Journal
This journal is a daily account of my creative labors and casual reflections. It’s intended to be frank, not entertaining. It’s also freewritten, quite unlike my other writing. The ideas jotted down here are thoughts going through my head rather than definitive views or even firm opinions, and they’re often completely wrong. In fact, it’s normal for me to work out completely wrong ideas at some length before I find correct ones. So don't take anything you read here seriously.
Listen to my most recent music:
November 3, 2026
Some years ago I bought a record called “Everyone Alive Wants Answers.” While I wouldn’t call it a classic or a masterpiece, it was a nice record, and the title has always stuck in my mind as well. It took ages for me to realize that the title was not, in fact, true. Most people don’t want answers. They either don’t care about answers or they want other things more. For instance, they want to believe things that are demonstrably false but that agree with their biases in some way.
The number of people who actually want answers is quite small; and when it comes to uncomfortable questions, it’s miniscule. One reason I was able to determine the truth about feminine beauty, and about various related topics, is that I actually wanted the answers and nobody else really did.
The disinterest in answers continues to mystify me. When I discuss these topics with various people I usually come to a point where I realize: oh. So they don’t actually want answers. Why are they bothering discussing the question at all then? Presumably they want to be told something or other regardless of whether it’s the answer, or want to make some proclamation regardless of whether it’s the answer (or whether it’s even their own genuine opinion).
Well - have you ever considered actually wanting answers?
November 1, 2026
I haven’t written much about the Iran War here lately. My predictions since late spring have been coming true, and based on the course of events my view on the war looks more correct than everyone else’s. I think it would be pretty funny if the author mainly known for posting about cute girls happened to correctly predict the course of the Iran War against the overwhelming weight of expert (and other) opinion while also writing Optimalism. Well, I’m not going to brag that that’s happened yet and chaotic global events could still derail my expectations, but we are on the way there.
My writeups in this journal haven’t been super coherent because I didn’t want to take the time to detail all the thinking behind my theory. However, I explained it quite clearly and in depth in my interview with Blackbeard (it’s near the end of that long interview). If you have a lot of time on your hands you can trace it back to the earlier entries in this journal and they should make more sense.
Still seems to me global depression by the end of 2027 is the central case, though that could be downgraded if we were to get a quick resolution in the next couple months. No, I don’t expect a quick resolution, but one can’t rule it out entirely. USG gets to decide how much pain to inflict, and max pain is my base case. Distorting price signals and using up reserves over the past six months has delayed adaptation. That makes it seem to casual observers that nothing is happening, but it can end up increasing the pain when reality finally hits.
This doesn’t mean the US economy has to collapse. According to my theory we should expect the government to pump an incredible amount of money in. More to businesses than consumers. They might wait till some crisis signs to justify and cover it, but they could start pumping now if they really wanted to. I don’t think this will mainly take the form of broad QE but rather dirigisme, though that’s just a guess.
Also looks likely to me that the big AI labs will need and get a bailout. AI is the future, but they probably can’t profit fast enough, especially against pressure from cheap Chinese models. Bailout will come for national security reasons. And rightly so. There will be a lot of grumbling about “too big to fail,” but that’s not the fundamental reason. Unless the math breakthroughs turn out to be a mirage or unrepeatable (they don’t look that way), US simply can’t afford to fall behind here.
Anyway, none of this is my area of expertise, but we’ll see how I do. For now, back to writing about cute girls.
October 30, 2026
Busy day today but I want to quickly respond to something Blackbeard sent me. I don’t have time to get in an extended philosophical debate and don’t like arguing so I’m just going to discuss it in general terms.
Basic question: some kids’ Youtube clip is a million times more popular than Beethoven, is it therefore actually “better” or “more valuable”?
My answer: Well, obviously not.
Couple mistakes being made in the argument in favor of the kids’ clip. First of all you really have to evaluate it by marginal value add. And not just on the basis of headline views. Views follow a power law and what ends up at the top is governed by random factors apart from quality. Both of those points have been demonstrated empirically. So views are misleading as to value add because the next best similar video might be down two orders of magnitude in views but mere chance factors could have flipped it, or the marginal value add of the top one could be extremely small.
Now, the above point is economically accurate but also beside the point because there is a second error in this kind of thinking. Namely that utilitarianism is already bunk. This is why I insist on D.2 of Optimalism which is a much more important principle than you might assume as reading it it just sounds like common sense, but philosophy has never been very good at common sense.
