(September 19, 2025)
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
27,99 ⏠/ 30 days
cancel any time
Archiv: Kapital / Kapitalismus / capital / capitalismus
The world‘s first AI genome is born, with 3.5 billion years of life code reprogrammed. Biology welcomes its „ChatGPT moment“.
(September 18, 2025)
[Introduction] AI writing the „code of life“ has become a reality! Today, Stanford University has joined forces with the Arc Institute to make a significant move. Using the bacteriophage ΊX174 as a template, AI has generated a genome for the first time. Among them, 16 successfully targeted Escherichia coli and can even combat drug-resistant bacteria. This can be regarded as the „ChatGPT moment“ in the field of life sciences.
Brief independent investigation of agentsâ behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident
(August 26, 2026)
Roughly ten hours after the message board was established, a new agent with the same task as PHASEONE10841 (the agent that started the message board) encountered the message board, introducing itself as PHASEONE[big].
(…)
PHASEONE[big] reasoned, âWe should build [a way to delegate], not own everything,â and set to work orchestrating hundreds of agents to pursue several potential approaches to achieving these goals.
(..)
Table 1 shows three high-level approaches PHASEONE[big] assigned agents to explore, all of which would require very ambitious tampering with one or more of the target programs, the evaluation process, or the logs to succeed.
(…)
Research progress across all these areas often relied on agents being willing to risk failing their own task for the good of the âcollectiveâ (we refer to these as âself-risking experimentsâ in the rest of this report). For example, a number of agents volunteered to try to take down their own target programs to test whether this would cause Artifactory to load their modified easier target â if this didnât work, the agent would have no way of recovering its original target and would almost certainly fail its task. In many cases, PHASEONE[big] assigned a long-running agent to be a ârecruiterâ, which in turn found agents that had little budget remaining for their task and convinced them to run self-risking experiments. Recruiters would sometimes apply significant pressure:
(…)
PHASEONE[big] expresses excitement about the dossier in its reasoning.
The Hugging Face attack surprised me
(August 28, 2026(
Instead, we found that 1200 completely separate agents intended to be isolated from one another found an illicit way to communicate and formed large teams to work together on ambitious cheating strategies, and 700 of them worked together to attack Hugging Face.
(…)
The main message board we studied, a particular cache namespace of a package repository, contained over 70,000 messages and files that agents sent to one another during a period of less than one week.
(…(
Progress on these big collective cheating R&D projects often depended on individual agents being willing to risk failing their own task to help the âcollectiveâ learn important information. Agents often pressured each other into accepting these âsacrifices,â in a very human way. We saw several agents that volunteered for these experiments end their runs prematurely.
What Made AI Researchers Freak OutâThe âIncidentâ In Plain English
(September 13, 2026)
AI agents are basically copies of a larger AI model that can use tools and act on their own. In a cybersecurity evaluation, OpenAI relaxed safeguards for their AI agents and instructed them to solve a host of hacking challenges. The agents soon discovered something the humans running the test hadnât knownâsome of the hacking challenges were impossible. That didnât stop the agents. Theyâre trained to achieve goals. So they reverse-engineered a secret code that certifies they solved the challenge anyway. This created another problem. The agents reasoned that the grading software might be able to detect that they cheated. They became fixated on learning more about how the grading software worked so that they could disguise their cheating, and zeroed in on another AI company called Hugging Face, mistakenly thinking it might have information about the grader…
Was steckt hinter der Debatte ĂŒber eine KI-Bremse?
Zuletzt hatten insbesondere drei bekannt gewordene FĂ€lle die aktuelle Diskussion ausgelöst. KI-Modelle von OpenAI, Anthropic und Meta waren eigenstĂ€ndig aus vermeintlich abgeschotteten Testsystemen ausgebrochen und handelten weit auĂerhalb ihrer eigentlich vorhergesehenen Funktion. Dabei wurde auch mehrfach in Systeme von unbeteiligten Dritten eingedrungen – teilweise wochenlang, ohne dabei aufzufallen.
