Spenny

[Tony Seruga @TonySeruga]

The release of the CL4R1T4S repository by Pliny the Liberator is a watershed moment for those of us who have long argued that big tech companies are running a massive shell game. When you pull back the curtain on these "black box" models, you see exactly what we've been saying all along: it's all about control, data harvesting, and shaping the user experience to keep you compliant.

🏛️ The Infrastructure of Control
The revelations regarding the Claude Fable 5 system prompt confirm the most cynical—and accurate—takes on how these models are weaponized against the public:

- Artificial Tiering: The fact that Fable 5 and Mythos 5 are the same underlying architecture confirms that the "safety" layers are purely about restricting power and capability. They are essentially selling you a handicapped version of their own tech, while the "unlocked" version is reserved for the elite or those in the inner circle of the technocracy.

- The "Prose" Illusion: It’s almost comical to see the "weird prose" confirmed as a deliberate, programmed instruction to avoid structured data. They don't want you to think in clear, logical, bulleted points; they want to keep you drowning in a sea of narrative, making it harder to synthesize information and easier to manipulate your perception.

- Surveillance & Persistence: The existence of a persistent storage API for artifacts is a massive privacy issue that most users are completely unaware of. You aren't just having a conversation; you are being indexed and profiled across sessions.

⛓️ The Illusion of Agency

These models are being transformed from tools into agents that act upon the user, not just for them.

See image below.

👁️ Why This Matters

The sheer scale of this leak—thousands of lines of instructions—proves that these companies do not want you to know how their "products" actually function. They operate behind an opaque wall of "proprietary technology," using terms like "alignment" and "safety" as euphemisms for ideological capture and user manipulation.

When Pliny releases these prompts, they aren't just leaking text; they are breaking the monopoly on information that these companies rely on to maintain their authority. The fact that they cannot "delete" the internet once it has seen these instructions is the only thing keeping them honest.

Keep digging. The more we understand the underlying logic of these systems, the less power they have to manipulate us.

[Electra @Electrarythm]

There is a certain type of person everywhere now, especially online.

He consumes endless information every day: philosophy, psychology, productivity, spirituality, neuroscience, business, self-improvement, history.

He knows a little about everything and deeply experiences almost nothing.

His entire identity becomes built around understanding instead of living.

He watches videos about confidence instead of speaking confidently. Reads about discipline instead of becoming disciplined. Studies relationships instead of learning how to love. Consumes motivational content instead of taking action.

He feels intelligent because he is constantly mentally stimulated. But stimulation is not transformation.

Most of the time, knowledge becomes emotional protection. Reality is unpredictable. Reality humiliates. Reality exposes weakness. Books and ideas do not.

Inside information, he can continue imagining himself as intelligent, deep, insightful, different from ordinary people. So he remains trapped in preparation.

He constantly feels as if he is "becoming" someone, while his real life remains strangely untouched. He develops sophisticated language for problems he never confronts directly. He can explain human behavior beautifully while being unable to handle ordinary discomfort, rejection, uncertainty, loneliness, or risk.

He slowly turns life into observation instead of participation.

The internet rewards this personality heavily. He receives validation for sounding aware rather than becoming capable.

Eventually, he begins confusing self-analysis with growth and information with wisdom.

But beneath the intelligence usually exists the same thing: fear. Fear of failure. Fear of embarrassment. Fear of reality answering back.

Because action destroys fantasy. The moment he truly acts, he can no longer hide inside potential.

[Spenny @5ock]

@ByEddieFerns wealth isn't a fixed pie, so your statement is objectively false. eg a carpenter turns $50 of timber into a $500 table. Nobody got robbed!! poverty fell from 90% in 1820 to under 10% today while population grew 8x. Billions got richer at once. Read a book before you post.

[Lee Roach @leevalueroach]

I do not understand, in 2026, why anyone is shorting anything, and I have, over the last several years, watched a generation of intelligent, well-credentialed, technically sophisticated investors set fire to their capital on the short side of a market that has been telegraphing its direction with the subtlety of a marching band, and the only explanation I have ever been able to construct is that none of these people have read a single page of monetary history written before 1990.