If we think about Husserl’s famous critique of Galileo in The Crisis and then generalize that we arrive at the observation that we tend to extract the most measurable elements from a gestalt and then reify them and ontologically demote the less measurable elements and this is actually an error because it is lossy and the loss is not reconstructable from the compressed signal. Phenomenology is all about this observation, the Galileo section is just the locus classicus. However phenomenology did not, in my recollection, state the abstract principle outright in this way even thought it was always latent there. In any case, whether it was ever stated or not is irrelevant, I am just pointing it out.
Now we can apply this same observation to ethics and generate a critique of utilitarianism which extracted the element of the good that was easiest to conceptualize as a scalar metric and then downgraded the other elements of the gestalt that were less tangible, once again giving lossy compression where the original signal can no longer be reconstructed. Same error. So, utilitarianism is bunk, and so are virtue ethics and deontological ethics. This observation gives you D.2 of Optimalism and yes, all of the above is implicit and motivated D.2.
For the case at hand, you can’t just say that the kids’ Youtube show created more net pleasure and therefore it’s better, you have to think about the gestalt of what is being created. And also through a value add lens. On this basis it is legitimate to say Beethoven is better.
No, I don’t even like Beethoven apart from a small number of pieces.
A further point is that the marginal value of great artworks is going to look different from that of a kids’ show with close competitors because there is a quasi-monopolistic aspect to great artworks due to their uniqueness that is not fully substitutable and in a sense they are more like land.
Apologies if this entry is obscure to readers without a philosophy background, you can ask AI for an unpacking.
P.S. Incidentally I recently watched an early episode of Fraggle Rock on a lark and it was an explanation of post hoc ergo propter hoc. Now we have kids’ Sharknado. How far we’ve fallen.
October 29, 2026
I’ve added an amusing biographical note to the contact page. After some thought on the problem of biography I decided something was better than nothing because in the absence of information people simply invent things.
I have also been helping some with the faces app. I am not sure how involved I want to be in this project or where it is going. Certainly I could add a great deal of value, but I am juggling too many things at the moment and distractions keep interrupting my novel. Committing to work on it intensively would involve a wrenching change in my life and the overall proposition would need to be very persuasive to justify that. I still hope to do some work on the novel this evening.
October 27, 2026
The founder I mentioned in my last entry decided to push his “official” launch date back a few days to iron out kinks and add features. So, I have not posted yesterday’s piece on social media yet. However, the site still works if you want to use it now.
Did some work on the novel today, though didn’t make as much headway as I wanted to.
Some quick notes on the reproduction rate of information below.
I’ve pointed out a few times how viral modeling works well for information. For something to spread to the relevant population by word of mouth, r0 must be = or > 1. That means the group that initially hears of the idea or work must collectively share it with a larger group. If they share it with a smaller group, it dies out without spreading meaningfully.
However, word of mouth isn’t the only way to spread. An algorithm or a sales team can artificially promote an idea or work, like you could artificially seed a virus. Then it can spread even if r0=0!
The power of controlling the information distribution network, in whatever form it takes (algo, sales team, propagandists) is immense for pushing ideas through a society because while one would certainly like r0>1, it’s not necessary. For instance, if you run Netflix and want to push idea Y, you can just keep reseeding it over and over until it reaches everyone.
One problem with my writing on beauty and sexuality is that r0<1, not necessarily because people don’t like it or find it useful, but specifically because they don’t want to share it - for one reason or another they feel uncomfortable doing so. And, I don’t have the scale to reseed it like Netflix propaganda.
It’s obvious how these same observations can be applied to a broad range of ideas. So you can see why controlling information distribution nodes (which include algorithms on social, Amazon, Google, etc. and AI models too), as I discuss in the “Missing Chapters” of Trust Networks, is of such critical importance and can force ideas on society even when they are both bad and unappealing.
This becomes even clearer if you think about it competitively. Suppose there is a good idea X and a bad idea Y which compete for the same space. In each case r0<1, and in fact Y has r0<<1. However, those who control critical information nodes repeatedly reseed Y. Eventually Y dominates society’s thinking. Once established, an idea’s dominance can be self-reinforcing.
A great deal of modern propaganda works in exactly this way, so it’s no surprise people believe such crazy things today.
