There's Something at the Edge of the Solar System

Planet Nine, Konstantin Batygin tells Lex Fridman, is a hypothesized body beyond Neptune: about five Earth masses, on an orbit of roughly 10,000 years, slightly tilted and slightly eccentric. No telescope has photographed it. The case is gravitational. Batygin, a planetary astrophysicist at Caltech, spends the opening of the conversation turning that ghost into numbers, then placing it inside a solar system that is much larger than the eight planets on a lunchbox.

He splits the system in two. The inner planets — Mercury, Venus, Earth, Mars — sit at about 40 percent, 70 percent, 100 percent, and 160 percent of Earth's orbital distance, and they are, to leading order, "heavily overgrown asteroids." Jupiter is 316 Earth masses; Saturn is 90, at about ten times Earth's separation from the Sun; Uranus and Neptune sit near 20 and 30 astronomical units. For a long time that was where the massive solar system was thought to end. The last thirty years added a second, icier debris field beyond Neptune: the Kuiper Belt. Pluto belongs to it. It is simply a large member of that population, not a ninth planet.

Pluto's demotion is a forty-year lesson in shrinking ignorance. It was found in 1930 because astronomers were hunting a seven-Earth-mass world beyond Neptune; they treated the discovery as that world, then immediately revised the mass down because the telescope could not resolve a disk. Over the next four decades the mass kept falling until Pluto was about 500 times lighter than Earth. Its surface area, Batygin notes, is almost exactly that of Russia — and Russia is big, but it is not a planet. Early in the twentieth century, Pluto stood for what we did not know about the edge. Planet Nine, he agrees, is the modern version of that same ignorance.

The first fingerprint is orbital clustering. Beyond a critical period of about 4,000 years, the longest-period Kuiper Belt objects point roughly the same way and tilt by about 20 degrees. Unstable bodies that Neptune still kicks around are scattered; the stable ones, which Neptune does not mess with, show the pattern. Five years before this conversation, Batygin and Mike Brown read that clustering as a gravitational one-way sign: a distant planet shepherding the orbits. One statistical approach, using a single survey's sky coverage, gives no confident answer. Their preferred method builds a map of every historical discovery — if something was found at a given sky position and brightness, someone looked there and could see objects that bright or brighter. Done that way, the false-alarm probability of the clustering is about 0.4 percent.

Discovering a Kuiper Belt object is easy in theory and hard in practice. Take a long-exposure snapshot, wait a night, do it again, wait another night, do it a third time. Galaxy stars stay put; solar-system ice moves, the same parallax that makes roadside trees pass faster than clouds. The bodies are giant icicles reflecting sunlight. Light has to travel something like forty times the Earth–Sun distance and bounce back, on the order of an hour, which is why they are so faint. The third night is there to catch acceleration: nearer asteroids can fake a small shift for two nights and then peel away.

What surprised Batygin once he started observing with Brown is how little three nights actually tell you. For the first year you know, with confidence, only where the object is on the sky and how far away it is. Almost all of that three-day motion is Earth's motion — Earth is the car, the object is the tree. An orbit has six parameters, so you come back a year later for another three nights, and in reality you need many more points to pin it down. Discoveries get a license plate: the first four digits are the year it appeared in the data, then a code for sky position. One object central to the Planet Nine story is VP113, nicknamed Biden because Joe Biden was vice president when it was found. About four thousand Kuiper Belt objects are catalogued now; the number could become millions.

The Vera Rubin Observatory, then still under construction and delayed somewhat by the pandemic, is built to scan the night sky automatically every day. Batygin says the survey is efficient enough that it might increase the census of the distant Kuiper Belt — the population he cares about — by a factor of a hundred. It might also simply image Planet Nine. The first detection of something hundreds of times farther from the Sun than Earth, reflecting sunlight, is where the story concludes. After that you study it.

If Neptune's orbit is one, the main Kuiper Belt sits near 1.3 to 1.5. Beyond that is the scattered disc: long, highly elliptical orbits that hug Neptune at perihelion and reach semi-major axes of about a thousand astronomical units. Keep going to 10,000–100,000 AU and you reach the Oort Cloud, a distinct, nearly spherical shell of ice whose outer edge is set by how far the Sun's gravity still wins. That edge falls roughly halfway to the nearest star. All long-period comets come from there. Bodies sitting some 30,000 AU out feel the galactic tide; their orbits slowly elongate until ice starts to sublimate near the Sun, and we see a comet. Batygin's guess is that collisions in the Oort Cloud over the age of the solar system have literally numbered zero. The physical radii of the objects are negligible compared with the gaps between them.