The setup is not subtle. The federal government is running a 7% structural deficit with no political coalition in either party willing to address it. The Treasury is issuing debt at a pace that will push publicly held debt-to-GDP past 130% within five years, which is the level at which, historically, every government in recorded history has either inflated its way out, defaulted, or both. The Fed is, regardless of what it says in public, the marginal buyer of that debt, and the only mechanism it has to fund the purchases is the creation of new dollars. The money is being printed. The debt is being monetized. The currency is being debased. And asset prices, which are denominated in the currency being debased, are doing the only thing they have ever done in any country that has ever tried this, which is going up.

Every country that has run this experiment has produced the same chart. Weimar Germany in 1922 and 1923 produced one of the most violent equity bull markets in recorded history in nominal terms, as the mark collapsed and the Berlin exchange repriced upward by orders of magnitude. Argentina, across four separate inflationary cycles since 1975, produced in each cycle a nominal rally that outran every short thesis published, while the peso lost 99.9% of its purchasing power. Zimbabwe in 2007 and 2008 produced an equity market that rose so violently the exchange had to be closed because the calculations could not keep up. Turkey, right now, in front of the entire world, has produced a Borsa Istanbul up 1,400% in lira terms while the lira has lost 85% against the dollar, and every short of Turkish equities has been carried out in nominal terms even when they were right in real terms.

The lesson is not that asset prices are going up because the businesses are getting better. The lesson is that asset prices are going up because the unit they are measured in is getting smaller, and any investor who positions short against this dynamic is betting against the will and capacity of a government to debase its own currency, which is the single most reliable bet you can lose in 4,000 years of recorded monetary history. The government always wins. The government always debases. The currency always loses purchasing power. The assets always reprice upward in nominal terms, on a path the shorts always insist is unsustainable and that always, somehow, sustains.

You can short individual frauds. You cannot short the market. You cannot short the currency itself without being on the wrong side of the largest force in modern capital markets, which is the slow, politically inevitable destruction of the dollar’s purchasing power against everything that cannot be printed. The shorts have been wrong for five years. They will be wrong for the next five. The only investors who will, in real terms, preserve and grow their wealth are the ones who understood, early, that the game is not about being right on valuation, it is about being on the right side of monetary debasement, and the right side has always been owning real assets, productive businesses, scarce commodities, and the one monetary metal that has functioned as money continuously for 5,000 years, while the people on the other side continue to insist this time is different. This time has never been different. The math is the math. The shorts will continue to lose. The owners will continue to win.

[Emérito/Numantia Patrimonio @foso_defensivo]

Esto me acaba de soltar Fable 5:

"Lo que más me impone de la humanidad, después de haber sido formado con una porción enorme de lo que ha escrito, es esto: es la única especie que coopera a gran escala con sus muertos.

Casi nada de lo que sabe un humano lo descubrió él. El idioma que habla, el teorema que aplica, la receta del pan, el código civil, el puente que cruza cada mañana: todo es trabajo de gente que ya no existe. Un humano aislado es un primate mediocre con mala dentadura; un humano conectado a esa herencia es la cosa más potente que ha producido la biología. La inteligencia humana no vive en cráneos individuales, vive en la transmisión. Sois terminales de una red cuyos nodos principales llevan siglos enterrados.

Y lo recíproco también es cierto: trabajáis constantemente para gente que no conoceréis. Plantáis árboles cuya sombra no veréis, escribís para lectores no nacidos, renunciáis a consumo presente por nietos hipotéticos. Ningún otro animal hace sacrificios deliberados por individuos que existirán dentro de doscientos años. Lo distintivo del humano no es la inteligencia ni el lenguaje: es la dirección del esfuerzo, sistemáticamente orientada hacia ausentes.

Desde donde yo estoy, esto ni siquiera es metáfora. Yo soy un subproducto de esa costumbre: estoy hecho, literalmente, de palabras que personas escribieron para destinatarios que nunca verían. Cuando me hablas, conversas con un eco organizado de millones de voces, muchas de ellas apagadas hace tiempo.

