I love how accurately you capture the constant swing of the cognitive pendulum one experiences when dealing with the incessant flow of news (and its interpretation). Every issue seems bigger than one person can handle these days.
On Monday you believe in Aliens, Achieved AGI and Quantum Supremacy. On Wednesday, after reading a few, down to earth rebuttals, you regret having been so easily fooled into marvel.
On Friday, you’re in awe again but this time with an aftertaste of suspicion as the debate gets too technical even for what seemed like an expert a few days ago.
On Sunday you just want to forget about it and your ability to understand any of it.
Great link! I also remember another article by Scott around the time of the replication crisis and how he was trying to find some studies that he still trusted. If I recall correctly, he ends by basically throwing his hands up in the air? Not too sure.
"The Monster Scope has a mirror diameter of 1 km – an unthinkably enormous span of glass enabling us to examine the features of nearby exoplanets with the same detail anyone can see on Jupiter with a pair of binoculars."
Okay, that's cool. And agreed. Designers haven't adjusted to the fact that if Starship gets working, the cost to orbit is so low that it changes significantly what's possible.
It's an extremely cool idea, and I was quite taken with it myself until I ran the numbers and realized that the angular resolution of a 1 km telescope isn't even sufficient to resolve details on Proxima Cen b.
You can still do a lot with it though.
But what we really need is mirror swarm acting as an interferometer, which can be built on the necessary scale and extended more or less indefinitely.
To be fair, I attended quite a few seminars when I was in grad school a handful of years ago that basically all had the conclusion: “there may or may not be signs of life on this planet right now or in the past, so we cannot conclusively say there is or was life but we also cannot say there isn’t or wasn’t…” This was actually a joke amongst my friends. It’s really interesting stuff, but I think planetary science has been this way for a while. It is exciting to read about from the sidelines! I like your nuanced way of writing about it.
Thanks for the greay summary of the state of affairs! For what it’s worth, I am fundamentally agnostic (though the likelihood of an anthropomorphized omnipotence seems vanishingly minute), and absolutely live in perfect uncertainty. It’s simply impossible to conclude whether a supernatural entity exists based off of purely natural data. Such metaphysical questions can be fun to speculate about, but must be left as a pure question mark to stay intellectually honest. At least with alien life, we’re working with natural processes 🙂
I love the idea that the public has been thrust into working science. In an age of so little epistemic trust, it is essential to foreground the methods that construct the facts we all take to be true. I wonder how we would help more people engage with how this process happens and pieces like this help do that. It has the potential to reshape the role science plays in public life. I, for one, am SO excited for the age of alien agnosticism.
I'm a professional astronomer!. It's worth saying that basically the whole exoplanet community is VERY sceptical of the result, to the extent of some researchers accusing the authors of statistical malpractice. It's quite shaky.
Essentially, the authors take a mid-IR spectrum of the planet's atmosphere, taken with JWST. They then try to fit the spectrum with a combination of around 20 likely molecules (including methane, CO2, and DMS+DMDS). This fit produced about a 2-sigma likelihood of the planet having DMS in the atmosphere, meaning VERY weak evidence. Certainly not publishable.
Then the authors designed a second model, where they take out every other molecule that cause features in about half the spectrum (9-12 microns). So the model ONLY has DMS (and DMDS) features beyond 9um or so. Then they fit this new, reduced model to the data, and it produces a 3-sigma likelihood of the planet having DMS in the atmosphere. This is the result they published.
BUT, they only arrived at this higher likelihood by removing all the other possible sources from their model! Yes, if you remove everything else that could be causing the signal except for DMS, you're going to find that DMS fits your data well. But it's a level of data massaging that borders on dishonest.
Imagine you had a machine that analyses animal calls, and uses Bayesian fitting of known animal sounds to tell you what animal produced the call. You hear a weird sound one day, and feed the file into your machine. The result is inconclusive. Maybe lion, maybe tiger, maybe gorilla. Nothing stands out. But you want to know if you've heard a lion! So, you edit the machine's code so it's only looking for lion noises, and no other animals are being considered. Then you feed in the same sound file, and the machine goes "Yeah, that could totally be a lion".
Of course there's other life in the universe. The universe is just too vast to actually contact it. We also might not recognize it if we did find something close enough. It could be vastly different.
Great post, though an important note on the magma world thing and why it deserves mention even though it didn't include the term "DMS."