Rogue interstellar visitors do join the party. Unlike solar-system bodies on ellipses, they travel on hyperbolas: they come in, say hello, and leave. The first was 'Oumuamua; the second was comet Borisov, found by an astronomer named Borisov. Their existence is unsurprising. Neptune is constantly ejecting Kuiper Belt objects into interstellar space; other planetary systems presumably leak ice the same way. Primitive life on those lonely bodies, though, Batygin puts at essentially zero. The interstellar medium is a hostile radiation environment. You do not even have to go that far: leave Earth's magnetic field and you are already in trouble. Panspermia is not entirely implausible, he says, but you have to twist a lot of parameters, and we still do not know how life starts — which makes survival in deep space a second-order question.

Are there aliens? He calls that a boring question, because the answer is of course there are. Planets orbit almost every star; life as an Earth-only accident is a statistical impossibility. The hard problems are prior: we do not even have a definition of life that would let you decide whether a rock that moved since yesterday is alive. Intelligent life he finds entirely plausible, given that Earth already has it — dolphins, pigs, humans. An Earth-like temperate architecture, though, is not a common outcome of planet formation, perhaps something like a one-percent effect. Large numbers still save you. Even a one-in-ten-million draw among 10^12 stars in the galaxy is a win.

Earth is uncommon in a specific way: it never acquired a thick hydrogen-helium atmosphere. Stars form with a disk of gas and dust. Giant planets, and many exoplanets, grab hydrogen and helium from that disk. Earth took about 100 million years to form, after the gas disk had already dissipated, and so missed the train. That lateness is why the sky is visible. An opaque atmosphere would be a greenhouse at the bottom; Venus, only a little closer to the Sun, is hundreds of degrees Celsius because of its thick air. The sci-fi picture of landing on random Earth-like worlds around other stars is, Batygin says, totally not true.

The most stable result of exoplanet searches is that most stars are encircled by a few-Earth-mass planets on orbits of days to weeks. That region of our solar system is astonishingly hollow. The Sun is not a special star, so the natural assumption is that the same formation physics ran here and then something erased the compact inner system. Batygin's calculations say such a population would not survive Jupiter's primordial wandering. Giant planets migrate: they carve a hole in the gas disk and change orbit by gravitational interaction with it. Jupiter plus Saturn makes the problem messy — you have to evolve the disk, Jupiter, Saturn, and their coupling together — but the common outcome is simple. Jupiter forms and its orbit shrinks; Saturn forms, catches up; then both move outward, all within the last million years of the primordial disk.

If those calculations are right, the inward-outward pattern pushes the inner few-Earth-mass planets onto the Sun. What remains is a mass-depleted remnant disk of only a couple of Earth masses, an annulus thin enough to flirt with being a ring. Over about 100 million years, collisions in that debris grow Earth, the Moon, and the rest of the terrestrial planets. The solar system we see is the aftermath of that violence. Jupiter–Saturn analogs themselves are rare: they are found around only about 10 percent of Sun-like stars, and a negligible fraction of the galaxy's numerous M dwarfs host giant planets. Why giants are rare remains, he says, one of the big unanswered problems.

Two bodies orbiting a star can be written on paper; you do not need a simulation. Three gravitationally interacting bodies have a chaotic solution. No matter how many runs you do, the answer loses meaning after a while. That jump from analytically solvable to unpredictable, Batygin says, should immediately tell you that engineering a calculation which forms the entire solar system from scratch and still yields a predictive answer is futile. The equations are not missing. Gravity, magnetohydrodynamics — Maxwell plus Navier–Stokes — and quantum mechanics for opacities are in hand. Put the same evolution on twice with an infinitesimal change in initial conditions, and you get a different solar system. Planetary systems are diverse for that reason: a disk of a given mass around a given star does not imply our architecture, only a spectrum of outcomes of which some percentage resemble it.

