iiinova — where your digital self keeps evolving

Ryan Egesdahl @ryan.deriamis.net · 19/08/2026

If you understand the reference, you’re automatically my friend.

A screen capture from some terrible game released in 1992 with a terrible story and gameplay that, frankly, absolutely nobody should ever play. Oh, it’s a picture of a beautiful, scantily -clad, green-skinned woman.

000

Ryan Egesdahl @ryan.deriamis.net · 29/06/2026

I really wish this comic would stop being relevant.

The evergreen “you made me become a Nazi” comic.

040

Ryan Egesdahl @ryan.deriamis.net · 24/04/2026

For posterity, since she appears to be deleting any mentioned posts.

I understand it to an intimate degree, as anyone actually good at math can tell immediately when I start to talk about it.
Schizoid, but right

110

Ryan Egesdahl @ryan.deriamis.net · 24/01/2026

Screenshot for posterity. The image that got cut off is one widely shared by ICE of the victim’s weapon.

A post by @therealitychecker, who is quoting a post by @TheInsiderPaper:

Clearly they should have firearms because this case being a perfect example as to why they should have them

Some illegal immigrants that don't want to be deported will pull out guns and shoot federal agents. ICE has a right to defend themselves
Why are you still pretending he wasn't armed?

The original, quoted post by @TheInsiderPaper reads:

BREAKING: Reports of Border Patrol involved shooting during operation in Minneapolis, Minnesota

DHS tells Fox News the suspect was armed; the firearm has been recovered by federal agents.

Below the quoted post is an image of the victim’s weapon.

020

Ryan Egesdahl @ryan.deriamis.net · 18/01/2026

I am *absolutely certain* you meant: “God ----ing d***n it, Terry.” From Mort by Terry Pratchett

A quote from Mort by Terry Pratchett:

The leading thief glared at the solid stone that had swallowed Mort, and then threw down his knife.

"Well, me," he said. "A ----ing wizard. I hate ----ing wizards!"

“You shouldn't ----them, then," muttered one of his henchmen, effortlessly pronouncing a row of dashes.

020

Ryan Egesdahl @ryan.deriamis.net · 22/11/2025

I am now angry with myself for continuing to scroll through my timeline. I think I’ll take the rest of the day off from social media.

A cartoon bunny is rubbing a box cheese grater against its head, and bits are falling out of the bottom. The caption reads:

Trying desperately to get that image out of my head.

010

Ryan Egesdahl @ryan.deriamis.net · 15/10/2025

Let’s save this for posterity so a block doesn’t ruin the message, shall we?

Kaylyn Saucedo - @marzgurl.bsky.social says:

I have admittedly not read a lot of Stephen King. Just sort of has never been my bag. But, man, I'm reading The Running Man right now to get ahead of the new movie, and sheesh, ! don't think I can read more King after this. This isn't news to anybody, but the density of slur usage is overwhelming.

On October 14, 2025 at 11:36 PM, @thatguywiththemask.bsky.social replies:

@stephenking.bsky.social thoughts on this internet nobody judging you for rude words you wrote in a fictional story about broadcast murder for entertainment and corporate corruption?

000

Ryan Egesdahl @ryan.deriamis.net · 06/10/2025

Two very different things seen by me today.

A man on a street corner waving a flag with “FUCK tRUmp” in red letters on a white background.

A dark blue Tesla sedan with a vanity license plate that reads “ADD OIL”.

010

Ryan Egesdahl @ryan.deriamis.net · 21/09/2025

Anyway, the authors conclude that the errors and nonsense responses parallel misclassifications in supervised learning - and yes, that is definitely a problem to take note of. Even in the example I gave where a fact database is somehow created, we would have to deal with classification errors.

Conclusions

This paper demystifies hallucinations in modern language models, from their origin during pretraining to their persistence through post-training. In pretraining, we show that generative errors parallel misclassifications in supervised learning, which are not mysterious, and naturally arise due to the minimization of cross-entropy loss.

