Untitled

This time it is about regular debris disks around regular main sequence stars. Thanks to recent infrared observations from JWST and the 1.3 mm ALMA radio interferometry survey, we have unprecedented details about the shapes of debris disks and gas profiles. As for the reason that astronomers study debris disks at different wavelengths, the wavelength of the scattered or emitted light is proportional to the size of the dust grains. So JWST will resolve the warm debris disks that are located closer to the star, and the small cold dust grains that are farther away than the mm grains due to radiation pressure. Anyways, these new, very detailed data are a testbed for a whole bunch of theoretical models. In broad strokes, three processes affect debris disk shapes. One, collisional cascade, in which collisions between planetesimals will shatter the objects into smaller pieces, triggering further collisions between the pieces. Two, Poynting-Robertson drag, which is caused by uneven emission of radiation after absorbing solar radiation due to the dust moving around the sun. Third, larger planetesimals or planets can perturb or stir debris disks due to their gravity. All these factors can affect the shape of the ring, dust size distribution, disk eccentricity (how oval it is), brightness asymmetry, and other asymmetric features. Finally, they have started probing on debris disk populations around M dwarfs, the stars with the lowest possible mass.

Even a bigger deal though, is that now we have a lot more studies and details on gas around debris disks. Something that even two years ago, while we did detect gas such as carbon monoxide (CO) were around disks, we did not have enough papers coming out to detail them. With current observations, we can begin resolving whether the gas is mainly primordial or replenished. Aside from further details on gas and dust compositions, we now know that the gas profile follow dust profile, and that there are gas asymmetry in the disks. They also observed that there is opacity to the radio waves from CO. Furthermore, since they know more about gas around debris disks, they can finally do fluid simulations, and model gas asymmetry and the dampening of dust velocity.

A notable result on debris disk gas that the talks loved to bring up was they detected fluorescence from CO around the star 49 ceti using JWST near infrared spectrograph. The emission even has evenly spaced peaks due to vibrational states of the CO molecule (Worthen et al 2025):

This curve is best explained by having some dust population fluoresce from UV light combined with some dust being hit by infrared.

A fun fact I found out from one of the talks. There is a debris disk picture from the Artemis II moon mission:

There are two broad concepts being talked today. First, it's protoplanetary disks, which is a disk of gas and dust in the process of forming planets and being accreted by the very young star still in the process of contracting. The gas is very important in a protoplanetary disk since interaction between the gas and dust greatly affects the way they behave. I particularly enjoyed the talk on streaming instability. One of the big issues in planet formation theory is the difficulty in getting 1 mm sized dust into kilometer sized planetesimals. Streaming instability solves that by taking account dust interactions with the gas. Since gas has pressure gradient, dust particles will face headwind against a slower moving gas, thereby creating drag. If there are a cluster of particles though, that will reduce drag locally behind its orbit and further clump matter in the area, which then combine by gravity. The talk went through various studies and how change in viscosity of gas, dust/gas ratio, and gas pressure gradient affected the planet formation models. And as always, when it comes to protoplanetary disks, people bring up the discussion of whether the gas that remains after the protoplanetary disk turns to debris disk is leftover gas or replenished.

More exciting for me though were the talks on white dwarf disks. While I have heard of it, this is the first time I am getting the details, and it is quiet exciting for me. From the talks, I get the impression it is a relatively new, burgeoning field, with first empirical evidence for it published in 1987, and the first working model developed in 2003. That particular model, the Jude model, assumes a Saturn like ring around a white dwarf: opaque, thin, circular, inner edge delineated by sublimation of silicates, outer edge by the Roche limit (the limit at which an object held together by gravity will break apart). Finally, Poynting-Robertson drag is a factor for accretion, drag from uneven heating by radiation due to the movement of the dust.

Graham Platner is fucking Count Olaf and half the left (mainly white people to be honest) were too idiotic and incompetent to see it. He might as well have called himself Captain Sham, since apparently having a Nazi tattoo was not disqualifying at all, people would have eaten it up anyways. Oh and certain content creators and reporters should stop talking about politics because if they fall for something this easy, they are obviously too unqualified to talk about politics.

-The physical book itself is really nice. The jacket is well designed, and the interior of it has a map. The hardcover gives it a legit textbook feel, and it comes with a bound ribbon bookmark.