He once had no intuition for where the Schrödinger equation came from. The professor's advice was not to worry and to calculate hydrogen energy levels. Years later, preparing a lecture on waves in self-gravitating disks, he imagined a disk as infinitely many concentric rings talking to each other gravitationally — a late-1700s Hamiltonian problem. In the continuum limit, the governing equation became Schrödinger's. He is careful: this does not unify quantum mechanics and gravity. There is nothing quantum about the Schrödinger equation itself. It is a wave equation; the quantum part is the interpretation. Waves, he says, go all the way down. Astrophysical disks are everywhere, from rings around potato-shaped Kuiper Belt objects the size of Los Angeles, to Saturn, to the solar nebula, to black-hole accretion disks and galaxies, because stuff rotates. On timescales between a few orbits and a disk's whole life, Schrödinger becomes a calculational tool that a week-long hydro simulation cannot reach.

How do you get from a hydrogen-helium disk to the first solids? Batygin reaches for the desert. Tumbleweeds roll because the wind rolls them, and if you watch carefully you sometimes see a family: a large one in front and smaller ones hiding in its wake, like ducks crossing a street. Cyclists in the Tour de France ride within ten centimeters of one another for the same reason — to cut collective air resistance, not because they are best friends. Centimeter-sized solids in a protoplanetary disk do exactly this. They spontaneously clump so that trailing grains sit in the wake of leading ones, minimizing the dust's aerodynamic interaction with the gas. That streaming instability is energetically favorable, so the clouds grow until self-gravity overwhelms internal support and they collapse. Those collapses are the first building blocks of planets, and the origin of the big asteroids. Qualitatively it is the same script as star formation: a cooling cloud contracts until gravity beats thermal pressure.

People do simulate this, and it is one of the active corners of planet-formation theory. The cartoon of a single instability hides a richly dynamical process, but the qualitative sequence is the same as the birth of stars: a cloud of hydrogen and helium sits in space, cools, and at some point contracts until gravity overtakes thermal pressure and collapses under its own weight. You get a little baby solar system. Asked whether a full high-resolution retelling of Jupiter clearing space is important, Batygin says the ideas matter and the method matters more: design each numerical experiment to test a specific aspect of the evolution, rather than pouring cycles into one gigantic run.

Could we one day simulate the full history of the solar system? "Yes, and it will be useless." Simulation is compression, so it must lose detail; simulate every detail and you have built an actual universe, not a model. Even setting information theory aside, an excruciatingly complete history would be no different from observing the real thing. Theoretical understanding, which is what he is after, comes from reducing complexity to a mechanism you can quantify. Cool visualizations are not extra understanding. Weather is the everyday proof: the equations of the atmosphere are known, and two-week forecasts are still meaningless, because typical weather systems have a horizon of a few days. Chaos grows without any injected randomness. The right move is to ask specific questions, compress the answers into theory, and then generalize to other stars and galaxies.

After a long detour through formation and simulation, Fridman pulls the conversation back. Planet Nine remains unseen. If it were not there, the clustered long-period orbits would disperse into a mutually symmetric spray. Its gravity instead holds them in a state roughly anti-aligned with its own orbit, oscillating on a timescale of about a billion years. That is only the easiest picture to visualize. Another line of evidence is Kuiper Belt objects tilted by as much as 90 degrees, almost out of the plane of the planets. Planet-formation simulations have never produced those without some extrinsic gravitational force. Planet Nine generates them readily, through a different gravitational effect. Batygin counts about five such independent tricks. Individually each is maybe a one-sigma result, not a reason to jump up and down. The attraction of the model is that one hypothesis explains a multitude of puzzles at once.

The outer solar system is also a museum of the system's childhood. Small meteorites are melted and differentiated — iron sank inward — even though they are too small to have melted from the heat of accretion. The solar nebula was polluted with aluminum-26, a short-lived radioactive isotope that dumps heat as it decays. The Sun cannot make enough of it; supernovae can. So the solar system did not form in isolation. It formed in a giant cloud of thousands of other stars, some exploding, leaving isotopic traces we still read. That fact invites panspermia, and Batygin does not love the question: it relocates the origin of life instead of solving it. His Caltech colleague Joe Kirschvink is something like 100 percent certain that life started on Mars and rode to Earth on Martian meteorites. Origin-of-life work, Batygin says, is fascinating and not physics.

