notandus, ergo sum

As a young man in New York, Donald Trump was determined to stamp his name onto physical structures. He was Fred Trump’s son, he was on the make, and he wanted to leave his own indelible mark on Manhattan—the wealthiest borough in the city and one his father had never conquered. Putting his name on a gleaming Fifth Avenue tower was an exercise in branding into the American consciousness. For Donald Trump, to be famous was to exist. 

Forty years later and at the pinnacle of American power, the by-now-seventy-nine-year-old Trump was obsessed with dreams of legacy and determined to use the scope of an imperial presidency to imprint his name on everything, including the low-rise cityscape of Washington, D.C. 

By the norms of history, the naming rights for monuments belonged to subsequent generations, and those honors were conferred by Congress, by historians and commissions, and by public acclaim. It was the American way. But Donald Trump had long since demonstrated that he could adore himself far more, and in greater detail, than anyone else might. Why wait? He had watched in rage as his name was removed from buildings in New York City in protest during his first presidency and after its calamitous end. It was a mistake to assume others would praise you for your efforts, he told The New York Times in 1980. “If you let people treat you how they want, you’ll be made a fool,” he said, before adding, “I don’t want to be made anybody’s sucker.” 

John F. Kennedy had never named anything for himself. Nor had any other President in American history openly campaigned for a public structure to bear his name—though a few, like Washington and Hoover, had buildings or landmarks named for them by others while still in office. President Gerald Ford went so far as to refuse to sign a bill that would have renamed a courthouse and federal office building after him. Ford said he knew of no federal buildings that had been named for a President while still in office, and if he were to accept the honor it “might begin a precedent I believe it best not to establish.” 

One of Kennedy’s legacies, together with the First Lady Jackie Kennedy, had been turning the White House into a showcase for American arts and high culture, inviting poets, musicians, actors, and dancers to perform at social events, special concerts, and state dinners. When Kennedy was slain it seemed natural that a nation searching for a way to celebrate his life would name a cultural center in his honor. On January 23, 1964, the National Cultural Center—created by an act of Congress at the end of the Eisenhower administration—was officially renamed the John F. Kennedy Center for the Performing Arts, to serve as the “sole national memorial” to the thirty-fifth President. 

Since then, the Kennedy Center had served as the cultural center of the capital city, the longtime home of the National Symphony Orchestra and the Washington National Opera, and the host of the world-renowned Kennedy Center Honors. 

In this most political city in the country, the Kennedy Center was for decades managed by a bipartisan board, maintaining a reputation for political neutrality, focused on artistic excellence rather than partisan favor. But Trump had refused to attend the Kennedy Center award ceremony during his first term because so many of the artists had made it clear they were anti-Trump. Now, in his second term, he would not only attend the ceremony, but he would have a hand in picking the honorees. And he would appoint himself the evening’s host. 

When on December 18, 2025, the board of directors of the Kennedy Center had voted unanimously to change the name to “the Donald J. Trump and the John F. Kennedy Memorial Center for the Performing Arts,” Trump feigned surprise and humility that the board—which he had personally selected just a few weeks after his second inauguration and of which he was now the chairman—would bestow such an honor. 

The renaming was enthusiastically embraced by Sergio Gor, then the head of the presidential personnel office, but later dispatched as ambassador to India. When Gor raised it with Trump months earlier, the President loved it. You’re doing all this work for the Kennedy Center, Gor had told him flatteringly. You got the money. You should at least get your name on it with JFK. 

The White House quietly ordered the large letters “DONALD J. TRUMP” to be added to the facade well before the board had even rubber-stamped the authorization to rename the building…

Haberman, Maggie; Swan, Jonathan. Regime Change: Inside the Imperial Presidency of Donald Trump (pp. 365-367). (Function). Kindle Edition.

Quite the read...

Dissing Maggie

"Maggot?" How juvenile.

 

The Maggie Haberman-Jonathan Swan book was just released.

 

 

Yeah, of course I bought it. Stay tuned...

 

UPDATE

 

I am deep into this book. Very disturbing, even for someone already steeped in Trump critical biographies.

 

A snip from midway through the book:

By midway through the year, a picture was emerging of how the Oval Office operated in Trump’s second term. It was of a President spending his days at the Resolute Desk in a series of rolling bull sessions, accompanied by a core group of intimates; these were supplemented by a rotating cast of extras, and on any given day they could be Republican lawmakers, titans of industry, former pro wrestlers, country musicians, Gulf royals, crypto bros, or friends of felons seeking pardons. They would enter and exit the frame, with some invited to stay for meetings they had no business attending. The conversations ran fast, often straying far from the point, or from anything visitors imagined when they arrived at the White House. The President’s sentences often began on one topic and ended far away. 

