> 03:32.078 It’s a 300 Megahertz span where the  galactic radio background is weakest,   03:32.078 --> 03:36.958 and right between an oxygen and hydrogen  emission lines that Drake guessed formed   03:36.958 --> 03:43.518 a natural framing of a galactic narrow-band  ham network. This sort of narrow-band radio   03:43.518 --> 03:48.158 search strategy drove many of the biggest  past and present SETI programs. And it   03:48.158 --> 03:55.598 seems a reasonable to try this. Except this  strategy hasn’t yielded anything yet. If the   03:55.598 --> 04:01.598 aliens are trying to talk to us, then they’re  not doing it how we think they’re doing it. 04:01.598 --> 04:07.198 According to Zuckerman, one of the keys to  rethinking our strategy is to take seriously   04:07.198 --> 04:12.238 the fact that any technological alien civilization  we’re likely to encounter will have been around   04:12.238 --> 04:18.558 for much longer than we have. That shifts  our thinking on each of our three points:   04:18.558 --> 04:24.398 the transmission tech, the targeting,  and the energy limits. I’ll come back   04:24.398 --> 04:29.278 to how we may need to shift those. But first,  let’s see if we can convince ourselves that   04:29.278 --> 04:35.118 any aliens that we can plausibly notice are  almost certainly more advanced than we are. 04:35.118 --> 04:40.478 Now it comes down to the average longevity of  a technological civilization. Let’s say that   04:40.478 --> 04:45.838 they tend to last for, I dunno, 10,000 years  after becoming able to send noticeable signals   04:45.838 --> 04:52.158 or probes or whatever. We, humanity, are in our  first century of this period. That’s the youngest   04:52.158 --> 05:00.878 1% of galactic civilizations. So, 99% would be  more advanced than us, and 90% are around 1000   05:00.878 --> 05:06.638 years into their technological phase. So also  much more advanced. Even if the civilization   05:06.638 --> 05:11.918 survival time is only 1000 years, most will still  be ahead of us. But if that survival time is much   05:11.918 --> 05:18.158 less than 1000 years then a good fraction of  civilizations are of comparable tech level to   05:18.158 --> 05:25.118 ours. But in that case then the window of their  noticeability is tiny. In order to notice them,   05:25.118 --> 05:32.558 our own 100 years of existence would have  to occur within their 100 or so years plus   05:32.558 --> 05:36.798 the travel time for their signals. And  that overlap window is a vanishingly   05:36.798 --> 05:42.638 small 1-in-100-million fraction of the 10  billion years of the Milky Way’s lifespan. 05:42.638 --> 05:48.398 Of course, the Milky Way may have many of  these razor thin expanding bubbles containing   05:48.398 --> 05:54.158 the desperate hails from newly emergent but soon  to be doomed intelligent species. But the hails   05:54.158 --> 05:59.198 themselves are doomed because no species lasts  long enough to exist during the passage of such   05:59.198 --> 06:05.278 a bubble. And that could be true, but if so the  case is hopeless and we should focus on finding   06:05.278 --> 06:11.038 the signals of species that last longer.  Or at least, so also advises Ben Zuckerman. 06:11.038 --> 06:17.118 So, what will an advanced civilization do  differently? As a relatively more advanced   06:17.118 --> 06:21.118 civilization ourselves—compared to  the 60s and only in very specific   06:21.118 --> 06:27.278 respects—what can we advise regarding  our key factors—tech, targets, energy. 06:27.278 --> 06:32.718 Regarding transmission technology: we know  how to do other things besides radio. With   06:32.718 --> 06:37.678 modern laser technology we can transmit highly  collimated visible and infrared beams of light.   06:37.678 --> 06:42.638 We haven’t built ones big enough to be space  telephones, but in principle we know that it’s   06:42.638 --> 06:48.878 possible and can track a path towards  this in a way that we couldn’t in 1960. 06:48.878 --> 06:54.718 And that helps with… Targeting. Remember that  the spread of a collimated beam is proportional   06:54.718 --> 07:00.158 to wavelength. Visible-light is several tens of  thousands of times shorter wavelength than radio,   07:00.158 --> 07:07.678 with proportionally higher collimation. The  beam collimation we could achieve with our   07:07.678 --> 07:12.718 1000km radio array can be achieved with a  1meter laser aperture or array. Or, if we   07:12.718 --> 07:19.278 say to hell with it and build a 1000-km laser  array—probably an interferometer in space—then   07:19.278 --> 07:25.918 we can focus it to the scale of a single  planet rather than the entire planet’s orbit. 