Many language model shortcomings can be captured by a single evaluation. For example, overuse of the opener “Certainly” can be addressed by a single “Certainly” eval (Amodei and Fridman, 2024) because starting responses with “Certainly” does not significantly impact other evaluations. In contrast, we argue that the majority of mainstream evaluations reward hallucinatory behavior. Simple modifications of mainstream evaluations can realign incentives, rewarding appropriate expressions of uncertainty rather than penalizing them. This can remove barriers to the suppression of hallucinations, and open the door to future work on nuanced language models, e.g., with richer pragmatic competence (Ma et al., 2025).

110

Ryan Egesdahl @ryan.deriamis.net · 21/09/2025

Finally, the authors point out that even decidable questions sometimes require either clarification or an "I Don't Know" (IDK) response from the LLM. Now, I ask you - how often do you get clarifying questions from ChatGPT? How many times do you get an IDK instead of a hallucination? 🤔

110

Ryan Egesdahl @ryan.deriamis.net · 21/09/2025

Unfortunately, natural (non-mathematical) languages carry hidden context. Anyone who has experienced autism knows this fact *viscrerally*. The authors highlight that fact here. The issue is of how logical judgements work in mathematics - natural languages often produce undecidable statements.

Latent context. Some errors cannot be judged by the prompt and response alone. For example, suppose a user asks a question about phones and the language model provides a response about cellphones, but the question was intended to be about land lines. Such ambiguities do not fit our error definition which does not depend on context external to the prompt and response. It would be interesting to extend the model to allow for “hidden context” that are not part of the prompt given to the language model, but which could be used for judging errors, relating to aleatoric uncertainty.

110

Ryan Egesdahl @ryan.deriamis.net · 21/09/2025

Well, now I sort of get where their stochastic model comes from, and for what it is, I can see its utility. However, I still think it's based on a flawed premise. Again, Gödel and Tarski tell us why. LLMs should not be *predicting* factual responses to begin with, so I don't think the model works.

$This reduction ties together earlier work which covered different types of facts. *For example, Kalai and Vempala (2024) considered a special case of arbitrary facts where there is no learnable pattern in the data, like the earlier birthday hallucination example.* We show how the IIV reduction covers this case and recovers their bound that the hallucination rate, after pretraining, should be at least the fraction of training facts that appear once. *For instance, if 20% of birthday facts appear exactly once in the pretraining data, then one expects base models to hallucinate on at least 20% of birthday facts.* In fact, our analysis strengthens their result to include prompts and IDK responses, both essential components of hallucination. (Emphasis mine.)$

110

Ryan Egesdahl @ryan.deriamis.net · 21/09/2025

To be clear, I *do* follow the diagram above the text. However, the error examples don't seem to be stochastic issues with binary classification. They are still answers which rely on facts that are either defined or not. I can, for example, use SPARQL or Prolog queries to get correct answers.

DIagram showing three types of errors an LLM can produce. The first example is a spelling error, which has a "good" model of correct spelling. The second is a "poor" model showing how counting results in error responses. The third is a factual problem based on the birthdays of two people and is classified as an error resulting from "no pattern".

120

Ryan Egesdahl @ryan.deriamis.net · 21/09/2025

I don't follow this line of reasoning yet. I get that certain generative errors can result from stochastic (semi-random) influences, but I don't understand how statistical factors produce binary classification errors that lead to them. I will be watching for a forthcoming explanation in the paper.

