[HN Gopher] Wasp-193B, a giant planet with a density similar to ...
       ___________________________________________________________________
        
       Wasp-193B, a giant planet with a density similar to that of cotton
       candy
        
       Author : geox
       Score  : 86 points
       Date   : 2024-05-14 13:20 UTC (9 hours ago)
        
 (HTM) web link (www.exotic.uliege.be)
 (TXT) w3m dump (www.exotic.uliege.be)
        
       | dberg wrote:
       | Here I was thinking this was a new LLM with 193B parameters.
        
         | spacemanspiff01 wrote:
         | Maybe it is, somewhere I'd the depths of the void there is a
         | giant planet sized llm, with cotton candy neurons.
        
           | dhosek wrote:
           | In an infinite universe, everything exists, it just takes
           | forever to reach it.
        
             | BenjiWiebe wrote:
             | In an infinite string of ones, does a two exist?
        
         | htrp wrote:
         | Matryoshka Brain....
        
         | qaq wrote:
         | same :)
        
         | world2vec wrote:
         | Actually a pretty cool name (and model size) for a locally run
         | LLM.
        
         | msikora wrote:
         | Yep, same!
        
         | dylan604 wrote:
         | I recognized the WASP prefix. It's a pretty cool program[0].
         | 
         | https://en.wikipedia.org/wiki/Wide_Angle_Search_for_Planets
         | 
         | The software to detect the transits seems like a fun challenge.
         | Locating each star in each of the images taken, then comparing
         | their brightness in each one. Seems simple enough until you
         | then realize your literally dealing with astronomically large
         | numbers.
        
           | JackFr wrote:
           | I love SPECULOOS as a network of robotic telescopes.
           | 
           | https://en.wikipedia.org/wiki/Speculaas
        
       | throwup238 wrote:
       | That raises some very important questions: what is the cotton-
       | candy/human crush depth equivalent? How much buoyancy would I
       | have?
       | 
       | Like, can I swim in this cotton candy planet? How deep can I get
       | in the cotton candy clouds or will I just sink and get candy
       | crushed?
        
         | lostlogin wrote:
         | > candy crushed
         | 
         | Someone was going to make the candy crush joke.
        
       | paulkrush wrote:
       | Maybe earth could look like this if we put enough satellites in
       | orbit.
        
         | nanomonkey wrote:
         | I'm imagining an orbital trailer park for nanorobots, with
         | wispy power cords and network cables trailing all over the
         | place as everyone participates in one big LAN party.
        
       | onion2k wrote:
       | _Exactly how a planet can inflate so much is a question that no
       | existing theory of planetary formation can yet answer._
       | 
       | Is there a reason to limit theories to natural formation? An
       | alien race could have built it as a hydrogen storage tank or
       | something.
        
         | jdiff wrote:
         | Because of all the things humanity has learned about the
         | universe, none of them has turned out to be aliens or evidence
         | of aliens. You can theorize it all you want, but so far every
         | time the universe has just turned out to be a big, strange
         | place.
        
           | RHSman2 wrote:
           | Aren't we the evidence for aliens?
        
         | klyrs wrote:
         | Until evidence to the contrary is found, it is not reasonable
         | to assume that aliens are going around making planets. I say
         | no, we have no reason to make such an assumption. If we think
         | we've done a good job enumerating all of the ways that a planet
         | can naturally form, that would provide reason to suspect that
         | we've missed something. But pinning it on aliens capable of
         | planet-scale engineering is quite the jump.
         | 
         | For example, rather than a hydrogen storage tank, perhaps this
         | was once a primordial rogue planet, a frozen ball of hydrogen
         | hurtling through space. If it was then captured by a star's
         | gravity well, and gently warmed by its newfound environment, it
         | could expand into the anomaly we see today. Perhaps in another
         | hundred years, we'll see that it's shedding its gas into a ring
         | around the star. This sounds improbable, but I find it less
         | incredible than immediately jumping to intelligent life.
        
       | netbioserror wrote:
       | The idea of a planet 1.5x the size of Jupiter, but 1/7th the
       | mass...is very unintuitive from a gravitational perspective.
       | Maybe it's spinning so ungodly fast that centrifugal forces are
       | keeping it puffed out?
        
         | __MatrixMan__ wrote:
         | I wonder if it's physically possible to start with a situation
         | that you're describing and end up with a toroidal planet.
        
