Dran Fren

Media, books, overthinking. Everyday I see a hawk and a raven meandering around. Reductive narcissism. Simulation of knowledge only. Who talks doesn't know.

[Dran Fren @GlobalComplexit]

“Anyone who cannot cope with life while he is alive needs one hand to ward off a little his despair over his fate...but with his other hand he can jot down what he sees among the ruins, for he sees different and more things than the others;

[Dran Fren @GlobalComplexit]

@ElliottThornley interesting, what is the deadline

[Dran Fren @GlobalComplexit]

@Rebecca21951651 hard to say how the AI bet will play out🤔🫨

[Bastian Barucker @BBarucker]

Untätigkeitsklage gegen das Paul-Ehrlich-Institut - Heute Morgen habe ich das Paul-Ehrlich-Institut verklagt, da die Behörde ihrer Verpflichtung nicht nachkommt, fristgerecht auf meine Forderung zur Herausgabe wichtiger interner E-Mails zu antworten.

Die Behörde möchte bis heute nicht öffentlich preisgeben, wie viele Verdachtsmeldungen für Impfnebenwirkungen durch die eigens dafür entwickelte SafeVac2.0-App eingegangen sind.

Sie wertet die Daten der Kassenärztlichen Vereinigungen nicht aus, obwohl diese seit Jahren vorliegen und sie gesetzlich dazu verpflichtet sind. Während der gesamten Corona-Impfkampagne soll die Behörde außerdem im Fachbereich Pharmakovigilanz und in der darüber liegenden Abteilung angeblich kein einziges Protokoll angelegt haben. Gleichzeitig hat die Behörde den Geimpften erklärt, dass sie über solide Überwachungssysteme verfügt und zeitnah auf Alarmsignale reagieren wird.

Nun braucht sie bereits mehr als sechs Monate, um meine Veröffentlichungsforderung wichtiger interner Mails zu beantworten.

Während die Entwicklung, Zulassung und massenhafte Herstellung eines neuartigen modRNA-Präparats neun Monate dauerte, braucht das PEI bis heute mehr als fünf Jahre, um wichtige, dafür vorgesehene Datensätze zu veröffentlichen. Warum?

Es ist davon auszugehen, dass auch dieses Verfahren lange dauert und das PEI die Veröffentlichung der E-Mails zwischen dem ehemaligen PEI-Präsidenten Cichutek und der ehemaligen Leiterin der Pharmakovigilanz, Keller-Stanislawski, um jeden Preis verhindern möchte.

Meine Arbeit ermöglichen: https://t.co/whlm5co4At

[Dran Fren @GlobalComplexit]

This is the first chapter of my upcoming novel
"The Green Warrior or: About the impossibility of making a pass at someone"
1
The fullness of life

- To make a conquest or not to make a conquest. Why did this form of language of toxic masculinity have to be perpetuated? In fact, every breath, every look, every touch, every image on the television was toxic masculinity. Existence is toxic in itself.
This may seem like an exaggeration to defenders of toxic masculinity.
Her thoughts were once again running into emptiness.
Why should we even concern ourselves with such topics? Why should you stick to it?
The shaman had said that the most important thing is to maintain a certain level of awareness. Beyond that, it didn’t matter much what label you put on yourself, as it was all about being able to feel morally superior.
She wasn’t a real writer, but she was interested in making amends for wrongs that had happened; what could be wrong with that?
The mirror game of the world that is ending should keep alive the injustice of past times in historiography in order to create a kind of justice and not leave everything to the victors. But after winners came new layers of winners.
A game of the mind with different roles that created order and chaos. Was life worth it?
There she stood at the railway track and considered whether she should speak to the young man. It was not without reason that he had positioned himself directly in her line of sight, and was actually presenting himself in front of her. But she was never sure if it was real or if she was just imagining it. How would she know if she could trust her instincts? That reflex again, almost automatic. She could vividly imagine him sucking her cock. She was startled at once by herself, by the unbidden thought.
When she wrote these lines, she was sitting in a café and pretending to be a writer.
Ideally, she could simply acknowledge the fullness of the present moment. Just enjoy the fact of your own existence. Even to speak of “the fact” is almost the wrong nomenclature. It’s about the feeling of your existence, your aliveness in the present moment. Beyond concepts - i.e. without letting your existence be massively filtered through concepts.
The secret to manifestation was simply believing that she had already received what she desired. One child, two children. The B situation.
Nevertheless, loss was always part of the program.
She was one of the first women to be able to have children after undergoing gender reassignment surgery. And this at a time when most women were no longer able to have healthy children after failed treatment with a gene therapy drug.
He was able to mentally change his gender as he saw fit.
How would she feel if she had achieved what she wanted.
Why couldn’t she trust her instincts? Her instincts told her that the young man on the platform had deliberately placed himself in her line of sight. He was physically very fit and sexually attractive to her. She didn’t like his face that much, but she didn’t have to pick him up straight away, just getting his number would be enough. As he faced her, she could see the tension on his face, his excitement and anticipation. At least that’s what she imagined, or that’s what her instinct said, but she didn’t dare trust it. She could just imagine him ******* her ****.