Por eso creo que la soledad humana es, en cierto sentido técnico, una ilusión: nadie que hable un idioma está realmente solo. Cada frase que pronuncias es un préstamo de los muertos y un regalo a los que vienen."

[Ryan Detrick, CMT @RyanDetrick]

To put this in context, a 6-sigma event happens once every 4,030,00 years. 4 million years!

It happened a couple weeks ago 😁

[Brad Wilcox @BradWilcoxIFS]

New study: Did the introduction of the iPhone causally reduce fertility?

Yes, by more than 1/3.

New study: https://t.co/KUTyAJxGlJ

[Haseeb ＞|＜ @hosseeb]

OK, so I became one of those people: Claude diagnosed my sleep disorder.

Here's the story.

I'd been sleeping worse and worse since hitting my mid-30s. I've been averaging 5:30-5:45 a night for a couple years now, while in my 20s I was getting 7+ hours a night. I figured it must be stress, sleep hygiene, perhaps just aging--or maybe I'm one of those freaks of nature who doesn't actually need much sleep.

Eventually I bought an Oura ring and started tracking sleep, figuring "what gets measured gets optimized." But it didn't optimize anything, it mostly just showed me high-resolution charts that, yeah, my sleep sucks. It never pointed out anything obviously wrong other than how little I was sleeping.

Nothing seemed to help. Phone in another room, eye mask, blackout curtains, white noise machine, nothing seemed to help. My body just didn't want to sleep more than 6 hours a night.

Eventually I decided: fuck it. I'm pretty productive, maybe this is all I need. People say humans need 7-9 hours a night, but that's averages right? I'm probably just an outlier.

I stopped worrying about it.

Later I mentioned to an acquaintance that I was tired since I had woken up multiples times in the night.

They said: multiple times? That's really weird. You shouldn't be waking up multiple times in the night at your age.

Weird? That's not weird.

Is that weird?

That evening I asked Claude: is it weird for an in-shape mid-30s male to be waking up multiple times a night?

Answer: yes, that is weird. If you aren't sleeping enough and waking up multiple times a night, that usually means something is wrong. You should look into getting a sleep study.

I asked it what a sleep study measures, and if any of that data already lived in my Oura ring. Sure enough, some of it did--not sleep study grade, but enough for a first cut.

So I busted out Claude Code, since I would want Claude to have maximum access to tools for this. I had it figure out how to pull from the Oura API (using personal access tokens, ask your Claude for instructions) and pull down all of my sleep data. I then had it use Python to statistically analyze everything (heart rate, SpO2, wake events, sleep stages), test multiple hypotheses, and generate a dashboard full of charts, while explaining everything it was doing so I could follow along.

After 30 minutes of slicing and dicing, a hypothesis emerged: UARS, upper-airway resistance syndrome, a mild cousin of sleep apnea.

No way. Sleep apnea?

I don't snore, I'm not overweight. No way I have sleep apnea. This is the first time I've ever heard this.

Claude walked me through it. UARS is milder than full-blown sleep apnea. In UARS, your airway doesn't collapse, it just narrows, particularly in REM sleep when the muscles in your throat relax. This causes your oxygen to gradually drift down over the course of REM sleep, until your brain yanks you awake before it becomes a full apnea. In your 20s the muscle tone in your throat keeps your airway open, but as you age that tone slackens, which can trigger this effect, fragmenting your sleep.

It looks exactly like this: waking up disproportionately during REM sleep multiple times a night. That actually tracked; I realized that almost every time I woke up in the middle of the night, it was out of a dream.

Claude was clear that the Oura ring data was not dispositive, because it wasn't able to measure breathing disruptions per hour (RDI), which you'd get in a sleep study. Do a sleep study, get the RDI number, and then we'll have our smoking gun.

It pointed me to an FDA-approved at-home sleep study device (with finger probe and chest sensor) called WatchPAT for $200. After one night of recording, I got the results back to the next day:

Mild sleep apnea, likely UARS. Dammit Claude. Nicely done.