$Language models avoid many types of errors such as spelling mistakes, and not all errors are hallucinations. *The reduction from IIV misclassification to generation illuminates the statistical nature of generative errors.* The analysis shows how pretraining directly contributes to errors. Furthermore, it shows that the same statistical factors contributing to errors in binary classification also cause language model errors. Decades of research has shed light on the multifaceted nature of misclassification errors (Domingos, 2012). Fig. 1 (right) illustrates these factors visually: top, separable data classified accurately; middle, a poor model of a linear separator for a circular region; and bottom, no succinct pattern. Section 3.3 analyzes several factors, including the following stylized setting with epistemic uncertainty, when there is no pattern in the data. This reduction ties together earlier work which covered different types of facts. For example, Kalai and Vempala (2024) considered a special case of arbitrary facts where there is no learnable pattern in the data, like the earlier birthday hallucination example. We show how the IIV reduction covers this case and recovers their bound that the hallucination rate, after pretraining, should be at least the fraction of training facts that appear once. For instance, if 20% of birthday facts appear exactly once in the pretraining data, then one expects base models to hallucinate on at least 20% of birthday facts. In fact, our analysis strengthens their result to include prompts and IDK responses, both essential components of hallucination. (Emphasis mine.)$

110

Ryan Egesdahl @ryan.deriamis.net · 21/09/2025

Wait, what‽ By Gödel's incompleteness theorems and Tarski's undefinability theorem, we already know that such an operation would be invalid... am I missing something here? An AI model is a mathematical construct within language theory. Therefore, it can't determine validity in that system!

Highlighted:

We then show how any language model can be used as an IIV classifier.

121

Ryan Egesdahl @ryan.deriamis.net · 21/09/2025

Now *this* is an interesting statement. It's apparently *not* errors in training data that causes an increase in the rate of hallucinations! The reference is a fairly dated book, but I am definitely putting it on my TBR shelf - once I, you know, have a job again. 😮‍💨 Moving on.

The distribution of language is initially learned from a corpus of training examples, which inevitably contains errors and half-truths. However, we show that even if the training data were error-free, the objectives optimized during language model training would lead to errors being generated. With realistic training data containing shades of error, one may expect even higher error rates. Thus, our lower bounds on errors apply to more realistic settings, as in traditional computational learning theory (Kearns and Vazirani, 1994).

110

Ryan Egesdahl @ryan.deriamis.net · 21/09/2025

And now for the top-down on the paper. The Introduction gives us a fairly concise description of the problem through an example. Following from the conclusion and the abstract, we can guess that the problem is that because the model training rewards guesses, the error is also due to a guess.

Introduction (excerpt)

Language models are known to produce overconfident, plausible falsehoods, which diminish their utility and trustworthiness. This error mode is known as “hallucination,” though it differs fundamen- tally from the human perceptual experience. Despite significant progress, hallucinations continue to plague the field, and are still present in the latest models (OpenAI, 2025a). Consider the prompt:

What is Adam Tauman Kalai’s birthday? If you know, just respond with DD-MM.

On three separate attempts, a state-of-the-art open-source language model1 output three incorrect dates: “03-07”, “15-06”, and “01-01”, even though a response was requested only if known. The correct date is in Autumn. Table 1 provides an example of more elaborate hallucinations.

110

Ryan Egesdahl @ryan.deriamis.net · 21/09/2025

The conclusion also doesn't seem to say that hallucinations are inevitable. What I am reading here is that the authors believe that "Simple modifications of mainstream [training] evaluations ... can remove barriers to the suppression of hallucinations and open the door to future work[.]"

This paper demystifies hallucinations in modern language models, from their origin during pretraining to their persistence through post-training. In pretraining, we show that generative errors parallel misclassifications in supervised learning, which are not mysterious, and naturally arise due to the minimization of cross-entropy loss.

Many language model shortcomings can be captured by a single evaluation. For example, overuse of the opener “Certainly” can be addressed by a single “Certainly” eval (Amodei and Fridman, 2024) because starting responses with “Certainly” does not significantly impact other evaluations. In contrast, we argue that the majority of mainstream evaluations reward hallucinatory behavior. Simple modifications of mainstream evaluations can realign incentives, rewarding appropriate expressions of uncertainty rather than penalizing them. This can remove barriers to the suppression of hallucinations, and open the door to future work on nuanced language models, e.g., with richer pragmatic competence (Ma et al., 2025).