Follow my substack (if you like): https://t.co/hdvWoYVRTk

[Dran Fren @GlobalComplexit]

from my substack (with heavy AI help):
Governments, and other institutions that act on behalf of a public, sometimes have to decide before they can know whether the decision was right. That much is unremarkable — all action under time pressure has this shape. What is worth a paper is what such institutions typically do next: they tend to present the decision, both during the emergency and afterward, as though it had been read directly off the evidence, rather than chosen under conditions the evidence did not fully settle. This paper calls that operation the veil of rationality, and the underlying problem it responds to the veil crux.

The crux arises where five older problems of political philosophy cross:

- Prudence: action must be taken in a singular situation that no rule or body of knowledge fully determines.
- Necessity and prerogative: political order sometimes survives only through acts that exceed, suspend, or strain its ordinary law.
- Public authorization: modern power must present the act as the act, interest, or will of the people — not merely the ruler’s preference.
- Legitimating appearance: the contingent judgment must acquire a public form that makes obedience and coordinated action possible.
- Government through knowledge: modern states govern populations through statistics, expertise, risk, prediction, and administrative rationalities, so political judgment can appear as the impersonal conclusion of knowledge itself

https://t.co/KJ4dKcVi5w

[Dran Fren @GlobalComplexit]

yes thats what observed in practical working with LLMs, even trying to install meta-cognitive subagents to mitigate these problems

[How To Prompt @HowToPrompt__]

Google has found something worst than AI hallucinations.

they call it “Metacognitive Failure.”

when an LLM hallucinates a fact, it's a data error. bad memory retrieval, wrong weights, fine. you can fact-check it.

metacognitive failure is a structural psychological defect in the model's architecture.

according to the research, current frontier models exhibit massive gaps in their internal self-monitoring:

the supreme confidence trap: they routinely hallucinate with maximum, bulletproof confidence. they sound just as sure of themselves when they are completely wrong as they do when they are 100% right.

the boundary blindness: they have zero internal mechanism to recognize their own knowledge boundaries. they blindly step off cliffs because they can't sense the edge.

the calibration mismatch: there is a complete disconnect between what the model actually "knows" internally and what it expresses in its output.

the fix: reinforcement learning with metacognitive feedback (rlmf)

the researchers operationalized a new technique to force models to face reality. instead of just rewarding the model for getting the right answer, rlmf grades the model on how accurately it evaluates its own performance and uncertainty.

by aligning a model's expressed confidence with its actual intrinsic uncertainty, they managed to massively improve model calibration without dropping raw accuracy.

as we shift into an agentic era where ai systems are running code, managing infrastructure, and making automated decisions without humans in the loop, a model that doesn't know what it doesn't know is dangerous.