07:25.918 --> 07:31.278 But is it even possible to target a single  planet 100s of light years away. Well maybe,   07:31.278 --> 07:37.918 yes. In the 60s we had no idea whether there  were planets around other stars at all. Now   07:37.918 --> 07:42.238 we know that essentially all stars have them,  and that Earth-sized planets around Sun-like   07:42.238 --> 07:48.078 stars are common. The Kepler mission told us that  this is a statistical necessity, even though we   07:48.078 --> 07:54.718 haven’t actually found exact Earth-analogs.  But we know how to find them in principle. 07:54.718 --> 07:58.158 Most importantly, as emphasized by  Zuckerman, we will ultimately be   07:58.158 --> 08:03.678 able to take pictures of exo-Earths. Using  space-based coronographs and star-shades   08:03.678 --> 08:08.718 and optical or infrared interferometry,  one day we’ll be resolving the surfaces   08:08.718 --> 08:13.758 of Earth-like planets within hundreds of  light years to actually see continents and   08:13.758 --> 08:19.278 oceans and alien forests even city lights.  Spectral information gathered this way,   08:19.278 --> 08:24.158 especially in the infrared, can even reveal  the chemical signatures of an active biosphere. 08:24.158 --> 08:28.638 Even back in the 1980s were gearing up  to build the Terrestrial Planet Finder,   08:28.638 --> 08:32.398 which would have been able to image  exo-Earths within 50 light years.   08:32.398 --> 08:37.998 It was canned due to politics and JWST  costing too much, but now the Habitable   08:37.998 --> 08:43.438 Worlds Observatory will pick up the mission  and get us our first pics in the 2030s. So,   08:43.438 --> 08:48.158 if we’re this close to finding out where the  local habitable and even inhabited worlds are,   08:48.158 --> 08:51.838 surely a civilization far more advanced  than us will have already done so. 08:51.838 --> 08:57.198 And this is one of Zuckerman’s key points.  An advanced species in the local region   08:57.198 --> 09:03.598 that cares at all about locating neighbors  will already know that we are here. And by   09:03.598 --> 09:10.478 “we” I mean life. Our technological signals  are only 100 or so light years out, so only   09:10.478 --> 09:16.238 those within that range know that some primates  figured out how to move electrons around wires. 09:16.798 --> 09:22.078 So now, not only does a technological neighbor  have the ability to send highly collimated   09:22.078 --> 09:29.038 signals, but they also know where to send them.  To the life-bearing worlds in their vicinity. 09:29.038 --> 09:34.398 Which brings us back to the energy issue.  An advanced civilization probably isn’t even   09:34.398 --> 09:41.118 energy limited in the way that the original SETI  programs assumed. We may be able to assume that   09:41.118 --> 09:46.798 signalers will be pumping more energy into  more efficient beams, enabling them to do   09:46.798 --> 09:52.798 various things—like transmitting messages across  many frequencies at once, rather than in single,   09:52.798 --> 09:58.878 narrow bands. And they could also get  their message to us from much further away. 09:58.878 --> 10:03.038 OK, let’s bring together what we’ve  learned. If we follow Zuckerman’s   10:03.038 --> 10:08.398 reasoning, the most likely signals are not  narrow-band radio with wide angular spread,   10:08.398 --> 10:14.558 but rather highly targeted beams that  could appear at any or all frequencies,   10:14.558 --> 10:20.078 more likely within optical or infrared  wavebands. And these could come potentially   10:20.078 --> 10:25.598 from much further away than is possible  with energy-limited radio broadcasts. 10:25.598 --> 10:31.838 He suggests a goal of observing all Sun-like  stars within around 650 Light years,   10:31.838 --> 10:38.158 with the range being chosen somewhat arbitrarily,  but feels like an OK definition of “local” in   10:38.158 --> 10:45.118 our 100,000 light year radius galaxy. In the 650  light year range, there are approximately 60,000   10:45.118 --> 10:51.038 Earth-like planets orbiting Sun-like stars based  on our Kepler projections. Zuckerman whittles   10:51.038 --> 10:57.838 this down to a very crude and maybe conservative  estimate of 600 that resemble Earth in various   10:57.838 --> 11:03.038 other important factors. Of course we don’t know  how close to Earth-like a civilization-spawning   11:03.038 --> 11:08.158 planet needs to be, but erring on the side  of conservative is the right way to start. 