           | itishappy wrote:
           | Yes and no. It's possible, but it's not a stable
           | configuration, so it's unlikely to form naturally.
           | 
           | https://www.aleph.se/andart/archives/2014/02/torusearth.html
        
         | rangerelf wrote:
         | I thought there wasn't such thing as a "centrifugal force" :-|
        
           | ordu wrote:
           | In an inertial frame of reference there is not. In a rotating
           | frame of reference there funny forces like centrifugal and a
           | Coriolis force.
        
       | NKosmatos wrote:
       | The density mentioned (0.059 grams per cubic centimeter) is the
       | overall density of the planet. Most probably it has a much much
       | denser core, solid or thick super hot fluid, like all gas giants.
       | It doesn't mean that the whole planet has the same cotton candy
       | density (0.05 grams per cubic centimeter) from the top layers
       | down to the core ;-)
       | 
       | On the funny side of this, what would happen if we kept adding
       | cotton candies at an empty/isolated place in space? After some
       | critical mass, the cotton candies would collapse under their own
       | weight/mass and a new asteroid would be created. If we kept
       | adding even more candies a further collapse would create a new
       | planet :-)
       | 
       | Imagine stacking/accumulating different materials. A planet made
       | just of bananas, a planet made of only water, a planet from
       | rice...
       | 
       | /note to self - check if there is such a planet simulator
       | available
        
         | keithly wrote:
         | Or a planet made of a mole of moles... https://what-
         | if.xkcd.com/4/
        
           | piker wrote:
           | > You might notice that we're ignoring the pockets of space
           | between the moles. In a moment, you'll see why.
           | 
           | Lol!
        
         | TrainedMonkey wrote:
         | You might enjoy a mole of moles - https://what-if.xkcd.com/4/
        
         | hbrav wrote:
         | As you added more and more cotton candy you would begin to
         | compress the centre under the pressure, and you would get some
         | heating from this. I think there's a good change you would melt
         | and cotton candy and eventually it would re-solidify into an
         | enormous boiled sweet core, with a diffuse envelope of cotton
         | candy around it.
        
           | prox wrote:
           | Would love to see a simulation of that haha
        
             | harry_ord wrote:
             | A project for future odd ball people/groups
        
           | datameta wrote:
           | They outer layer would be an excellent insulator.
        
         | _boffin_ wrote:
         | I'm sensing a new XKCD--how much cotton candy would it take
         | until it becomes an asteroid and what would the core be after.
        
         | roywiggins wrote:
         | "Carl Sagan's Cosmos: 'The Meat Planet'"
         | 
         | https://www.youtube.com/watch?v=ZP7K9SycELA
        
         | tuanx5 wrote:
         | Sounds like you should visit Magrathea for a designer planet!
        
         | euroderf wrote:
         | This sounds like a task for one of the Culture's GSVs.
        
         | MisterBastahrd wrote:
         | What's the minimum material density necessary for a collapse
         | such as this? Is it possible to have a planet with a core of
         | solid oxygen, for example?
        
           | pixl97 wrote:
           | I mean, when we talk about the Jovian planet we believe it
           | has a core of solid metal hydrogen, the only real problem
           | getting an oxygen planet is there is much more
           | hydrogen/helium in the universe so that's what it's going to
           | be made out of.
        
         | Rumudiez wrote:
         | the Katamari game series plays on this idea of creating themed
         | planets by rolling up everyday objects on Earth
        
         | jon_richards wrote:
         | What fascinates me is that if you had _enough_ cotton candy, it
         | would be a black hole. I don 't mean it would collapse into a
         | black hole, I mean even at a uniform density of 0.05 grams per
         | cubic centimeter, it would _already be_ a black hole.
         | 
         | We're so used to surface area scaling at r^2 and volume scaling
         | at r^3 and the weird effects that can have (never scale up an
         | exothermic reaction), but the maximum amount of matter that can
         | exist in a volume without creating a black hole scales by r^1.
         | Even with cotton candy, that r^3 is going to out-scale r^1 at
         | some point.
         | 
         | It's an interesting thought experiment for algorithm complexity
         | as well. Can you actually retrieve an element from an array in
         | constant time regardless of the size of the array? In the
         | extreme case, the drive containing the array must have r
         | proportional to the size of the array to avoid becoming a black
         | hole. Assuming the query and element travel along the drive at
         | the speed of light, retrieving the element still takes time
         | proportional to the size of the array.
        