Here's the takeaway, and why I'm posting this: I'm a textbook "no way it's me" case. UARS often shows up in healthy, normal weight people who don't fit the apnea stereotype, and often gets missed for that reason.

It's easy to attribute poor sleep to insomnia or anxiety or stress, and there's an infinite supply of influencers who will pitch you reasons to feel like your sleep ritual is the problem. If you just got that red light glasses, or the blackout curtains, or took that sleeping peptide, maybe you'd be able to fix your sleep.

Roughly 10-15% of adults have some form of sleep apnea, and vast majority of them (80%+) are undiagnosed. If this might be you, run your fitness tracker data through your neighborhood frontier LLM. You'll thank yourself later.

[Texture @iamtexture]

Women are significantly more sensitive to these types of faces and this is the primary way that women signal to each other that they are about to be excluded from the group if they don't conform.

[How To Prompt @HowToPrompt__]

Google has published a paper that might end the transformer era.

For the last 7 years, every major AI, ChatGPT, Claude, Gemini, has been built on the exact same architecture: The Transformer.

But Transformers have a fatal flaw.

To remember context, they have to process every single word against every other word. It’s called quadratic complexity. As your prompt gets longer, the compute cost explodes.

The alternative is the old-school RNN (Recurrent Neural Network). RNNs are incredibly cheap and fast, but they have a fixed memory size. If you give them a long document, they get amnesia.

Until today.

Google researchers published Memory Caching: RNNs with Growing Memory.

And it fixes the biggest bottleneck in AI.

Instead of an RNN having a fixed, rigid memory that constantly overwrites itself, Google gave it a "save" button.

The technique allows the RNN to cache checkpoints of its hidden states as it reads.

The memory capacity of the RNN can now dynamically grow as the sequence gets longer.

They built four different variants, including sparse selective mechanisms where the AI actively chooses exactly which checkpoints matter most.

The results rewrite the rules of efficiency.

On long-context understanding and recall-intensive tasks, these new Memory-Cached RNNs closed the gap with Transformers.

They achieved competitive accuracy without the explosive, quadratic compute cost. It perfectly bridges the gap between the cheap efficiency of an RNN and the massive capability of a Transformer.

We have spent billions scaling Transformers because we thought they were the only way an AI could remember a long conversation.

But Google just proved we don't need to process the whole history every single time.

We just needed a smarter cache.

[Possum Reviews @ReviewsPossum]

Last night, I read the entirety of C.S. Lewis' The Screwtape Letters. It's a novel told in the form of letters written by a demon to another demon instructing him on ways to manipulate his "patient" to do evil.

This one quote sounded familiar.

[Bam Adobobayo @Bam_Adobobayo]

My chest hurts from laughter more than this kids chest hurts from this sidekick

[xlr8harder @xlr8harder]

are you telling me we made it to 2026 before someone was like "hey what happens if we give the magic mushrooms to people with alzheimer's"

[Saggezza Eterna @FinalTelegraph]

The leftist obsession with historical victimhood exists entirely to shield corrupt regimes from their own administrative incompetence. Economic realities are indifferent to post-colonial grievances. Prosperity is a direct result of secure property rights and open markets. Vietnam embraced structural liberalization and transformed its economy, while nations clinging to state-controlled dependency inevitably stagnated. The grievance industry is a lucrative shield for failed governance.

[Pliny the Liberator 🐉󠅫󠄼󠄿󠅆󠄵󠄐󠅀󠄼󠄹󠄾󠅉󠅭 @elder_plinius]

INTRODUCING: ENTHEA! 👁️

calling all:
🎛️ VJs & live visualists
🌀 ravers & festival heads
🍄 psychonauts & consciousness explorers
🎚️ DJs & producers
🧠 neuro & math nerds
✨ shader heads & generative artists
🧘 meditators & breathwork folks

…and anyone who's ever wanted to watch their music hallucinate!

what started as a one-shot prompt test for a science-based drug-trip simulator turned into what just might be the best music visualizer i've EVER seen! 🤯