The point of mentioning the other compositions for K2-18b, like a magma ocean scenario or the gaseous mini-Neptune one, is not to say that these models included a source of DMS. It's to say: "we don't even know if this planet is habitable — heck, we don't even know its composition, so how in the universe are we supposed to accurately model its geochemistry and atmosphere with enough confidence to rule out abiotic sources?" It's a nod to the problem of unknown unknowns which, while probably not 100% solvable, can be mostly dissolved by gathering more and more information over time.
Of course, there's no such thing as certainty, especially when you're talking about planets you'll never get to visit. We have enough trouble figuring out what geological processes are going on in our own solar system, sometimes even on our own planet. It's only recently that we convinced ourselves that there's active vulcanism on Venus, the planet that's closest to us. We can't even "see" most exoplanets, including this one. We study them by watching their stars flicker as they pass in front of or behind them.
There are also legitimate claims challenging the detection altogether (some have accused the 3 sigma of being a case of statistical hacking), so stay tuned for more uncertainty.
Any chemistry buffs want to speculate whether these concentrations might be produced through unknown pathways in high hydrogen + high pressure atmospheres? Or link to any papers that look into this?
I'd definitely be interested in such proposals. My understanding is that one major issue to an abiotic explanation (assuming the signal is real, of course - there's always a possibility it too doesn't replicate) is that DMS decays quickly, and to be maintained would need low-levels of carbon dioxide. But they observe high levels on K2-18b. So the model would need to fit the various atmospheric conditions we do know about. I wouldn't be surprised if the Reed lab (here: https://iopscience.iop.org/article/10.3847/2041-8213/ad74da/meta) follows this up with a more specific experiment to try to more precisely mimic K2-18b's atmosphere.
Could indeed be - whoever is "first" here goes down in history books, so researchers have a strong confirmation bias. But I'm always of two minds about ego in academia. Of course, we associate ego with being a bad thing, but also historically (and even among the people I've met) ego in academia can be a good thing. There's lots of headwinds in academia, so it requires a bullish self-confidence to do truly novel work. And also, the pay is low, so there's only so many motivators to go around.
Yes, egos motivate scientists. But overstating and hyping speculative scientific results to stroke them is not helpful and detract attention and resources from those doing thoughtful verifiable, reproducible research. Many researchers at the forefront of exoplanet research report high quality results, including Dr. Laura Kreidberg and Dr. Erik Petigura. As a journalist, your job is to request their opinions, on or off the record, to learn the truth.
As a former science writer, fwiw, it has always astonished me how the science community religiously poo-poos massive troves of observational-albeit-anecdotal data regarding such topics.
To cite as fact a video game designer-turned professional debunker's armchair theory which posited eye-witness testimonies from multiple trained air force pilots were caused by their failure to understand the parallax effect is de facto odd and suggests the writer has not studied the issue beyond cherry-picking data that serves their premise.
It is an issue that has repeated itself throughout history. Take gorillas; first observed in the 5th century, were treated as a mythical joke by European academics - even after countless reports, bones and pelts were produced. Gorilla beringei aka mountain gorillas, remained lumped with unicorns and other cryptids until 1902.
The default mode here seems to be: "Things other people have seen which I have not seen must be what I believe them to be because I place my observations above everyone else's without merit and / or perhaps suffer from a debilitating case of normalcy bias.
Fascinating read! I’m Harrison, an ex fine dining industry line cook. My stack "The Secret Ingredient" adapts hit restaurant recipes (mostly NYC and L.A.) for easy home cooking. Dm me if interested in a recommendation swap — we’re growing fast!
I love how accurately you capture the constant swing of the cognitive pendulum one experiences when dealing with the incessant flow of news (and its interpretation). Every issue seems bigger than one person can handle these days.
On Monday you believe in Aliens, Achieved AGI and Quantum Supremacy. On Wednesday, after reading a few, down to earth rebuttals, you regret having been so easily fooled into marvel.
On Friday, you’re in awe again but this time with an aftertaste of suspicion as the debate gets too technical even for what seemed like an expert a few days ago.
On Sunday you just want to forget about it and your ability to understand any of it.
https://slatestarcodex.com/2019/06/03/repost-epistemic-learned-helplessness/
Great link! I also remember another article by Scott around the time of the replication crisis and how he was trying to find some studies that he still trusted. If I recall correctly, he ends by basically throwing his hands up in the air? Not too sure.