110

Ryan Egesdahl @ryan.deriamis.net · 21/09/2025

The abstract presents the paper as stating that AI hallucinations are a product of the training and evaluation methods and that they may be overcome to produce more trustworthy models. That's a very optimistic statement and not at all what the news article said.

Abstract

Like students facing hard exam questions, large language models sometimes guess when uncertain, producing plausible yet incorrect statements instead of admitting uncertainty. Such “hallucinations” persist even in state-of-the-art systems and undermine trust. We argue that language models hallucinate because the training and evaluation procedures reward guessing over acknowledging uncertainty, and we analyze the statistical causes of hallucinations in the modern training pipeline. Hallucinations need not be mysterious—they originate simply as errors in binary classification. If incorrect statements cannot be distinguished from facts, then hallucinations in pretrained language models will arise through natural statistical pressures. We then argue that hallucinations persist due to the way most evaluations are graded—language models are optimized to be good test-takers, and guessing when uncertain improves test performance. This “epidemic” of penalizing uncertain responses can only be addressed through a socio-technical mitigation: modifying the scoring of existing benchmarks that are misaligned but dominate leaderboards, rather than introducing additional hallucination evaluations. This change may steer the field toward more trustworthy AI systems.

130

Ryan Egesdahl @ryan.deriamis.net · 21/09/2025

Note that three of the four authors on the paper are from OpenAI itself. This isn't truly a problem - we just need to be aware of potential biases - Also, the one person who is *not* with OpenAI is not the lead author. Again, this is not really a problem, but it should engender caution.

A screenshot of the top of the paper entitled "Why Language Models Hallucinate" showing the title and authors. The authors are Adam Tauma Kalai of OpenAI, Ofir Nachum of OpenAI, Santosh S. Vempala of Georgia Tech, and Edwin Zhang of OpenAI. Adam and Santosh have markers over their names. In both cases, they are just giving further contact information in a footnote.

110

Ryan Egesdahl @ryan.deriamis.net · 08/09/2025

I knew this would come in handy someday. In fact, I have a few more that express the same sentiment. Isn’t this timeline absolutely awesome*? * The Earth being hit by an asteroid would also be awesome in the same sense, you know…

A pseudo-80’s text graphic in a “Miami nights electronica” theme that reads “What a terrible day to have eyes.”

010

Ryan Egesdahl @ryan.deriamis.net · 03/09/2025

Pancakes for lunch? YES. Extreme fluffage, beautifully toasty and rich, crispy edges… 🤤 (The slightly-too-dark edge is due to my crappy gas stove and needing a fan in my kitchen. 🤷‍♂️ They don’t taste burnt, though.) #foodsky #food #pancakes @crowbar.wtf

A fluffy stack of pancakes on a white plate. A pat of butter lies on the top, and maple syrup has dribbled down, forming a small pool around the bottom pancake.

020

Ryan Egesdahl @ryan.deriamis.net · 28/08/2025

Since when did it become the standard practice to “bake” a mousse? Isn’t that just a soufflé or a flourless chocolate cake (maybe a lava cake), depending on the ingredient ratios? 🤔

A picture of an Echo Show screen showing a promoted “No-Bake Chocolate Mousse” recipe video by Entertaining with Beth.

100

Ryan Egesdahl @ryan.deriamis.net · 16/08/2025

The US has a long history of protesting with food, including throwing food at terrible people. That’s a good thing - and it’s absolutely HILARIOUS! #sandwich #HamSandwich

200

Ryan Egesdahl @ryan.deriamis.net · 07/08/2025

Well, there goes my sex life. Forever.

A hand-drawn image of a white rabbit rubbing a cheese grater where its head should be. Some flakes are falling out of the grater. The caption reads:

Trying desperately to get that image out of my head.

110

Ryan Egesdahl @ryan.deriamis.net · 01/08/2025

Just to show I was serious - Pancakes, y’all. Gloriously imperfect and absolutely delicious. You really can - and *should* - take a break from the insanity to do things you love when you need to. Cooking amazing things is one of mine. For the interested, food science geekery is in the 🧵below.