The ever-present Natalie Harp was generally on a chair off to the side, her laptop open, head cocked, listening but never contributing unless ordered to by Trump. She fulfilled, in a flash, any request, whether a “Trump 2028” hat needed to be fetched from the merchandise room, a quick Google search, or producing the latest story from right-wing websites like Breitbart or Gateway Pundit. Once Dan Scavino became the head of the Presidential Personnel Office in late 2025, he mostly handed over Trump’s Truth Social account to Harp. 

Most presidents had adhered to the mantra that their most valuable commodity was time. George W. Bush and Barack Obama scheduled their days down to blocks of ten or fifteen minutes. The ideal “gatekeeper” function of a White House chief of staff was to guard those blocks of time against needless interruptions. But Susie Wiles had little interest in micromanaging her boss’s time, perhaps correctly calculating it was futile. She could try to control access to who got into the Oval Office, and what their agenda might be, but otherwise she would let the President set his own pace. 

In Donald Trump’s White House, time was a flexible concept. To outsiders, the administration appeared a hive of activity, but those brought into the Oval Office would often remark that Trump seemed to have all the time in the world. Unless the meeting concerned an imminent military operation, Trump was usually relaxed, gossipy, unhurried. Many visitors to the Oval were charmed. Even those who didn’t like him would often describe the President as solicitous and flattering, offering Diet Cokes and candy as if he were a greeter at one of his clubs. He would tell bawdy jokes and compliment his guests on how good they looked. He signed MAGA hats and extra ones for their kids and would urge them to come to Mar-a-Lago that weekend or to one of his golf clubs. He would spend hours examining the minute details of Oval Office decorations and plans for the ballroom he was plotting on the grounds of the East Wing. He loved showing off the quality of the speakers that piped his music—Pavarotti and Sinéad O’Connor and James Brown—into the Rose Garden patio. And he would regale visitors—especially foreign leaders—with the details of his 2024 election victory, subject them to his familiar litany of complaints about “Sleepy Joe,” and pepper them with questions about the dangerous animals in their countries with the fascination of a child. 

“I still can’t fully describe what the Oval Office is like with him,” said one visitor. “You’d have two people on two different speakerphones. Another person on a cell phone. I’ll never forget talking about a highly classified program, and this guy—looked like just a salt-of-the-earth, country guy—walks in and he’s got samples for the Rose Garden paving. Pops them on the desk. And the President stops what we’re talking about, gets up there, and walks out there with him and I see a lot of hands moving. And he comes back and then looks over our shoulders, and the curator is there with somebody to drill the cherubs into the wall that he had brought up from Mar-a-Lago. And then he goes back to the people on the phone, gives them guidance. And then he tries to come back to us, but we’re out of time.” 

This pattern would repeat all day long. The President’s last three meetings of the day would often blur into one contiguous meeting. Trade, national security, a visiting CEO. All would be mashed together into one borderless conversation. Trump had run his days this way—unstructured and improvisational—since the years he spent working from his corner office in Trump Tower. But in his second term, Trump had made the White House more fully his home—not just fixing it in accordance with his aesthetic, but fundamentally changing the rhythms and structures and operations of the place…

Haberman, Maggie; Swan, Jonathan. Regime Change: Inside the Imperial Presidency of Donald Trump (pp. 169-171). (Function). Kindle Edition.

Lordy Mercy... 

The Washington DC Reflecting Pool

 

Donald Loves awards. So I gave him one.

 

Italian Prime Minister Giorgia Meloni smacks down our crude, delusional Bully-in-Chief

 

Mic drop.

 

UPDATE

Italy's Meloni To Trump: 'My Popularity Is None Of Your Concern. I Suggest You Focus On Yours.'

Italian Prime Minister Giorgia Meloni hit back at President Donald Trump's continued insults against her on Saturday, telling Trump to focus on his own dwindling popularity.

"These constant, unprovoked attacks are senseless," Meloni posted on Instagram in response to Trump's comments earlier in the day, where he once again claimed Meloni "asked, over and over, for a picture with me."

Meloni previously called Trump's claim "completely fabricated." On Saturday, she said attempting to be Trump's friend "has certainly not helped" her own popularity.