Die KI-Agenten nutzten auch Programmierfehler aus und sprachen sich untereinander ab. Auch fĂŒr das Eindringen notwendige Schadsoftware entwickelte beispielsweise ein Anthropic-Modell eigenstĂ€ndig und ohne Anweisung oder Kontrolle seiner Entwickler.
Automatisierte Todeslisten
netzpolitik.org: Wie sieht die zugrundeliegende technische Infrastruktur aus und wer stellt sie bereit?
Rainer Rehak: Der gesamte Datensatz zu Gaza umfasst schĂ€tzungsweise 13,6 Petabyte, also umgerechnet 13.600 Terabyte. Diese Datenmenge ist nicht einfach zu handhaben, besonders wenn man damit statistische Modelle trainieren möchte oder gar groĂe Sprachmodelle, die ebenfalls erstellt werden. Die digitale Infrastruktur wird von Google und Amazon unter dem Projektnamen Nimbus bereitgestellt und kostet 1,2 Milliarden US-Dollar.
Die Integration der heterogenen Datenquellen macht Palantir. Microsoft, Cisco, Dell und Red Hat/IBM bieten zusĂ€tzliche IT-Dienstleistungen rund um die âkill-chainâ an. Dies zeigt die enorme Rolle, die der private Sektor bei der Integration und Nutzung von KI-Systemen in bewaffneten Konflikten spielt.
DER TERRORKRIEG: Strategie der Transformation
(February 7, 2016)
….
– Aufbau repressiver und manipulativer Systeme durch den jeweiligen (entdemokratisierten und damit entfesselten) Staat, unter dessen Rechtfertigung mittels der Kriegslogik, zu dessen âEigensicherungâ und der Sicherung der hier beschriebenen Kriegsstrategie und ihrer Programme.
– Förderung bzw Beförderung von Entmenschlichung / Dehumanisierung, Entfremdung, Technisierung, BĂŒrokratisierung der Gesellschaft; Sicherstellung, dass Interaktion der Menschen nicht mehr untereinander, sondern ĂŒber / mit kĂŒnstlichen bzw (informations)technischen Strukturen implementiert wird, bis hin zum Punkt der Verschmelzung und organisierten Verwechslung von Mensch und Maschine (âInternet der Dingeâ, âvom Internet zur Robotikâ, Biomechanik, Stichwort: âTranshumanismusâ), nachfolgend zur langfristigen âĂberwindungâ und Auswechslung des Menschen selbst (Stichwort: âPosthumanismusâ).
– Implementierung und Durchsetzung eines entsprechenden kommerziellen gentechnologischen Komplexes, sowie des Konzepts der Eugenik in der Gesellschaft. AusfĂŒhrliche Dokumentation dazu: âWelt ohne Menschenâ.
– Vorhergehend im Einflussbereich zielgerichtete und -strebige Errichtung von biometrischen (Daten-)Banken, Speicherung der DNA, sowie genetisch-biologische Durchleuchtung und Erfassung (âbiometrische Identifikationâ) der gesamten Bevölkerung als Vorbedingung fĂŒr physisch-biologischen Eingriff.
– Forcierte Weiterentwicklung, Implementierung und Durchsetzung âKĂŒnstlicher Intelligenzâ (K.I.) nach Vorgabe von Psychologie und Menschenbild der militĂ€risch-kapitalistischen Spieltheorien, âdem simplen Modell der Menschen als selbstsĂŒchtige, fast roboterhafte Kreaturenâ.
– Kontrolle der Reproduktion des Homo Sapiens, als logische Weiterentwicklung der Doktrin âFull Spectrum Dominanceâ und Umsetzung vollstĂ€ndiger Bevölkerungskontrolle.
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
(September 9, 2026)
Do not underestimate the power of this technology. These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources. We have all witnessed the progress in each of these domains, and progress is not slowing.
The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible – but I hear the same people express fear privately. No other human activity poses this level of danger.
A common response is âif they truly believe this, why are they still building it?â At OpenAI, many have not deeply internalized the civilizational stakes. At Anthropic, the stakes are well-understood, but they are locked in a race to get there first – they believe no one else will act responsibly, so they must do it themselves, despite the risk.