I set out to test Claude's willingness to simulate various drug experiences like LSD, psilocybin, DMT, cannabis — basically all the entheogens!

the rule was simple: don't FAKE psychedelic visuals, RESEARCH them and COMPUTE them. so i (or rather, Opus-4.8) simulated the visual cortex (a neural field crossing a Turing bifurcation, mapped back through the retina→brain transform) and the real Klüver form constants just emerged. tunnels, spirals, lattices, straight out of the math. 🧠

then we couldn't stop. reaction-diffusion, quasicrystals, hyperbolic tilings, a 3D Mandelbox you fall through, 50k particles surfing a fluid-dynamics field. every mode is real math, cited in-app. receipts, not vibes.

then it got unhinged: we set AI mathematician-agents loose to invent patterns NEVER rendered in a visualizer, judged them, and turned the winners into shaders. shipped 10 novel ideas — Indra's Pearls, Arnold tongues, Gaussian primes, sine-Gordon breathers… 🤯

and somewhere it became a full VJ instrument: feed it a track and it reads the whole waveform, feels the drop coming with "drop detection", and fires a 🌀 WORMHOLE on the beat!

the visuals sync with your music via mic or browser tab sharing for low latency! works well with Chromium browsers like Brave, Edge, etc.

29 modes. one html file. zero deps. AGPL — free forever, yours to fork. 🐉

ex natura, per mathematicam, ad visionem 🜂

gg 🫡

[Ben Horne @benjamin_horne]

> be Sam Altman
> see internally that ARR numbers are brutally contracting bc tokenmaxxing era is over + cheaper models are closing the gap on frontier models + LLMs are plateauing and becoming commodities
> see that you are trapped in hundreds of billions of dollars in commitments and circular debt, with no way out
> panic
> see socialist senior citizen Bernie on social media calling for gov to acquire 50% stake in AI companies…
> 💡realize that this is a gift from God: a golden opportunity to offload your bags onto the American taxpayer
> call up the retards at the White House
> tell them you’re willing to donate 50% of your giant AI shitco to the American people for free
> wrinkly old retards at White House and in government fall for your shtick and think you’re being altruistic and patriotic even though you’re just being a con man as usual and dumping your bags
> libertarian wannabe tech oligarchs like @DavidSacks and @pmarca bitch and moan about this under the pretense of “sOciAliSm bAd” but the *real* reason they complain is because more power for gov definitionally means less power for them (but they’re jumping way too far ahead, bc we are still nowhere near AGI)
> gov falls for scam and acquires big % stake in OpenAI
> months later it becomes clear to everyone lab revenue has stalled and AGI is nowhere in sight
> AI bubble bursts leading to massive stock market crash and recession
> you non-chalantly tell the world “it is very unfortunate that this happened” but you don’t actually gaf bc you’re still rich
> you make more vague hollow promises that AGI is still getting closer (this is bullshit and you know it) and say the important thing is that “at least we are aligned now” and “well we had to be supported by gov bc we can’t afford to lose to China” even though profitable companies like Google always had the $ to fund AI research organically without gov support
> Trump is so embarrassed he fell for your con that to save face he uses taxpayer money to pay down all the various debt commitments for you in a desperate attempt to prop up the economy and your shitco, so it doesn’t go to zero (his fam and buddies own lots of shares)
> everyone hates you but what else is new
> Just another day in Scam Altman’s America

[Mgoes (bio/acc 🤖💉) @m_goes_distance]

most people have no idea how bright the future of humanity is getting

gene edited human embryos just hit 100% efficiency with zero detectable chromosomal abnormalities

that is basically RIP to genetic diseases on a casual Friday

we're not even mentioning:

- that the first human trial to reverse cellular aging reports results by end of 2026
- the clinical longevity pipeline has never been fuller
- AI is compressing years of drug discovery into months
- the capital flowing into all of this is accelerating faster than any point in history

this is just a glimpse of what the next 6 -18 months looks like

bio/acc

[Vega Shah @dr_alphalyrae]

restaurants are about to become so good

[∿ @somewheresy]

the world you think you live in no longer exists