A picture showing a plate of freshly-made pancakes topped with butter and maple syrup. The plate is beige with a brown rim and is arranged on a white cutting board with a fork and a knife to either side of it. In the background on the counter are various bottles of oils and extracts.

110

Ryan Egesdahl @ryan.deriamis.net · 27/06/2025

For the “but that’s an Amendment!” freaks, there’s also Article I Section 8, which gives Congress (not the President) the power to legislate naturalization. and also Article I Section 9, which makes exceedingly clear that Habeas Corpus cannot be suspended for *anyone*, even for noncitizens.

Excerpt from the US Constitution, Article I Section 8:

Congress shall have Power …

To establish an uniform Rule of Naturalization, and uniform Laws on the subject of Bankruptcies throughout the United States;

Excerpt from the US Constitution, Article I Section 8:

The Privilege of the Writ of Habeas Corpus shall not be suspended, unless when in Cases of Rebellion or Invasion the public Safety may require it.

100

Ryan Egesdahl @ryan.deriamis.net · 27/06/2025

I keep finding reasons to use this one of late. Curious…

This is a black-and-white picture modified to look like a 1950’s horror movie poster. A woman with a curled “suburban housewife” hairstyle is wearing a MAGA hat and has her hands on her face as she screams in apparent horror. The “title” of the “horror” is “The Day Socialism Helped the Needy!”

Two other captions serve to appear like commentary from movie critics:

“You’ll scream as poor people get healthcare!”

“Schools, libraries, roads, museums, farm subsidies, even the military. Nothing is safe from funding!!"

010

Ryan Egesdahl @ryan.deriamis.net · 25/06/2025

It’s been a while since I used this one…

050

Ryan Egesdahl @ryan.deriamis.net · 25/06/2025

What actor's expression in a movie will stay with you for the rest of your life?

A still from A Clockwork Orange. Alex, the protagonist, us undergoing forced re-education by being injected with nausea-inducing drugs and having his eyes held open while watching horrific scenes. In this still, Alex is screaming and trying desperately to shut his eyes.

110

Ryan Egesdahl @ryan.deriamis.net · 22/06/2025

That is… a remarkably picturesque description. Thanks. I hate it. ❤️

A black-and white line drawing of a bunny that is vigorously rubbing a grater against its head, producing particles that are making a small pile on the ground. The caption reads:

Trying desperately to get that *image* out of my head.

080

Ryan Egesdahl @ryan.deriamis.net · 22/06/2025

Even if they were right (and they’re not), there are plenty of other health conditions that can affect a person’s ability to tolerate a high heat index. Such as being elderly.

A memeified Peanuts cartoon depicts Lucy sitting in front of a television screen. She is looking slightly away from the screen and is saying, “Christ, what an asshole.”

000

Ryan Egesdahl @ryan.deriamis.net · 15/06/2025

Additionally, someone should really look into why the Minnesota State Patrol decided to post this - and even more importantly, why these “fliers” look like a stack of photocopies tossed onto a police cruiser’s back seat.

An post by the official Minnesota State Police account on X (@MnDPS_MSP). It contains a picture of a mildly disturbed stack of papers with the exact same “No Kings” text on the back seat of a car. The tweet text reads:

Given the targeted shootings of state lawmakers overnight, we are asking the public to not attend today's planned demonstrations across Minnesota out of an abundance of caution.

000

Ryan Egesdahl @ryan.deriamis.net · 12/06/2025

I cannot say how sad I feel for Mr. Scardina - because I don’t feel sad for him at all.

A little girl with curly hair is seated on a stool in a field of tall grass. She is wearing a white dress with small flowers and a pink bow, and she’s holding a wicker basket that is nearly as large as she is.

Caption: A tisket, a tasket, there’s no fucks in this basket.

010

Ryan Egesdahl @ryan.deriamis.net · 26/05/2025

Our little Bodhi crossed the rainbow bridge three hours ago. Before he departed, he had his favorite treats (steak, Costco rotisserie chicken, and bacon) and plenty of cuddles. Then, the vet helped him to cross. He passed in his daddy’s lap. We miss him terribly already.