Conclusive evidence does not require a probe. The moment Planet Nine is imaged from Earth, it is done: something reflecting sunlight, known to be hundreds of times as far from the Sun as Earth. In principle there could be more than one such object; there is no evidence for a flock of them. The Oort Cloud itself probably holds five to seven Earth masses of material. The historical analogy is Neptune. Uranus, found in 1781, refused to follow its predicted orbit; in the mid-1800s Urbain Le Verrier calculated where a more distant planet had to sit — and then they found it. Planet Nine is different in one respect. Batygin and Brown can predict the orbit and mass relatively well. They cannot say where the planet is on that orbit, because the relevant Kuiper Belt objects have not completed even one of their 4,000-year laps. Fridman jokes that exploration is just getting good at waiting. Batygin will not take waiting without a second skill: "You can't just be good at waiting. You got to know how to chill."

A recent paper with Brown asked what Planet Nine does to objects in the inner Oort Cloud. Standard simulations had treated the trans-Neptunian solar system as sourced entirely from the inside: Neptune scatters the Kuiper Belt outward, Planet Nine aligns orbits, and you evolve for the age of the system. During the pandemic they did something they regularly do — ask how to disprove themselves, and what the simulations were missing. The missing piece, obvious in retrospect, was isolation. The solar system formed in a cluster of thousands of stars and, in that phase, built an almost spherical population of icy debris at a few thousand AU, maybe closer.

Without Planet Nine that population is dormant: the objects slowly orbit and nothing interesting ever happens. With Planet Nine, the planet can gravitationally grab some of them and re-inject them into the distant solar system. Numerical experiments showed the process not only works but follows roughly the same pathway as objects Neptune is scattering out. The picture becomes a two-way river: some ice coming out, some moving in. If the observed sample is not entirely stuff that left the inner solar system, but also stuff that was fished back from farther out, the best-fit Planet Nine orbit becomes slightly more eccentric. A hypothesis that revises itself is a hypothesis that is still alive. Planet Nine, in this telling, is a natural bridge between a sleeping reservoir of ice and the bodies astronomers are starting to observe directly.

Some physicists have suggested Planet Nine is not a planet but a primordial black hole. Ordinary black holes form when a star above about 1.4 solar masses runs out of fusion fuel and collapses. Primordial black holes would have condensed from the Big Bang across a whole mass spectrum, from asteroids up through five or ten Earth masses. The small ones evaporate over cosmic time; five Earth masses is large enough to remain. Batygin's calculations cannot rule this out, because they return only orbit and mass. A five-Earth-mass cup, a five-Earth-mass hedgehog, or a black hole would all do. The gravity of a black hole is no different from the gravity of a planet. If the Sun became a black hole tomorrow it would go dark, but Earth would keep orbiting. The idea that black holes suck everything in is, he says, a sci-fi notion. Mass is mass.

The person most associated with hoping it is a black hole, in this conversation, is Ed Witten, who floated a fleet of probes accelerated by radiation pressure, launched isotropically, with the gravitational field inferred from trajectory deflections. Batygin thinks the idea is cool and, after talking with JPL engineers, much harder than it looks. At that precision, chaotic solar flares drive unpredictable radiation-pressure gradients, and you do not know where every asteroid is. Possible, not trivial. He would rather the object be a planet. On a logarithmic scale of complexity, planets sit between stars and insects; an insect is far more complicated than a star, which is "somebody playing with a spring." Black holes are simpler still — macroscopic particles. The galaxy is excellent at making few-Earth-mass worlds, the most common exoplanet type, and the solar system has no analog: it jumps from one Earth mass to poorly understood Uranus and Neptune. If Planet Nine is a planet, going there is the closest window on that dominant class. New Horizons at Pluto is the cautionary tale against assuming a distant iceball is boring: at about 40 kelvin, water ice behaves like metal, but carbon monoxide ice flows into glaciers nobody would have imagined.

The broader hardware story is already shifting. Flagship NASA missions still have enormous development cycles; CubeSats offer a few million dollars and a couple of years, conceivably a PhD that designs, flies, and publishes within five or six years — not fully executed yet, but on this decade's horizon. SpaceX and the commercial sector have taken over an industry that used to be a government, Cold War project. That drop in cost is a boost to the human condition; Starlink is the invoice. Camping in the Mojave, Batygin watched a string of satellites appear and vanish. He is not signed on to Mars as a safe haven. "Mars sucks." You can preview the experience by camping in that same desert. He grants the magnetism of exploration and the nonlinear payoff of Apollo, and he is glad people want to go. He is not moving there permanently. The leading-order engineering problem, he thinks, is not abandoning Earth.