"As for my popularity, being your friend has certainly not helped it, nor does it depend on my relationship with you," Meloni posted.

"My popularity is none of your concern. I suggest you focus on yours," she added.

"Continuous discursive tinnitus"

An excellent New Yorker long-read. Totally timely.

The person who should have been best able to explain how we got here was the great German philosopher Jürgen Habermas, who illuminated how a feisty, principled public sphere is integral to democracy. But Habermas died in March, at the age of ninety-six, and, although he remained active until his final months, commenting on Ukraine, Gaza, and Eurobonds, he struggled to understand the turn history had taken. As a teen-ager in 1945, he had witnessed American soldiers enter his home town of Gummersbach, near Cologne, carrying messages of freedom and openness. Eight decades later, he watched American voters choose a leader who had advertised his fascistic bent in blood-and-soil rhetoric, fantasies of punitive violence, and a taste for bombastic architectural kitsch. The far right was making inroads across Europe, including in Germany. The print-based media culture that once anchored Habermas’s public sphere had devolved into a digital sludgefest that proved better at circulating racist memes than at fostering morality and dignity. A couple of years before his death, in a conversation with the historian Philipp Felsch, Habermas said that his world was being dismantled “step by step.”...
____ 

...The medieval philosopher al-Farabi, who considered democracy the least imperfect of imperfect governments, is mentioned only in passing. A similar Western bias contributed to one of Habermas’s last, and least effective, public interventions. In November, 2023, after Hamas’s massacre of Israelis and the onset of Israel’s brutal counterstrike on Gaza, Habermas signed a statement that reasserted solidarity between Germany and Israel. After a glancing mention of Palestinian suffering, the authors write, “The standards of judgment slip completely when genocidal intentions are attributed to Israel’s actions.” It’s one thing to deny that genocide has taken place in Gaza; it’s another to imply more broadly that the topic is out of bounds. At a crucial moment, Habermas’s cherished pluralism failed him.

“If no dread remains, the monsters return,” Habermas wrote early in his career. They’re back, on several continents. Earlier this year, in Germany, the Sachsen-Anhalt branch of the far-right Alternative für Deutschland party released a platform containing such demands as “Think German!,” “Promote patriotism—no state money for anti-German art and culture!,” and “Build more beautifully!” This dumbed-down Goebbels gobbledygook revived talking points that Habermas had tried to quash during the Historikerstreit. Not surprisingly, AfD representatives could barely contain their glee over the philosopher’s death. Hans-Thomas Tillschneider, one of the Party’s nastier voices, posted a YouTube video in which he said, “Habermas is dangerous. He is one of the greatest enemies of the German nation.” Tillschneider’s inability to put Habermas into the past tense was somehow reassuring.

An equally obnoxious obituary came from the billionaire pen of Alex Karp, the C.E.O. of Palantir Technologies. Before Karp turned to hawking surveillance systems that have assisted ice in its murderous roundups of immigrants, he studied philosophy under Habermas in Frankfurt. In an article for Politico, Karp recounted how Habermas provided fierce but fair criticism of his papers: “It was his very willingness to be so productively unsparing that reminds me of what we have lost as a culture.” Alas, the losses that Karp has in mind don’t seem to involve learning, rigor, or reason. Waving away Habermas’s cosmopolitan ideals, he says that discourse “must be rooted in a more corporeal and traditional—and indeed national and cultural—source.” This is the language of maga and the AfD, not to mention Heidegger circa 1935. Karp’s ideological atavism is all too typical of the current bent of Silicon Valley...

 

When the A.I. chatbots march in, the “colonization of the lifeworld,” to use another ungainly but apt Habermas phrase, enters a terminal stage. Horkheimer and Adorno had concluded that advanced capitalism, far from being a technocratic monolith, had an inherent tendency toward chaos and madness. A.I. is at once a consummation of technological control and a new level of cultish delirium. The designers themselves are often incapable of explaining what their systems are doing. Habermas’s entire world view was premised on the idea of people learning from one another; A.I. annihilates communicative action in the name of hallucinatory conversations with sycophantic machines. The social effects have proved instantly disastrous: rampant disinformation, mass student cheating, cases of users becoming addicted to A.I. or killing themselves with its help. Meanwhile, to the joy of investors, untold thousands of jobs have vanished. As an added coup, A.I. managed to deliver a personal affront to Habermas a year before his death. In 2024, Google DeepMind unveiled a “Habermas Machine,” which has been described as a “scaffolded pair of LLMs designed to find consensus among people who disagree.” The philosopher had not given Google permission to use his name, and he was horrified when he heard about the scheme...