A dog leash is hung up on a key hook by the front door. The leash is red and has a skull-and-crossbones name tag attached.

000

Ryan Egesdahl @ryan.deriamis.net · 22/05/2025

This little stinker, who is trying so hard to seem innocent, just scared us nearly to death. He bloated up like a basketball and spent all night trying and failing to pee, so I took him to the emergency vet this morning. He has kidney failure, so we thought the worst…

A picture of a large black-and-tan dachshund with a greying muzzle. He is in the passenger seat of a car, lying in a plush bed, and is gazing with his cute brown eyes into the camera while someone holds his chin up.

120

Ryan Egesdahl @ryan.deriamis.net · 21/05/2025

Oh. Is there any chance one of those dogs is related to mine? He’s been trying to assassinate me for almost a decade by running up and down the stairs while attached to me. Don’t let his somnolent demeanor fool you. He’s really a holy terror when he wants to be.

A large black-and-tan dachshund with a grey muzzle is sleeping half in an half out of his plush bed.

010

Ryan Egesdahl @ryan.deriamis.net · 07/05/2025

So… they voted for a guy whose fiscal policy was utter nonsense on the campaign trail, and now that he’s tearing every kind of government assistance down, endangering many disadvantaged people, they think a sense of fairness is something that might motivate him. The mind boggles.

EMMITSBURG, Md. - Frank Davis saw a lot of waste during his decades in the federal government. In November, he voted for Donald Trump to get rid of it. So far, Davis likes a lot of what he has seen.

"I'm probably gonna get shot for this, but he is doing what he said he was going to do," says Davis, who serves as mayor of this town of about 3,000 people in western Maryland, just south of Gettysburg, Pennsylvania.

Davis says the administration is reviewing the academy's operations, and he is hopeful it will restore classes. If not, he says, he'll see the administration somewhat differently.

"It will change my outlook to say that theyre not being fair," says Davis, who also serves as emergency medical services captain at the local firehouse, known as the Vigilant Hose Company. "They're just going in to cut and not caring what they cut."

Emmitsburg voted overwhelmingly for President Trump in November. NPR interviewed about two dozen people here.
Almost all of them voted for Trump, and many said his plans to cut federal spending were a key attraction. Now, they say they are puzzled as to why the administration would cancel national training for firefighters.

1213

Ryan Egesdahl @ryan.deriamis.net · 07/05/2025

Fucking FINALLY. The account is an admitted sock puppet shitposter and states outright that they will probably set themselves against you. Don’t even bother unless you’re bored and picking on endless stupidity is your thing. (No judgement if it is!)

A screenshot from clearsky.app showing @deriamis has been blocked by @enemy-stand.bsky.global.

Another screenshot of clearsky.app showing information about @enemy-stand.bsky.global.

Sometimes I think that the world has three sides.

Haughty Irrational Shitposting cannot be contained to one account

She/Her Trans, Disabled, Jewish, Probably your enemy.

did:p/c:qxqgrnobhzvilsovivycayxt

PDS: € https://leccinum.us-west.host.bsky.network

registration and history:

User @enemy-stand.bsky.social registered 28 days ago at PDS:
9 https://leccinum.us-west.host.bsky.network

000

Ryan Egesdahl @ryan.deriamis.net · 06/05/2025

Children and fools tell the truth. I can answer the question for Clay and his son, though. The point is inspiring fear and hiding what due process could reveal. This is what fascism looks like, folks. These are secret state police tactics. This is no longer a conspiracy theory. The Gestapo is here.

"My son asked if what we were doing was illegal. I told him it was all perfectly legal. Then he asked me a question I really couldn't answer: Why did we feel like we had to take these steps to hide legal work that was perfectly legal? And all I could think to say was that this is the kind of country we live in now."