Would human society have developed as it did if Earth had an opaque atmosphere and we could not see the stars? Drastically different, Batygin says, if it developed at all. An opaque atmosphere is hot at the bottom; we would be different beings. Early human exploration used the stars as a navigation tool, and astronomy then shows up in essentially every scientific development until the twentieth century. Newton developed calculus in part to explain Kepler's laws. Replacing a fire god in a chariot with a mechanistic night sky abandoned a non-secular view of nature and, he argues, put humanity on the course of the last 400 years. We would be a very different species without astronomy.

He extends the point past telescopes. People ask why anyone should fund paleontology or the mating patterns of penguins. There is a tremendous underappreciation for the usefulness of useless knowledge — a phrase he borrows from a little book by the founder of the Institute for Advanced Study. Maxwell was not sitting in the 1800s hoping to invent microphones so two people could have this conversation. Electronics on the table still run on his equations. If you are a purely pragmatic person who does not care about the human condition, the accounting is still favorable: the "useless" things have created way more capital than the useful ones. Failure is part of exploration, not a verdict that the pursuit was worthless. A good way to have good ideas, he recalls from Linus Pauling's generation, is to have a lot of ideas. There is no science police. Better to propose interesting things and be wrong some of the time than to make incremental progress your whole life.

Could 'Oumuamua be space junk from a distant civilization? You cannot immediately say no, Batygin answers. Anything on the night sky could in principle be junk, and capture of debris — natural or otherwise — would have been easy while the solar system still sat in its birth cluster. Look at the facts, though: a weird elongated shape, and an extra acceleration. The most daring non-alien resolution, from his friends Daryl Seligman and Greg Laughlin, is that it is a piece of hydrogen ice. Giant molecular clouds of hydrogen and helium float between the stars and eventually become stellar nurseries; in starless clouds, their calculations say, you can condense hydrogen icebergs. Sublimation would thin the body like a bar of soap, explaining the elongation. Escaping hydrogen would provide a jet, explaining the kick, and would be invisible, explaining the missing coma. Batygin likes the idea because it is a new class of astrophysical object and because it is falsifiable. When Vera Rubin comes online, many more interstellar visitors should appear. If they lack the hydrogen-ice signatures, the hypothesis dies. "Aliens" is fundamentally hard to say no to.

Fridman argues that a successful civilization would produce orders of magnitude more junk than queens, so SETI should look for garbage. Batygin is unconvinced on energetic grounds. Humans have launched thousands of satellites; almost none will ever leave the solar system — Voyager 1 and 2, maybe Pioneer, and Elon Musk's Tesla Roadster will not. Interstellar travel has a cost that is hard to overcome. We also tend to search for whatever we ourselves are doing this decade. Radio leakage has diminished tremendously in fifty years; computation has not. Something like a few percent of the electrical grid went to mining bitcoin in the previous year, and the energy cost of an AI ecosystem is becoming a tangible bottleneck. Space-junk SETI, he thinks, is still soaked in 1950s sci-fi in which everyone would be flying to the Moon by 2020. He has also never seen an alien civilization. Trying to imagine one or two orders of magnitude ahead is almost limiting: ask people in 1870 what they wanted, and they wanted faster buggies.

He still grants the magnetism of the wild-idea hat. Some Jules Verne novels did a phenomenal job of predicting the future, and science fiction does not only forecast — it plants seeds that create the future. Batygin is a sucker for sci-fi on his runs, including the implausible kind. Numerical experiments already dominate about 70 percent of his work: he writes "terrible code" that sets up a simulation, interacting with the computer as something that generates insights you cannot get from analytical mathematics, the shift Stanisław Ulam was pointing at when computers stopped being human arithmetic. Going to Mars, or sending probes to Planet Nine, will also be harder than imagined. We do it anyway. Fridman would be among the early thousands and is fine dying there. Batygin is not. Death he calls fundamentally boring, and he would rather not do it alone. Northern New Mexico — one of the most beautiful places he has been — would do, returning to nature with family.