11:08.158 --> 11:13.518 OK, so let’s say we locate our target planets  to monitor. What are we really looking for? 11:13.518 --> 11:17.678 Beyond the manner of transmission,  there are also many questions about   11:17.678 --> 11:23.358 what would flag a signal as technological in  origin. Many people have thought about this,   11:23.358 --> 11:29.358 and maybe we cover it in another episode. In  general, we have a good idea of what natural   11:29.358 --> 11:32.718 sources of electromagnetic radiation  look like. They tend to consist of   11:32.718 --> 11:38.718 a combination of a fairly well understood  family of spectra—from hot gas and plasma,   11:38.718 --> 11:43.998 from charged particles moving in magnetic fields,  from electrons hopping between atomic energy   11:43.998 --> 11:50.798 levels. Frequency spikes in unexpected parts  of the spectrum and other unexpected intensity   11:50.798 --> 11:57.198 patterns can signal a non-natural origin. Or we  might expect a strange time-dependence—patterns   11:57.198 --> 12:02.398 can also be encoded in how the signal changes  over time. Sudden and-or periodic changes   12:02.398 --> 12:08.158 in signal strength that may reveal encoded  information not possible for natural sources. 12:08.158 --> 12:13.118 If the signal really does come from another  planet, and if the signal has a distinct   12:13.118 --> 12:18.958 frequency structure, then that should at  least be clear. It’ll move back and forth   12:18.958 --> 12:24.158 slightly in frequency due to Doppler  shift as the planet orbits its star. 12:24.158 --> 12:29.118 The truth is we just don’t know what a  technological transmission might look   12:29.118 --> 12:35.678 like and we can’t rely on aliens thinking  how we think they should think. We’ve been   12:35.678 --> 12:41.598 teased by weird stuff in the past. Like with the  regular pulsation of the first pulsar or radio   12:41.598 --> 12:47.918 spike in the Wow signal—but so far all have been  explained as combinations of natural processes.   12:47.918 --> 12:53.918 The hope is that if our astronomical observations  get smart enough and broad enough, we’ll know a   12:53.918 --> 12:58.798 technological signal when we see it. But for now,  what we need to do is look for anomalies. And   12:58.798 --> 13:04.078 modern SETI is becoming more and mor about anomaly  detection rather than targeted searches for the   13:04.078 --> 13:10.798 digits of pi encoded in a doppler-varying water  hole spike or whatever we imagine “they” will do. 13:10.798 --> 13:15.518 While it’s good to keep an open mind, the problem  with looking for generic anomalies is that the   13:15.518 --> 13:22.398 possibility space is enormous. Back when Frank  Drake did his first SETI search it wasn’t possible   13:22.398 --> 13:28.318 to just look for anything weird coming from  anywhere. Now though? Now it’s not just possible,   13:28.318 --> 13:34.238 we’re kind of already doing it. For example,  the European Space Organization’s HARPS program   13:34.238 --> 13:40.718 uses a telescope in the Chilean Andes to look for  exoplanets around nearly 3000 stars by measuring   13:40.718 --> 13:46.398 the tiny Doppler shift in a star’s spectrum  caused by the similarly tiny wobble in the   13:46.398 --> 13:52.638 star’s motion induced by orbiting planets. But in  a study published last year, Benjamin Fields and   13:52.638 --> 13:58.318 Jason Goodman showed that this same data could  be sensitive to laser communication from these   13:58.318 --> 14:04.078 orbiting worlds. They didn’t find anything … but  it’s a good proof of concept that technosignatures   14:04.078 --> 14:10.718 may be hidden in data we’re already taking.  And maybe in the data we’ve already taken. 14:10.718 --> 14:16.878 This whole idea of piggybacking SETI on existing  astronomy surveys is being called commensal SETI,   14:16.878 --> 14:21.518 and it’s something that Zuckerman highlights  also. The sorts of signals that he expects,   14:21.518 --> 14:26.958 and many others besides, may be observable  in our current and very near future surveys.   