           | lainga wrote:
           | I have struggled to find an established name for this value.
           | I thought it would be something called the "Schwarzschild
           | density", but no luck. Famously, the [someone's
           | name??]-density of the observable universe just happens to be
           | very close to the density that would tip the universe over
           | into being a very large black hole
        
             | littlestymaar wrote:
             | But given that the universe is expanding, if we're close to
             | this density doesn't that mean that the density used to be
             | higher than the threshold at some point?
        
               | pixl97 wrote:
               | This is why when talking about the big bang physicists
               | talk about the inflation field at the start of it. The
               | primitives in the pre microsecond universe have to be
               | different from a black hole otherwise there's no way to
               | un black hole yourself.
        
             | jon_richards wrote:
             | Wouldn't it just be the inverse of the Schwarzschild
             | radius? https://en.wikipedia.org/wiki/Schwarzschild_radius
        
           | dTal wrote:
           | Worth noting that even setting aside weird black hole
           | physics, in the real world the amount of storage available
           | given some information density and some maximum constant
           | access latency is bound in r^3 by the speed of light, and
           | therefore array access is not _really_ constant time but
           | O([?]n). This isn 't some abstract ivory tower thing but the
           | way computers actually work - due to cache hierarchies you
           | will find that working with a 100 byte array is much faster
           | than working with a 100 megabyte array, which is much faster
           | than working with a 100 terabyte array. Every time the data
           | gets larger, it gets further away...
        
       | Finnucane wrote:
       | A giant planet actually made of cotton candy would be way more
       | interesting.
        
       | fnordpiglet wrote:
       | Clearly it's a hollow sphere world.
        
       | nico wrote:
       | I thought this was an announcement of an LLM model called Wasp
       | that had 193B parameters
       | 
       | Was happy when I clicked the link and it was about a newly
       | discovered planet 1200 light years from Earth
        
       | chimpanzee wrote:
       | The article doesn't mention this, but the planet is often
       | speculated to be the home planet of the Klowns.
       | 
       | https://youtu.be/plUXguATsTQ
       | 
       | https://en.wikipedia.org/wiki/Killer_Klowns_from_Outer_Space
        
       | abecedarius wrote:
       | How do you get an object of 44 Earth masses with the average
       | density of styrofoam?
       | 
       | The obvious starting point I guess is a core of very light
       | elements, and a really really extended atmosphere. But an
       | atmosphere thins out exponentially, more steeply the higher the
       | gravity. So for an atmosphere to help much, those 40ish Earth
       | masses of hypothesized core need to be spread out a whole lot
       | already. It seems like a difficult planetary engineering problem
       | -- spin it up almost to bursting?
       | 
       | So taking that to be too silly, the core must be only a small
       | fraction of the mass: a big gas cloud with a nugget in the
       | center. I'm not sure that could work, but it's what I'd try to
       | work out next.
        
         | jameshart wrote:
         | Dyson sphere around a neutron star.
        
         | perihelions wrote:
         | It's basically explained in the article: it's probably just
         | very hot--the atmospheric scale height[0] is proportional to
         | temperature. The open question (per the article) is _why_ it
         | got so hot, because they don 't have a model for that. ( _" It
         | certainly requires a significant deposit of energy deep into
         | the planet's interior, but the details of the mechanism are not
         | yet understood.""_)
         | 
         | [0] https://en.wikipedia.org/wiki/Scale_height
        
       | jandrese wrote:
       | Maybe it is something like a small planet that has a giant ring
       | complex that happens to be aligned perpendicular to us?
        
         | astrolx wrote:
         | Usually, you would see the sign of this in the data - because
         | we can actually detect the gap between the planet and eventual
         | rings. But if the gap is extremely small, then it could be a
         | possibility.
        
       | ck2 wrote:
       | It's weird to me they can make such assessments with confidence
       | when we didn't even know the density/composition of Jupiter to
       | any certainty until Hubble was around just in time to observe
       | Shoemaker-Levy plow into it and eject enough surface matter
       | 
       | https://en.wikipedia.org/wiki/Comet_Shoemaker%E2%80%93Levy_9...
        
         | cdelsolar wrote:
         | We've known the mass and the size of Jupiter since way before
         | the shoemaker-levy impact.
        