AI will likely make it easier to manipulate the average person because bias can be built into training in various ways and AI is a strong persuader. I am not sure that competing models will fix this, because high-end training and regular updates may remain expensive and under the control of a small number of nodes that can easily be captured by motivated actors. Of course, whether things will really work out that way remains to be seen. And I am not sure what competing open-source strong persuaders would lead to, as that’s an unprecedented scenario…
October 26, 2026
Surprise event today (or tomorrow, depending on where you live). One of my readers has been building an app based on the faces theory from DBL. I’ve known about this for a few months but I didn’t realize till yesterday that the release was imminent. I’m posting a thread to help him whip up some interest in the app because I think it has huge potential. Going to post it below first even though technically the release is a few hours away, and then on Twitter when I wake up. I can’t promise the links will work immediately. If they don’t, just wait a few hours.
To be honest, I’m super excited about this app. I’m glad someone is finally doing something with the discoveries I made in Dispelling Beauty Lies. There’s a huge amount of latent value in DBL, and basically I’ve been trying to hand out gold to people on a silver platter for years, and until now nobody has even wanted to pick it up because everyone’s so retarded. I hope this app is successful, because the world needs someone to unlock that value. I have more ideas about what you could do with such an app than I even have time to list here.
Anyway, here is the introductory thread I’ve written.
How Understanding Your Type Can Make Dating Easier And More Fun
Are “types” simple fun like horoscopes - or can learning about them actually make our relationships better? Well, the answer is yes and YES. Keep reading and I’ll give you some news you can use.
A few years ago I noticed something really interesting about faces. While we all agree that some faces are prettier than others, we DON’T agree about which we consider our favorites. In fact, most people have a distinctive “type.”
When I asked people to send me their four favorite faces, they typically chose faces that look like four sisters or brothers - without even realizing it! Just take a look.
That was a fascinating discovery. But at first I didn’t know what to do with it.
Eventually I decided to quiz my readers on some generic types made from layering similar faces. This time, I asked them to choose the three they rated highest.
This time I made a second discovery. While we each have distinctive favorites, our favorites aren’t evenly distributed. Some types are a lot more popular than others - even if they’re completely free of blemishes and asymmetries. (The popular types are numbered in green.)
This sounds discouraging. Maybe unpopular people are just out of luck, and they can’t do anything about it. Maybe types don’t really even matter!
However, after I thought about the results for a while I found a silver lining: even the LEAST popular facial types were ranked highly by SOMEONE. Nobody scored a zero. That means every type on the chart has a chance to form a satisfied, happy couple.
There’s a second encouraging implication that’s less obvious: it’s in your best interest to pursue the LEAST POPULAR facial type you rate highly. This counterintuitive principle is actually the most useful and important conclusion of my faces research.
Your competitors aren’t chasing that type, which means your chance of success is higher - and when you do succeed, you’ll feel like you’re getting a great deal. Your relationship will be stickier too.
Most people don’t chase their least popular favorite intentionally. But as I explained in DISPELLING BEAUTY LIES, they often end up with the same result accidentally - after a painful period of trial and error. (Imagine if we could avoid most of that trial and error!)
Let’s sum up.
If you can figure out which faces are YOUR type but NOT everyone else’s type, then you’ll ALSO know the most promising partners to chase. The ones that give you BOTH a high chance of success AND a high chance of being an ideal match.
Unfortunately, my chart of facial types was WAY too simple to finish the job. So back then I concluded with the observation that we need a massive AI study of everyone’s facial preferences, incorporating millions of ratings, in order to put my theory into practice.
And that was where I left it, for a couple years.
Then a few months ago I received a surprise DM from one of my readers. @mathyouf had studied the theory - and decided to build the missing machine: an AI-driven app that gathers everyone’s facial preferences and reports on how common they are.
Now, to be clear, I wasn’t on the dev team. But I did go over some details of the theory with Matthew and helped beta test the app. I don’t even have time to explain in this thread all the ways it could help everyone. The potential is enormous.
It’s launching today - and it’s free. It’s called MyType. In the rest of this thread I’m going to explain how to use it to figure out who you should pursue to have the best chance at a happy relationship.
First, though, I want to share a link to my original theory and a link to the app.
Here’s the theory - “The Most Important Secret About Faces,” from DISPELLING BEAUTY LIES.
https://www.jsanilac.com/dispelling-beauty-lies/#vii-the-most-important-secret-about-faces
And here’s the app: MyType.
https://mytype.fyi/
Ok, now I’m going to go over the best way to use the app.
Basically, the app is going to show you sets of faces. All you have to do is pick your favorites.