Asked for the greatest song of all time, Batygin jokes with Blink-182's "What's My Age Again," then gets serious. Songwriting-wise the Beatles came close; "In My Life," written by a 24-year-old, is one of the most introspective pieces of music he knows. As a Pink Floyd fan he puts *The Dark Side of the Moon* near the pinnacle of what can be created — "Time" is a great song — and *The Wall* can transport you into another place. Greatness, for him, has a uniqueness test. Blink-182, Sum 41, and a thousand lesser-known pop-punk bands shared a sparkling production; Pink Floyd does not sound like anyone else. If it disappeared, the world would miss it. A lot of music produced today feels like H&M clothes: not bad, worn three times, thrown away. That we are still talking about Pink Floyd in 2021 is itself the point.

He would not call himself a great musician, but music is essential to everything he does. If he stops playing — if he is traveling — he notably loses creativity in every other part of life. He does not view playing music as a separate endeavor from doing science. It is part of the same creative process, and he would say it is the thing that enables the science: without it, the science would suck even more than it already does. Inspiration is actually rare; most of the time everyone is just doing what needs to get done. When the spark arrives, if the thing in your hands can be put on hold, put it on hold. Answering email in that window is a disservice. He keeps a social-media presence but mostly stays off it, because he does not enjoy the mental cycles it robs. He plays guitar most days, including made-up songs about the cat while putting the kids to sleep. Before a gig the band sharpens up, not to make money, but for the satisfaction of doing something well.

Batygin was born in Russia, moved to Japan in 1994, and arrived in the United States at the very end of 1999, almost fourteen. Elementary school was a regular Japanese public school; he was the non-Japanese person in the class. In a stretch of Russian, he told Fridman about first meeting capitalism in 1991–92: Pepsi appeared, then Coke, and as a child he was confused that one had stolen the other's product. (Pepsi had reached the Soviet Union first, in a deal that at one point involved Soviet submarines.) McDonald's was a six-hour line, a packed room, and a Big Mac whose sesame-seed bun seemed like culinary art — he still wonders how they inject the seeds. The early 1990s were rough; his parents were near the bottom of the financial spectrum. Trouble since then does not register the same way. The immigrant advantage, as he names it, is having already done the humiliating full environment swap — isolated, made fun of — so the next one, middle school in the U.S., was easier. He would not trade that for anything.

He is not a career coach and does not have it all figured out. What he does see is a perpetual illusion of a template for success: high-school students pouring themselves into straight A's and a stuffed CV so it looks impressive. That, he thinks, is a poor way to optimize a life. Do the thing that fills life with passion and interest, perpetually. A straight-A student is impressive and also somewhat boring. Fridman offers the split he saw in Max Tegmark at MIT: spend a small fraction of effort so the CV looks adequate, then put most of your time into the work you actually care about, letting Planet-Nine-like side effects slowly feed the résumé. Batygin agrees that the baseline matters, then pushes further. Grades are not a tangible product. You cannot take your A and have it live a separate life. Code can. Music can, if somebody listens. Doing things that can be separated from you is exceptionally attractive — and, he thinks, more impressive to others than another checkbox.

He wonders whether academia's current gauntlet — bachelor's, PhD, one or two or three postdocs — is the only shape the profession can have. Faraday did not know algebra; he drew diagrams of magnetic fields, and Faraday's law came entirely from intuition. The advisor of the advisor of Batygin's undergraduate advisor did not have a PhD. The internet makes it easier than ever to show the work itself, whether or not a certificate exists. Grades and certificates, Batygin and Fridman agree, mean less in fields where you can put code, videos, or papers into the world and let them walk around without you.

Asked for the meaning of life, he offers 42, then a real answer: human beings are fundamentally stochastic, and also stubbornly interested in expanding their understanding of the world and in building tools that enable that understanding. Even people who claim to be anti-science still ask how things work. He is in the privileged position of having that curiosity as a job. As the economy shifts, creative work is taking a larger share — not yet dominant, but the tide is visible if you compare it with a factory repeating the same motion. Life, he says, is becoming more fun. He cannot wait to chill: wait for Kuiper Belt objects to finish an orbit, wait for a photograph of Planet Nine, and in the meantime keep the guitar, the calculations, and the people. Confidence, he notes on the way out, is essential to being a rock star.