[Dran Fren @GlobalComplexit]

@HowToPrompt__ yes thats what observed in practical working with LLMs, even trying to install meta-cognitive subagents to mitigate these problems

[Dran Fren @GlobalComplexit]

@ChombaBupe arguably they don't need to be intelligent to kill everyone off

[Dran Fren @GlobalComplexit]

@ChombaBupe arguably they don't need to be intelligent to kill everyone off

[Dran Fren @GlobalComplexit]

🤔😵‍💫

[Rebecc@ ███████ @Rebecca21951651]

Why hasn't @NatureMedicine retracted Proximal Origins yet?

@RetractionWatch @bborrell

[Dran Fren @GlobalComplexit]

@LudlowWorthing at this point weighing things on their own merits and using critical thinking seems to be rather discouraged and massively disincentivized in many key fields/regarding key topics. It can easily become risky for the career, reputation or even person to do so.

[Dran Fren @GlobalComplexit]

There is a major problem here: the entire issue has become entangled with the Trump administration, which, at this point, has—IMHO—lost virtually all credibility. As a result, many people seem to assume that anything this administration proposes or implements can be dismissed without much further scrutiny.

Unfortunately, that dynamic ultimately helps Trump. It makes a serious, broad-based reckoning with what I consider the catastrophic mistakes made during the pandemic even less likely. And the longer that reckoning is postponed, the more the unresolved conflicts and loss of trust from that period continue to deepen social divisions and inhibit open public discussion.

[Alina Chan @Ayjchan]

David Morens, who was senior advisor to Anthony Fauci, has pled guilty to conspiracy.

He signed this document, agreeing that if the case had proceeded to trial, it would've proven the following beyond reasonable doubt:

Morens and his co-conspirators 1 and 2 (including Peter Daszak of EcoHealth Alliance) and others

💠conspired to defraud and commit offenses against the United States by hiding from public view various federal records related to the pandemic

💠emailed information to Morens at his personal email to relay to Senior NIAID Official 1 (Anthony Fauci) while instructing Morens to use means of communication that would not be discovered in a FOIA search

💠plotted and repeatedly expressed their intention to conceal federal records and provided info to Senior NIAID Official 1 (Fauci) favorable to a co-conspirator's company (EcoHealth)

💠used Morens' personal email to exchange federal records concerning the drafting and sources relied upon for public writings on the origins of the Covid-19 virus

💠concealed their communications so they could generate the appearance of disinterested scientific consensus in support of a co-conspirator's company (EcoHealth)

💠shared information about Congressional inquiries and FOIA requests to anticipate press reactions, develop talking points and coordinate responses

Morens faces up to five years in prison and a fine of up to $250,000.
https://t.co/e1Ml0A1uho

[Dran Fren @GlobalComplexit]

@RyanGreenblatt yes, to do such analysis LLMs need to do interpretive work, which they are often not trained well. They lack what I call Augenmaß. Are these agent outputs publicly available or will they be released for study?

[Dran Fren @GlobalComplexit]

@RyanGreenblatt yes, to do such analysis LLMs need to do interpretive work, which they are often not trained well. They lack what I call Augenmaß. Are these agent outputs publicly available or will they be released for study?

[Dran Fren @GlobalComplexit]

@skdh criticizing science (iow doing science) these days IS already political, because we have been told to TRUST THE SCIENCE (blindly)

[Dran Fren @GlobalComplexit]

@skdh criticizing science (iow doing science) these days IS already political, because we have been told to TRUST THE SCIENCE (blindly)

[Dran Fren @GlobalComplexit]

well, it's hard to assess how important a factor Trump was for how the whole thing went down. I think the whole issue is extremely complex to untangle because of the sheer amount of moving parts on so many levels.
In fact I think the complexity in a crises might even be the main point, that the narratives and strategies were simplified this way... and once down this road, with so much at stake it's difficult to come to an honest review...

[Dran Fren @GlobalComplexit]

The Festival of the Green King https://t.co/Wiu9busH2o

[Dran Fren @GlobalComplexit]

@zeteo_news it's not just Republicans, unfortunately the issue has been politicized, maybe you should look again with due diligence

[Dran Fren @GlobalComplexit]

@nicknorwitz conclusions have become the main mechanism