This one maybe?
https://slatestarcodex.com/2014/04/28/the-control-group-is-out-of-control/
Yup, that's the one. Thanks!
☯️
We really should build Casey Handmer’s Monsterscope. It feels like the only worthwhile space mega project to really invest into in the upcoming decades. https://caseyhandmer.wordpress.com/2024/11/30/it-is-time-to-build-the-monster-scope/
"The Monster Scope has a mirror diameter of 1 km – an unthinkably enormous span of glass enabling us to examine the features of nearby exoplanets with the same detail anyone can see on Jupiter with a pair of binoculars."
Okay, that's cool. And agreed. Designers haven't adjusted to the fact that if Starship gets working, the cost to orbit is so low that it changes significantly what's possible.
Yeah not to mention cosmology/black holes/dark matter/basically way higher resolution everything in astronomy
and can be modular and upgradable including its total size in the future
Time to think big (although I saw some people disputing the claims to resolution)
It's an extremely cool idea, and I was quite taken with it myself until I ran the numbers and realized that the angular resolution of a 1 km telescope isn't even sufficient to resolve details on Proxima Cen b.
You can still do a lot with it though.
But what we really need is mirror swarm acting as an interferometer, which can be built on the necessary scale and extended more or less indefinitely.
https://barsoom.substack.com/p/what-big-eyes-you-have
To be fair, I attended quite a few seminars when I was in grad school a handful of years ago that basically all had the conclusion: “there may or may not be signs of life on this planet right now or in the past, so we cannot conclusively say there is or was life but we also cannot say there isn’t or wasn’t…” This was actually a joke amongst my friends. It’s really interesting stuff, but I think planetary science has been this way for a while. It is exciting to read about from the sidelines! I like your nuanced way of writing about it.
I never knew that exoplanet excretion could be so very exciting!
Thanks for the greay summary of the state of affairs! For what it’s worth, I am fundamentally agnostic (though the likelihood of an anthropomorphized omnipotence seems vanishingly minute), and absolutely live in perfect uncertainty. It’s simply impossible to conclude whether a supernatural entity exists based off of purely natural data. Such metaphysical questions can be fun to speculate about, but must be left as a pure question mark to stay intellectually honest. At least with alien life, we’re working with natural processes 🙂
I love the idea that the public has been thrust into working science. In an age of so little epistemic trust, it is essential to foreground the methods that construct the facts we all take to be true. I wonder how we would help more people engage with how this process happens and pieces like this help do that. It has the potential to reshape the role science plays in public life. I, for one, am SO excited for the age of alien agnosticism.
Hi Eric,
I'm a professional astronomer!. It's worth saying that basically the whole exoplanet community is VERY sceptical of the result, to the extent of some researchers accusing the authors of statistical malpractice. It's quite shaky.
Essentially, the authors take a mid-IR spectrum of the planet's atmosphere, taken with JWST. They then try to fit the spectrum with a combination of around 20 likely molecules (including methane, CO2, and DMS+DMDS). This fit produced about a 2-sigma likelihood of the planet having DMS in the atmosphere, meaning VERY weak evidence. Certainly not publishable.
Then the authors designed a second model, where they take out every other molecule that cause features in about half the spectrum (9-12 microns). So the model ONLY has DMS (and DMDS) features beyond 9um or so. Then they fit this new, reduced model to the data, and it produces a 3-sigma likelihood of the planet having DMS in the atmosphere. This is the result they published.
BUT, they only arrived at this higher likelihood by removing all the other possible sources from their model! Yes, if you remove everything else that could be causing the signal except for DMS, you're going to find that DMS fits your data well. But it's a level of data massaging that borders on dishonest.
Imagine you had a machine that analyses animal calls, and uses Bayesian fitting of known animal sounds to tell you what animal produced the call. You hear a weird sound one day, and feed the file into your machine. The result is inconclusive. Maybe lion, maybe tiger, maybe gorilla. Nothing stands out. But you want to know if you've heard a lion! So, you edit the machine's code so it's only looking for lion noises, and no other animals are being considered. Then you feed in the same sound file, and the machine goes "Yeah, that could totally be a lion".
Of course there's other life in the universe. The universe is just too vast to actually contact it. We also might not recognize it if we did find something close enough. It could be vastly different.