The app suggests you go through twenty sets. That’s probably supposed to be encouraging, but based on my beta testing, it’s not nearly enough. I recommend around 150 sets.
150 sets sounds like a lot. But consider how complicated faces are. If you reject a great face because you hate the eyebrows, AI can’t read your mind to figure out the reason. The more sets you do, the better it can understand what you like and dislike.
Plus, this isn’t hard work. Choosing faces is fun.
Here are my own results. There’s a pretty obvious type, eh? Only one of them really doesn’t fit (which I crossed out). The ones I rate highly I circled in green.
When I did 100 more my results were even more uniform, with the same faces and facial structure recurring across ethnicities - a phenomenon I’d previously noted in DISPELLING BEAUTY LIES. (There are interesting ethnic breakdowns in the app, but I’m not going to discuss them here.)
You can compare the app results to my four favorites, selected long beforehand. Not bad. And for our purposes here, the AI doesn’t need to learn your EXACT favorites.
It’s fun to contrast these with the negatives. Pretty stark! I don’t know what you boys are thinking, but I’m gonna have to give long-faced blonde cheerleaders a hard pass.
Ok, now let’s talk about how to use the results page.
Figuring out that you have a type is actually NOT THE POINT. The point is to figure out the DIFFERENCE between your type and everyone else’s. THAT’S where the “relationship alpha” is.
The app doesn’t foreground this important part of my theory QUITE as much as I’d like yet but (maybe thanks to this thread) I suspect that could change in the future.
So, if you click through the faces on the results page it will tell you how highly everyone else rated them as well as how highly you rated them. What you’re looking for are GAPS.
These faces that I rated highly are super popular with everyone. They might be my type, but there’s no relationship alpha there.
However, this next group of faces that I rated highly are significantly LESS popular with the crowd. The gap between my rating and everyone else’s rating is a sign that I should be chasing these faces more.
Now, to be clear, I wouldn’t rate this set at “100%” in the usual sense. That only means these are faces I picked as especially pretty. The REALLY valuable knowledge here is that other people are less keen - and other people being less keen means MORE FOR ME.
Here are the LEAST popular faces I liked. Again, I wouldn’t rank these as the best faces out there or call them 99% “perfect,” but in theory they’re the ones that give me the maximum relationship alpha - because I like them a lot more than other people do. And now I know!
Ok, to be fair I need to note a few quibbles. Matthew’s done a great job on this app but it’s just launching and it’s not perfect yet.
The labels are not totally accurate. The app tells me my main likes are “indie film girl, art girl, dark hair, French it-girl, beautiful and sad,” and my main dislikes are “cheerleader, sorority girl, gym girl, coastal preppy.” Ok, fair enough, that’s all true. I don’t have big complaints about vintage Sophie Marceau.
However, you’ll inevitably notice some faces have the wrong labels. I rated these faces well, but I’m not convinced “heroin chic” is the best description of them? Fortunately, the labeling doesn’t actually matter. The labels are there for fun and don’t determine the value of your results.
A second quibble is that it’s not clear just how precise the crowd ratings are yet. I picked a few faces that look like Nanno of “The Girl From Nowhere.” That was a surprise to me as I didn’t realize I liked this face. But what was more surprising (to me) is that it gets fairly mid ratings from the crowd.
One of my earlier findings is that we’re bad at recognizing when faces we like aren’t liked by others. Is the Nanno type actually sorta mid in the eyes of the masses? It seems unlikely to me, but I honestly don’t know. And without this app, nobody else will know for sure either.
Precision is the kind of thing that will get better and better with time. Because the more people use the app, the more precise the numbers become. So, let’s get at it! And if you’re in doubt, check your results again later.
Well, I have a lot more to say about the potential here. I’ve only scratched the surface. But this thread is already long so I’m going to wrap up. Try the free MyType app and get your relationship alpha. And thanks to @mathyouf for making theory into reality!
I’ll post the two links again below.
Faces theory - “The Most Important Secret About Faces,” from DISPELLING BEAUTY LIES.
https://www.jsanilac.com/dispelling-beauty-lies/#vii-the-most-important-secret-about-faces
And the new MyType app!
https://mytype.fyi/
October 25, 2026
Today I’ve been thinking about the increasingly convincing AI fakes of… well, everything. It’s almost impossible to verify authenticity in the end-product anymore. Yet people still want to feel like they’re in contact with other people, and the arts and entertainment are an important way to make that contact.
Given the above, I think “signing” is going to matter more than it did in the past. And unusually, signing is something that can happen in degrees.