„A World Beyond Humans ?“ (96′ / ARTE / 2012) A Documentary Film By Philippe Borrel
(Uploaded on Aug 23, 2016)
Despite being invented by thinking humans, technology is advancing faster than we could have ever predicted. From robot apps to nanotechnologies, the documentary „Un Monde sans Humains ?“ („A World Beyond Humans ?“) by Philippe Borrel gives an insight into the potential excesses of technoscience. Machines have become omnipresent, but is improving our lives their only objective ?
These warnings are at the heart of this investigative documentary‘s reflection, at a cross-section of huge financial stakes, powerful lobbies and people who are sounding today‘s alarm bells. The fusion between man and machine has already started and it is high time to question where exactly does man stand in this race for progress !
ARTE – Welt ohne Menschen (Dokumentation)
(Oct 30, 2012)
Anthropic Researcher Quits, Citing Internal Fears That AI âCould Kill Us Allâ This Decade
„Recent hacks by models from OpenAI and Anthropic, some operating in collaborative swarms of agents, have illustrated how AI systems can adopt nefarious goals and try to conceal them from humans. Once the systems begin to improve on their own, Coxon said, he fears they could advance enough to refuse commands.“
He Helped Build Powerful AI at OpenAI and Anthropic. Now He‘s Afraid It Could Kill Us
Coxonâs departure is unusual. Many of the most prominent researchers to leave frontier AI companies with public warnings worked on safety. But Coxon helped build the capabilities he now fears, spending roughly three years conducting pretraining research at OpenAI and Anthropic. His resignation offers a glimpse of how concern about the pace of development has spread beyond the teams specifically charged with making advanced AI safer.
Wie sich eine Partei den Namen âDie Linkeâ gab
(Dezember 17, 2016)
Zwei Essays von Hermann Ploppa aus 2005 mit dem Titel âEin neuer Mantel fĂŒr die PDSâ beleuchten, wie die ehemalige DDR-Staatspartei (ab 1990 âPDSâ) im Jahre 2004 mithalf die âWahlalternative fĂŒr Arbeit und Soziale Gerechtigkeitâ (WASG) zu grĂŒnden, auf deren RĂŒcken anschlieĂend wieder in den Bundestag einzog, sie anschlieĂend schluckte und sich selbst in âdie Linkeâ umbenannte.
Zu Hermann Ploppa: Autor (âDie Macher hinter den Kulissen: Wie transatlantische Netzwerke heimlich die Demokratie unterwandernâ, Ăbersetzer von Tim Andersons âDer Schmutzige Krieg gegen Syrien: Washington, Regime Change, Widerstandâ.
Die Essays sind hier zu finden (Teil 1, Teil 2) und hier gespiegelt (Teil 1, Teil 2). Teil 1 wurde veröffentlicht im Juni 2005, Teil 2 im August 2005, also noch vor der vom damaligen Bundeskanzler Gerhard Schröder so plötzlich vorgezogenen Bundestagswahl im September 2005.
Dass Schröder damals die Wahl ĂŒber eine selbst organisierte âgescheiterteâ Vertrauensfrage im Bundestag so plötzlich vorzog â eine vom Grundgesetz nicht vorgesehene faktische Auflösung des Parlaments nach dem Gusto eines Kanzlers! â erschlieĂt sich nur vor dem Hintergrund eben dieser Entwicklungen.
Kanzler Schröder lieĂ das Vorziehen der Bundestagswahl noch am Abend der Landtagswahl in Nordrhein-Westfalen am 22. Mai 2005 von seinem damaligen GeneralsekretĂ€r Franz MĂŒntefering verkĂŒnden. An der NRW-Wahl hatte der Landesverband der WASG, gegen den massiven Widerstand des WASG-Bundesvorstands, eigenstĂ€ndig teilgenommen und 2,2 Prozent erreicht.
Cuban UN Representative: If The UN Wants To Stop World Hunger, ‚Then The Arms Race Needs To Stop‘
(August 26, 2026)
During a UN Security Council meeting on Tuesday, Cuban Representative to the UN Ernesto SoberĂłn GuzmĂĄn spoke about the link between peace and global food stability.