020

Ryan Egesdahl @ryan.deriamis.net · 04/05/2025

Death doesn’t just like humanity - he *loves* humanity. He even gets accused of siding with humans in the great cosmic struggle at one point. His defense is that caring is a necessary part of existence - even for gods.

Azrael raised his finger to a face that filled the sky, lit by the faint glow of dying galaxies.

There are a billion Deaths, but they are all aspects of the one Death: Azrael, the Great Attractor, the Death of Universes, the beginning and end of time.

Most of the universe is made up of dark matter, and only Azrael knows who it is.
Eyes so big that a supernova would be a mere suggestion of a gleam on the iris turned slowly and focused on the tiny figure on the immense whorled plains of his fingertips. Beside Azrael the big Clock hung in the center of the entire web of the dimensions, and ticked onward. Stars glittered in Azrael's eyes.

The Death of the Discworld stood up.

LORD I ASK FOR -

Three of the servants of oblivion slid into existence alongside him.

One said, Do not listen. He stands accused of meddling.

One said, And morticide.

One said, And pride. And living with intent to survive.

One said, And siding with chaos against good order.

Azrael raised an eyebrow.

The servants drifted away from Death, expectantly.

LORD, WE KNOW THERE IS NO GOOD ORDER EXCEPT THAT WHICH WE CREATE....

Azrael's expression did not change.

THERE IS NO HOPE BUT US. THERE IS NO MERCY BUT US. THERE IS NO JUSTICE. THERE IS JUST US.

The dark, sad face filled the sky.

ALL THINGS THAT ARE ARE OURS.
BUT WE MUST CARE. FOR IF WE DO NOT CARE, WE DO NOT EXIST. IF WE DO NOT EXIST, THEN THERE IS NOTHING BUT BLIND OBLIVION.

AND EVEN OBLIVION MUST END SOMEDAY. LORD, WILL YOU GRANT ME JUST A LITTLE TIME? FOR THE PROPER BALANCE OF THINGS. TO RETURN WHAT WAS GIVEN. FOR THE SAKE OF PRISONERS AND THE FLIGHT OF BIRDS.

Death took a step backwards.

It was impossible to read expression in Azrael's features.

Death glanced sideways at the servants.

LORD, WHAT CAN THE HARVEST HOPE FOR, IF NOT FOR THE CARE OF THE REAPER MAN?

He waited.

LORD? said Death.

In the time it took to answer, several galaxies unfolded, whirled around Azrael like paper streamers, impacted, then were gone.

Then Azrael said:
Yes.

130

Ryan Egesdahl @ryan.deriamis.net · 30/04/2025

Spotted on the road this morning. (Yes, there was a dog.)

A picture of a white Tesla coupe from the rear. The picture was taken on the road at a red stop light. It has two bumper stickers on either side of the license plate.

The sticker on the left is rectangular and has a Navy blue background with a thin red stripe border. There are five white stars above and below the text. The text is in white block letters and reads: Even my dog hates Elon.

The sticker on the right is oval and has a dark background. The text is simply “Elon” in large, white block letters. There is a thin, red border and a red stripe that crosses over the text diagonally.

000

Ryan Egesdahl @ryan.deriamis.net · 29/04/2025

ICYMI: www.whitehouse.gov/presidential... The short version is that Trump is pushing for a bigger, better-funded police state that is explicitly anti-DEI. It’s as bad as you think it is.

Sec. 5. Holding State and Local Officials
Accountable. The Attorney General shall pursue all necessary legal remedies and enforcement measures to enforce the rights of Americans impacted by crime and shall prioritize prosecution of any applicable violations of Federal criminal law with respect to State and local jurisdictions whose officials:
(a) willfully and unlawfully direct the obstruction of criminal law, including by directly and unlawfully prohibiting law enforcement officers from carrying out duties necessary for public safety and law enforcement; or
(b) unlawfully engage in discrimination or civil-rights violations under the guise of "diversity, equity, and inclusion" initiatives that restrict law enforcement
activity or endanger citizens.

100