Obviously we know a painting can be signed. But what if you just have AI paint it and then sign it because you gave the instructions? Well, in a sense, I think the parts of a work are now going to be signed for authenticity too. Maybe this will work and maybe it won’t, but that is what people will wonder in various ways: how much of what I’m consuming comes from a person?
Artists always resisted this in the past. There were good reasons for that. You could do a big sales routine about your life but fail to deliver in terms of actual artistic quality. In effect, you could use your personal human narrative to obscure deficiencies in the art. So the most dedicated and demanding artists would always distance their lives from their work and try to ensure the work stood on its own. Well, I don’t think that’s going to work anymore.
This journal is a way of signing my work so that you can see it’s thoroughly tied to a person. In the past I wouldn’t be keen on that for the reasons just stated - and also because my characters are a long way from me and I don’t want you to confuse the two - but it’s the new reality; and I’d rather you know how much comes from me to you than assume you’re interacting, partly or mostly, with a machine.
October 24, 2026
Feeling quite lethargic today - I’m trying to see if cutting my coffee dose will improve my mood without impairing my productivity. Probably not, but I’m going to try anyway as I have really been unable to conjure up a positive mood for ages. Will be a bit off while I adapt, naturally.
Judging by correspondence, there seems to be some more interest in “The Pleasure Gap,” though it’s hard to tell exactly what’s going on there or where it’s coming from, as I don’t have analytics.
October 23, 2026
“The Pleasure Gap” has been garnering more attention, even though still very few have been willing to share it in public. I reviewed it again and made a number of minor improvements. While I remain extremely skeptical, it’s technically possible that, in a few exceptional cases, word-of-mouth is sometimes temporarily an actual thing for a very brief time. Either way, it would be a strange irony if this forthright piece (which I almost didn’t publish) is what finally ends up drawing people to my site.
October 22, 2026
Yesterday I spent the whole day on plotting for the new novel and its two planned sequels. Today I had a productive day and drafted more than a couple thousand words for chapter three, which is unusually fast for me, as I typically write very slowly. Probably it was just an easy sequence that will need revision. I am hoping if I string together a few more good days like this I will feel with it enough to write some music again too.
I made a brief appearance on social media, mainly because I was getting tired of the degree of total isolation. However, it is really very hard for me to say anything substantial in that format, and most of my followers there only care about my writing on beauty, so that’s what I’m used to posting about - even though I’ve spent hardly any time writing about it this year. I would like to post about other things, but 280 characters is useless, and long unread threads are too time consuming.
Have also been thinking about the sacred, but that is going to have to wait for another day as I’m afraid it would be too distracting to write about here at the moment.
October 20, 2026
Working on plotting for the novel today. Not much to say about that yet.
After a discussion with a friend today I reread the “Missing Chapters” and thought I had not written them very clearly. Well, that is baked in the cake by the fact that they are “missing.” I’m not going to take the time to perfect them. Nevertheless, I added the paragraphs below as a clarifying conclusion.
As an illustration of these points, look back over the last few entries. The “slutship paradox” is a useful explanation of some illusions presently damaging gender relations. But because people are dumb and have moreover had their virtues degraded or never developed in the first place (namely, they want to be angry and rush to judgment and do not bother trying to resist these impulses), an explanation of this sort has much less social reach than simpler and wronger alternatives which play more effectively on the emotions. Yes, this is always true to some degree, but degrees matter. In the end a society has the beliefs it deserves. And simple stupidity is not the only cause.
Brief Summary Of Generalized Ponerology
The evil have a competitive advantage over the good because they are less constrained by moral norms or concern for general welfare. By consequence, they eventually rise to the top of any social aggregate that is insufficiently defended against them. To increase their power, they then corrupt it further. This process is recursive. Although structure matters, a purely structural analysis cannot capture it and indeed dangerously obscures it, because it is first and foremost a hysteretic change in the character of the people occupying the structure.
Progressive infection by evil afflicts social aggregates of every type and level, whether they are formally organized or not. Its downstream effects include a decline in the accuracy of beliefs as well as a decline in values, virtues, and general prosperity, none of which can be sustainably reversed while the root cause remains. To escape this fate one must design and implement mechanisms that ensure the good have a persistent advantage, which is far more difficult than usually assumed, and impossible when (as now) the infection runs too deep.
October 18, 2026
Some quick notes on a topic Zoomers are always asking me to address.
The Slutship Paradox