14:26.958 --> 14:33.678 Because those surveys are insane. Our ability  to put extremely sensitive, gigantic cameras   14:33.678 --> 14:38.398 on ever-larger telescopes, and to put some of  those telescopes in space has opened up a new   14:38.398 --> 14:45.118 era in astronomy in general—an era of big  data, of all-sky, time domain surveys. 14:45.118 --> 14:50.878 For example, the Rubin Observatory, also in the  Chilean Andes, is about to start its 10-year   14:50.878 --> 14:57.358 survey in which it images the entire southern  sky every 3 days. The Euclid and Grace Roman   14:57.358 --> 15:03.278 telescopes bring this wide-field capability  to the crystal clarity of space, the SKA is   15:03.278 --> 15:07.758 old-fashioned radio, but will be so powerful that  it feels like it’s getting close to the “advanced   15:07.758 --> 15:13.358 civilization” radio telescopes that people like  Frank Drake were imagining back in the 60s. 15:13.358 --> 15:19.438 These facilities are just beginning or about to  begin their work of downloading the universe with   15:19.438 --> 15:25.918 a breadth, depth, and resolution both in space and  time that has never been seen before, and that few   15:25.918 --> 15:31.358 imagined back in the 60’s. And while the main  science drivers are to characterize the natural   15:31.358 --> 15:37.758 universe, they’re also going to be better SETI  programs than any dedicated SETI program to date. 15:37.758 --> 15:43.438 In a paper published in July last year, Eleanor  Gallay and team outlined what the possibilities   15:43.438 --> 15:48.958 might be with the Rubin Observatory. The  data flow from this survey is gigantic,   15:48.958 --> 15:53.118 and there aren’t enough graduate students  in the observable universe to look at every   15:53.118 --> 15:58.718 pixel for every anomaly. But systems are  now in place to automatically monitor the   15:58.718 --> 16:04.638 20 terabytes that come from the telescope  every night to spot interesting changes.   16:04.638 --> 16:10.958 This is done by teams known as data brokers, who  gain quick access to the data and run their clever   16:10.958 --> 16:15.678 algorithms—many of them leaning heavily into  machine learning—to look for different types of   16:15.678 --> 16:20.958 “events”. They’re really focused on natural stuff,  like supernovae and asteroids. But unnatural   16:20.958 --> 16:25.918 variation can also be picked up, especially  with some light patching of the algorithms.   16:25.918 --> 16:30.718 The Gallay team show how the existing  alert structure might also be co-opted   16:30.718 --> 16:37.278 to find a variety of technosignatures, and  that Rubin-LSST is going to enable a scale   16:37.278 --> 16:44.558 of search that may allow us to spot even the  most excruciatingly rare technosignatures. 16:44.558 --> 16:50.158 Commensal SETI isn’t restricted to the  optical-IR-focused surveys that Zuckerman’s   16:50.158 --> 16:56.238 paper emphasizes. For example, it really could be  that the aliens use radio. The square kilometer   16:56.238 --> 17:00.718 array is currently being assembled in South  Africa and Australia and will ramp up its   17:00.718 --> 17:07.118 sensitivity towards its completion date in the  early 2030s. The SKA will take as much data every   17:07.118 --> 17:13.678 second as Rubin does every night. That means the  challenges are of the same family as for Rubin,   17:13.678 --> 17:20.158 but even crazier. How do you pick out potential  signals without having to store every byte?   17:20.158 --> 17:25.838 Several research teams have been working hard  towards building the right alert algorithms. If   17:25.838 --> 17:31.198 aliens do send radio signals, the SKA  is our best change yet to catch them. 17:31.198 --> 17:35.918 Our eyes on the universe are clearer  and wider than ever before. They’ll   17:35.918 --> 17:39.278 show us the beginning and end of  the universe and the lives and   17:39.278 --> 17:44.798 deaths of stars and galaxies and strange  natural phenomena we’ve never imagined.   17:44.798 --> 17:49.278 But a few of us will be watching that  feed from the universe for something   17:49.278 --> 17:58.078 else. Someone else. Signals that we’re not the  only intelligence in this corner of space time. 17:49.278 --> 17:53.758 Father’s day gift or just want one for yourself,  there’s links in the description below and if you   17:49.278 --> 17:59.438 use the code PBS there’s 15% off all hoverpens  for the next 72 hours and 10% off after that.