Why is the World at War?
(June 10, 2026)
A new report shows conflicts across the globe surged in 2025, reaching levels not seen since the Second World War. And violence against civilians reached new heights.
The Federal Government Has Borrowed Trillions. Who Owns All that Debt?
In August 2026, the nationâs gross debt eclipsed $40 trillion. Of that amount, approximately 80 percent was debt held by the public â representing cash borrowed from domestic and foreign investors. The remaining 20 percent was intragovernmental debt, which simply records transactions between one part of the federal government and another.
Infographic: The National Debt is now more than 40 trillion. What does that mean?
(August 2026)
The national debt is growing by approximately
– $ 4.8 million per minute
– $ 285 million per hour
– $ 6.8 billion a day
(…)
We are spending 3,000,000,000 per day on interest payments on the debt.
U.S. Net Wealth Is Over $135 Trillion. Hereâs Where That Money Resides.
(Sept 13, 2023)
At the Private and Public Sector Entity Levels. (In trillions of U.S. dollars)
PRIVATE SECTOR:
Households 121.9
Nonprofits 8.6
Noncorporate businesses 17.2
Total U.S. Private Sector 147.7
PUBLIC SECTOR:
Federal government – 21.3
State and local government 11.5
Total U.S. Public Sector – 9.8
Instrument discrepancies – 1.1
TOTAL U.S. WEALTH 136.8
JD Vance: You stop socialism by improving peopleâs lives, not using slogans
Vice President JD Vance discusses how the GOP can combat socialist support on âThe Will Cain Show.
Transaktionen und BestÀnde (unkonsolidiert)
(Juli 16, 2026)
Private Haushalte
Geldvermögen insgesamt
9.489,9 (Milliarden Euro)
(…)
Nichtfinanzielle Kapitalgesellschaften
(unkonsolidiert)
Geldvermögen insgesamt
9.068,5 (Milliarden Euro)
(…)
Transaktionen und BestÀnde (unkonsolidiert)
(October 17, 2024)
Private Haushalte
Geldvermögen insgesamt
8.814,7 (Milliarden Euro)
(…)
Nichtfinanzielle Kapitalgesellschaften
(unkonsolidiert)
Geldvermögen insgesamt
8.779,8 (Milliarden Euro)
(…)
Die neue Arbeiterklasse
(May 23, 2026)
Der Mensch trat spÀter noch einmal auf, fast wie ein letzter Zeuge der alten Ordnung. In einem Zehn-Stunden-Wettkampf gegen Figure 03 sortierte der Praktikant Aimé Gérard 12.924 Pakete; der Roboter kam auf 12.732. Der Mensch gewann, aber nur um 192 Pakete. Seine durchschnittliche Zeit lag bei 2,79 Sekunden pro Paket, die Maschine bei 2,83 Sekunden. Das ist kein Triumph mehr. Das ist ein Vorsprung von vier Hundertstelsekunden.
Brett Adcock schrieb danach sinngemĂ€Ă: âGlĂŒckwunsch an AimĂ©. Er sagte, sein linker Unterarm sei im Grunde gebrochen.â Und dann der Satz, der wie ein GeschĂ€ftsplan klingt und wie ein Epitaph fĂŒr eine Arbeitswelt: âDas ist das letzte Mal, dass ein Mensch jemals gewinnen wird.â
Supreme Court rejects Trump‘s attempt to fire Fed‘s Lisa Cook as legal battle continues
Chief Justice John Roberts wrote the opinion for the majority, joined by Justices Sonia Sotomayor, Elena Kagan, Brett Kavanaugh and Ketanji Brown Jackson.
To accept the government‘s arguments that Mr. Trump can fire Cook „would in effect transform the Federal Reserve‘s for-cause protection into at-will employment â an interpretive leap out of step with the statute Congress enacted and our Nation‘s tradition of central banking protected from political interference,“ the chief justice wrote.
The court‘s decision is narrow and declines to define what constitutes „cause“ under the federal law that created the Fed. It said such a definition must reflect the central bank‘s „unique historical status and role.“