-It's a better read than I expected, since it isn't just an encyclopedia of facts, but it is in the view point of a specific character in Terra. That it is in Erikson's viewpoint, a Victorian scholar and noble, albeit a more progressive minded one, is integral to the text's interpretation. It is an aspect that is continuously pointed out by both Erikson himself, and Kaltsit's notes in the margins. So while he supports the rights of the infected, and views Sarkaz as people, he also has specific views on technology, progress, and civilization that an upper class Victorian would have. The facts that he does have has a lot to do with the travels he did through various countries. It is also a book that he wrote with a specific purpose, as told by his correspondence with Kaltsit. It is not meant to be academic and tries to go softer on controversial topics such as oripathy and infected so that people are eased into it. Great care has been taken by the irl author so readers are aware of what types of viewpoints and biases are in the book by the fictional authors. Basically, it's not an objective book on Terra, and it is what I like about it.

-Speaking of technology, there is a particularly salient section in the book regarding the in universe book "A Land Without Originium". Basically, the inability for people to imagine technology without originium speaks more of the people of the specific place and era than what technological development actually was like in Terra.

-It is obvious some of the details in the world building have not been properly filled out because there are segments that are very broad strokes. Book also likes to repeat its point in some sections a bit too much for my liking.

-Some of my favorite aspects are the little stories peppered throughout the book. Like the story about the guy who built Gaul's first nomadic city, the cannonade on catastrophe experiment, or the story of the infected Victorian defectors in the Reunion side booklet.

-The section on the history of Sarkaz is instead told by a Sarkaz school teacher who died during the civil war, letting a Sarkaz tell the story of their people.

-My favorite part was the Aegir section, a series of science fiction short stories written by an Aegir writer who himself has an interesting story.

-The Yan section is my least favorite. When other countries are complicated and nuanced regarding history, demographics, culture, etc (for example, think of how Ursus's flag originated from given its history, or the history of migration and the unique polity in Rim Billiton), everything before and after the 100 clan rebellion feels like rosy propaganda tbh. Should have instead collapsed like 10 times just like real life China.

-Ah yes, the three branches of the Columbian government, the executive branch (the federal administration), the legislative branch (the federated congress), and the Department of Defense.

-The Seven Cities Federation part is really cool as it is about some of the last remaining city states trying to fend off being absorbed into larger states. There is a tragic irony that the first nomadic city was meant to be a tool of liberation, but was co-opted by large nation states.

-While a couple of typos is something I can overlook, since there is a lot of text so this is bound happen, there are two major textual mistakes that I feel like editorial should have caught going through it.

So the last couple of days I have been attending a conference on the usage of Large Language Model (LLM) on astronomy and astrophysics. I usually don't do this, but I came away from the conference simultaneously impressed and thinking that AI proponents are just not beating the allegations, even in a professional, scientific setting.

Now, before I go on some of my fundamental disagreements, I do want to mention, a lot of people are doing good research work. A lot of them are testing the limits of LLMs and benchmarking them. They have developed tools so that LLM agents will interface with external or internal sources, and draw real data from the right place. The formatting will not be consistent, but the numbers drawn are real. A lot of talks and people have also been honest about what doesn't work about it, and what needs to be improved. And legitimately, some works, like a way to query astronomical data from database, could legit be a game changer. I personally know that going through all these disparate astronomical databases is a pain in the ass. Or if there are quick ways to classify humongous amount of data, say, galaxy images, with certain amount of reliability, it would be cool. I hope they succeed.

The other aspect I liked is that I learned quiet a bit. I learned more about how LLMs worked and some of the softwares used. There was a talk about how LLMs think of tokens rather than letters and words, and why they make the mistakes they do, and I really enjoyed that one. I actually came away from day 1 of the conference feeling pretty positive. They also did hack sections. While I wasn't very good at it, it was fun thinking of stuff. Even if the ideas are silly, it's okay, you are learning.

Oh and I did like how one of the host of the conference handled a potential power dynamics situation. The hosts were running the event smoothly, they did a great job.

At the same time, some of the stuff presented makes me think, why? As an example, why would I want an LLM summary of a paper? Every paper literally has a summary, it's called an abstract. If you want more details, go read the conclusion. Instead, we want to spend a massive effort trying to correct all the fake stuff when the amount of stuff being published is faster than the available man power for it? The most ridiculous example of this was someone proposing to mint NFTs as a way to verify AI videos as science accurate. In the year 2026. I don't even want to go into why this is such a stupid idea. That said, I always allow for one crank idea for every conference, even if this one made me particularly angry.