Great post, though an important note on the magma world thing and why it deserves mention even though it didn't include the term "DMS."
The point of mentioning the other compositions for K2-18b, like a magma ocean scenario or the gaseous mini-Neptune one, is not to say that these models included a source of DMS. It's to say: "we don't even know if this planet is habitable — heck, we don't even know its composition, so how in the universe are we supposed to accurately model its geochemistry and atmosphere with enough confidence to rule out abiotic sources?" It's a nod to the problem of unknown unknowns which, while probably not 100% solvable, can be mostly dissolved by gathering more and more information over time.
Of course, there's no such thing as certainty, especially when you're talking about planets you'll never get to visit. We have enough trouble figuring out what geological processes are going on in our own solar system, sometimes even on our own planet. It's only recently that we convinced ourselves that there's active vulcanism on Venus, the planet that's closest to us. We can't even "see" most exoplanets, including this one. We study them by watching their stars flicker as they pass in front of or behind them.
There are also legitimate claims challenging the detection altogether (some have accused the 3 sigma of being a case of statistical hacking), so stay tuned for more uncertainty.
The 3-sigma result is 100% a case of statistical hacking. I posted a comment above :https://www.theintrinsicperspective.com/p/alien-poop-means-we-are-not-alone/comment/110395991
I feel like this deserves more attention!
Any chemistry buffs want to speculate whether these concentrations might be produced through unknown pathways in high hydrogen + high pressure atmospheres? Or link to any papers that look into this?
I'd definitely be interested in such proposals. My understanding is that one major issue to an abiotic explanation (assuming the signal is real, of course - there's always a possibility it too doesn't replicate) is that DMS decays quickly, and to be maintained would need low-levels of carbon dioxide. But they observe high levels on K2-18b. So the model would need to fit the various atmospheric conditions we do know about. I wouldn't be surprised if the Reed lab (here: https://iopscience.iop.org/article/10.3847/2041-8213/ad74da/meta) follows this up with a more specific experiment to try to more precisely mimic K2-18b's atmosphere.
bloody knew it, everything is meme or poop
Much ado about nothing. Except egos in academia.
Could indeed be - whoever is "first" here goes down in history books, so researchers have a strong confirmation bias. But I'm always of two minds about ego in academia. Of course, we associate ego with being a bad thing, but also historically (and even among the people I've met) ego in academia can be a good thing. There's lots of headwinds in academia, so it requires a bullish self-confidence to do truly novel work. And also, the pay is low, so there's only so many motivators to go around.
Yes, egos motivate scientists. But overstating and hyping speculative scientific results to stroke them is not helpful and detract attention and resources from those doing thoughtful verifiable, reproducible research. Many researchers at the forefront of exoplanet research report high quality results, including Dr. Laura Kreidberg and Dr. Erik Petigura. As a journalist, your job is to request their opinions, on or off the record, to learn the truth.
Yep, just a total waste of useful brain power…
But poop
As a former science writer, fwiw, it has always astonished me how the science community religiously poo-poos massive troves of observational-albeit-anecdotal data regarding such topics.
To cite as fact a video game designer-turned professional debunker's armchair theory which posited eye-witness testimonies from multiple trained air force pilots were caused by their failure to understand the parallax effect is de facto odd and suggests the writer has not studied the issue beyond cherry-picking data that serves their premise.
It is an issue that has repeated itself throughout history. Take gorillas; first observed in the 5th century, were treated as a mythical joke by European academics - even after countless reports, bones and pelts were produced. Gorilla beringei aka mountain gorillas, remained lumped with unicorns and other cryptids until 1902.
The default mode here seems to be: "Things other people have seen which I have not seen must be what I believe them to be because I place my observations above everyone else's without merit and / or perhaps suffer from a debilitating case of normalcy bias.
Otherwise, nice article;)
Aliens are a whole lot more likely than AGI. It has no effect on gdp for the same reason the one horned skunk ape has no impact on gdp
I do too agent mulder😎
Fascinating read! I’m Harrison, an ex fine dining industry line cook. My stack "The Secret Ingredient" adapts hit restaurant recipes (mostly NYC and L.A.) for easy home cooking. Dm me if interested in a recommendation swap — we’re growing fast!
check us out:
https://thesecretingredient.substack.com