         | astrolx wrote:
         | AFAIK the Shoemaker-Levy impact was useful for other studies,
         | but not so for the density/mass of Jupiter (also as per your
         | link).
         | 
         | Jupiter is full of surprises still (mostly about its interior).
         | However, the assessments made in this article about the mass
         | and radius of the planet are using very basic physical laws
         | (mostly Kepler laws from the 1600s).
        
       | abc_lisper wrote:
       | Or it could be alien made. Like a dyson sphere - but a planet,
       | with solar capture on the shell. Build everything without all the
       | mass, add magnetic shoes and other tech to make living in 0g a
       | reality.
        
       | cedws wrote:
       | >In data taken between 2006 and 2008, and again from 2011 to
       | 2012, the WASP-South observatory detected periodic transits, or
       | dips in light, from the star WASP-193. Astronomers determined
       | that the star's periodic dips in brightness were consistent with
       | a planet passing in front of the star every 6.25 days. The
       | scientists measured the amount of light the planet blocked with
       | each transit, which gave them an estimate of the planet's size.
       | 
       | I'm not a physicist nor astronomer so I apologise if I sound
       | arrogantly dismissive about something I am clueless about. But
       | how can they be so confident about something from a single
       | measurement? Couldn't there be other things causing dips in
       | light, like dust clouds/asteroids also passing in front of the
       | star at the same time? It seems like a flimsy way of figuring the
       | size of another astronomical body.
        
         | feoren wrote:
         | When we think about making an observation with a telescope, we
         | might think of someone in a crow's nest looking at a far away
         | island, and think: it could be debris, it could be a boat, it
         | could be a smudge on the telescope, it could be a whale, etc.
         | etc. There are several reasons why this intuition breaks down
         | looking at the stars.
         | 
         | For one, astronomical systems are _extremely_ regular. The same
         | thing will happen, over and over, almost the exact same way,
         | millions or billions of times before anything appreciably
         | changes. It 's extremely rare to find a system that is in the
         | midst of significant change. We do find them, of course, and
         | study them, but we do so after passing over countless "boring"
         | areas of the sky.
         | 
         | This regularity means that anything we look at is extremely
         | likely to be in a very stable configuration, and we can use our
         | knowledge of orbital mechanics to rule out all sorts of things
         | like weird dust clouds orbiting the star, which would not be
         | stable (or would look very different if they were). Realize
         | also that we stare at these systems (or at least check up on
         | them periodically) for _years_ at a time -- in this case they
         | said 2006 to 2008, and again 2011 to 2012. So we can easily
         | rule out some random thing passing between our telescope and
         | the star precisely every 6.25 days for 3 years. It must be
         | something stably orbiting the star.
         | 
         | We can also use the exact shape of the "light profile" to be
         | pretty certain about the shape of the object passing in front
         | of it. Over dozens or hundreds of observations, we see the same
         | characteristic dip in light at the exact same period every
         | time, and it matches the shape of a circle passing in front of
         | another circle. It doesn't happen immediately, because the
         | planet spends a bit of time only partly obscuring the star, and
         | we can use this to discern that it is roughly circular in
         | shape. We use this profile and the amount of reduction to
         | calculate the size of the planet. Asteroids would be far too
         | small to detect (we can't even detect all the asteroids in our
         | own solar system!)
         | 
         | Also realize that when they say "light", they're usually
         | looking at a whole spectrum of light. It's not one number that
         | dips, it's a whole waveform that subtly changes. A dust cloud
         | would be partly transparent, differently for different
         | wavelengths, and change the light in a different way. In the
         | extreme cases we can even use this to guess whether the planet
         | might have an atmosphere, because it changes the waveform
         | differently at its edges.
         | 
         | Space is also _really_ empty. A bird might pass in front of
         | your telescope on land, but in space, it 's almost certain that
         | _nothing_ is going to randomly pass between you and what you
         | 're looking at, even over years.
         | 
         | And finally, even with all this, they spend _years_ analyzing
         | the data before they can be confident enough to make claims
         | like this. There is a _lot_ of data and a _lot_ of time spent
         | on it, and it 's the culmination of many different lines of
         | evidence. It's certainly not "flimsy".
        
           | RHSman2 wrote:
           | How very well explained
        
         | sniggers wrote:
         | IANAPNA either, but a solid object like a planet blocks light
         | differently than dust clouds. Dust clouds, even dense ones,
         | still allow some wavelengths through that planets do not. I
         | would guess that's what allows them to be confident that it's a
         | planet.
        
       ___________________________________________________________________
       (page generated 2024-05-14 23:02 UTC)