The worst part of this conference is the insidious nature of its worst ideas. The presentations are slick, the words sound reasonable, but they stand for the promotion for lowering the standards and hypocrisy. Nowhere was this more evident in a talk where I was lectured about what LLMs can and cannot achieve, the fact that LLMs don't even save that much time in a lot of tasks, all the while using trashy, clearly AI generated slides that, while sleek on the surface, were aesthetically clashing and incomprehensible to interpretation (and I have a suspicion they are not the only person to do it by the way). Now, I have seen plenty of bad talks, including from yours truly, but you know, at least I can say that my slides could reasonably be followed by human beings. Earlier in the conference, another speaker gave a talk about how current ethical principles could be used to ground LLM research without new guidelines. Well, they were burying the lede here. Because as the other speaker admitted, they thought the power point looked good enough when asked about it. Clearly too many people are too oblivious and lack enough self-awareness that we do, in fact, necessitate new guidelines. After all, someone did go in front of everybody to present that they were writing a book that would normally take four years in one year. To be honest, I do not trust these people enough to think it is going to end well. After all, so many things I saw and heard made me think a lot of these people don't really care about science communication. Which as an educator, well, it kind of pisses me off. In addition, there was an argument that went, when every astronomer adopted MCMC thanks to the emcee package, a lot of people did not due their due diligence to check burn-in and convergence, and that is okay. Now, I agree with the part that it's a good thing you don't have to learn every little thing about MCMC. That does not mean that the argument should extend such that the former is okay. Maybe the referees should raise their standards a bit regarding MCMC then?

Now, the Victorian arc was very ambitious, at times overly so. I think it's safe to say there are a few dead ends and turn backs in the story as the writing group tried to figure out how to handle this behemoth of a premise. It's the writing group's attempt to create a fictional civil conflict with a foreign invasion, which if you have followed any real life conflicts like that, they are actually extremely complicated and confusing to follow. While I personally like the attempt, it is not going to endear a lot of people to the storyline, and the story does creak at its attempt at juggling the details of a billion different things. It is also not helped by the fact that I think chapter 9 and 10 were both pretty messy introductions to the conflicts for different reasons for me personally. Chapter 11-14 though, each tries to bring focus to the story by aiming for a specific theme, each ending bringing the end of certain storylines to simplify the situation. In addition, you have Firelight and Babel bringing in much needed context. So, with all that in mind, chapter 14: Absolved will be the Seekers, the end of the Victorian arc, had a grand assignment. It has to close all the remaining different stories in a satisfying manner, all following an overarching theme that will satisfy its own story, but also the wider six chapter arc. It has to cover the fallout from Babel and integrate the inevitable exposition dump on the truth of originium into the wider emotional truths of the arc and the series. It has to create a rational geopolitical result for the status quo change for the Sarkaz, the Exemplars, the dukes, Dublinn, and even in a minor way, Reunion even if chapter 13 for the most part concluded its story.

So for my part, they have managed to land the plane pretty well, with a few rough spots aside. The best handled storyline is the one on the Sarkaz side. It was a good idea for Wisadel to be the emotional anchor to the overall theme of generational trauma, fate, and inheritance. She is kind of the main character of the chapter, along with Amiya and Siege. Her moment to moment dialogue is pretty funny, and her dynamic with Mon3tr and Doctor are some of the most fun I had reading this chapter. Really though, she has great dramatic moments too. I got emotional to her finally processing her grief over Theresa, and her finally getting to properly talk with Theresa in the assimilated universe. I find her character arc is beautiful, she inherited a weapon of war and got her name via war, but now she is given a hopeful name and inherits Theresa's ideal. Maybe the intent is, some day, Kazdel can follow her story writ large.

On the Theresa side of the story, she takes ownership of the fates of the Sarkaz by exorcising the Myriad Souls, the literal manifestation of generational trauma via scifi mystery technology. I found the actual concept of the myriad souls terrifying actually, a reality where some data of you is forever reliving the same trauma, and then passing that on to your descendant. The fact that seeing the truth of the myriad souls pacifies the Revenant, who ironically, his entire life was spent living in a similar state and jealous of those who passed on, was low key but very effective in driving that home. Anyways, I think it's interesting that Theresa after death chose to be ruthlessly pragmatic, and backed Theresis in exchange for being able to do this. This makes her reunion with Rhodes Island both sweet and tragic. And god do they milk the drama for all it's worth, with my personal favorite moment being the moment Amiya reveals how she knew who betrayed Babel the whole time, and that she has been grappling her image of who the Doctor is. Anyways, the Rhodes Island side unpacks the legacy Theresa left behind, how they are carrying on with her legacy despite the previous bad end, and how they have to move beyond her. As an aside, I really love Logos in this. He is badass, idealistic, a goof, and takes his responsibilities very seriously. A representative of sorts of the new generation of Kazdelians raised with different ideals via Theresa's new educational system back then, and now having the means to fight for the new ideals.

-

-

-

-