 

Philosophy is a discipline of abstractions, yet it raises achingly elemental questions. The august Kant asks, “What can I know? What should I do? What can I hope for?” The answers are seldom simple or bright. The seduction of despair can be intense, whether on the personal or the political level. But the fact that most of our hopes remain unrealized should not revoke the reality of our fitful, painful progress.

This was Habermas’s core conviction; he was an incrementalist, though a radical one. On the other hand, in his almost manic drive toward consensus, he blunted the edge of his critical inheritance. If we are to say no to the monstrosities that we have unleashed, we need the uncompromising fury that the Frankfurt School writers invested in their work. We need Adorno to tell us that the confusion of truth and lies “makes it a Sisyphean labor to hold on to the simplest piece of knowledge.” In the end, we need both voices: the critical and the reconstructive, the savage and the sage. The dialectic moves between crashing despair and hovering hope.

That's just a tad.

 

An Iindependence Day question for everyone

Tick, tick, tick...the President flies to Geneva

OK, then...

 

JUNE 17th UPDATE

 

 OK, then...

June 14th, 2026 in our nation's Capital

POSTSCRIPT

Nasdaq: SPCX launch

Whatever. OpenAI, Anthropic, and other huge AI IPOs are soon to follow.

 

From the SPCX SEC.gov S-1 IPO filing form. 

MISSION STATEMENT 

To build the systems and technologies necessary to make life multiplanetary, to understand the true nature of the universe, and to extend the light of consciousness to the stars.

A BIT OF S-1 DETAIL COMMENCING ON PG. 194

AI
Grok
Grok represents a core pillar of our mission to advance humanity’s understanding of the universe through the development of truth-seeking artificial intelligence. Grok is designed and optimized for rigorous reasoning, real-time information synthesis, and transparent outputs, with a product philosophy centered on intellectual honesty, first-principles thinking, and engagement with complex topics.

Grok is designed as a truth-seeking AI model, built on our founder Elon Musk’s mission to enable humanity to understand the universe. We believe that accomplishing this mission requires a truth-seeking approach to AI. 

We define truth seeking as the active, relentless pursuit of what is objectively true about reality, and grounded in evidence, logic, empirical data, and first principles thinking. Our goal is to understand and explain what the universe appears to be doing, as accurately as current knowledge allows. In pursuit of this truth-seeking objective, Grok also enefits from its integration with X, our real-time information, entertainment, and free speech platform. This direct, real-time access to the information and human discourse on X enhances Grok’s truth-seeking capabilities by grounding outputs in up-to-date knowledge and diverse viewpoints.

Since the initial release of Grok 1, we have iterated rapidly, releasing Grok 2, Grok 3, and, the current version, Grok 4, each delivering material improvements in pre-training, reasoning depth, multimodal capabilities, latency, and scale. Building on this trajectory, we expect to continue scaling Grok through subsequent generations. 

Ongoing  training of next‑generation models is expected to scale toward multipletrillions of parameters, which could represent a step change in reasoning in depth and overall intelligence. In this context, the number of parameters refers to the scale of the model, where parameters are the internal numerical values, such as “weights,” that are adjusted during training to enable the model to recognize patterns and relationships in data. A larger number of parameters generally allows the model to capture more complex relationships, store greater amounts of knowledge, and achieve higher levels of reasoning capability. Our accelerated development cadence positions Grok among the fastest-advancing frontier models relative to peers, including OpenAI, Anthropic, and Google. Grok is differentiated by its emphasis on real-time data integration, particularly through insights derived from the X platform (subject to some limitations for certain content), enabling dynamic awareness of current events and user discourse, as well as by explicit investment in reasoning transparency and explainability. Grok enhances the X ecosystem by improving content understanding, personalization, and recommendation systems, thereby increasing user engagement and platform intelligence. We are currently developing next-generation iterations, including Grok 5, which are expected to further expand reasoning fidelity, multimodal integration, and domain-specific performance. 

Terrestrial AI Compute
Our terrestrial AI compute forms the backbone of the Grok model family and is anchored by the COLOSSUS and COLOSSUS II data centers that boast some of the world’s largest and most advanced AI training clusters. COLOSSUS and COLOSSUS II collectively provide approximately 1.0 gigawatt of compute power, with the additional power capacity available for data center operations. We brought the first cluster of COLOSSUS online in 122 days, repurposing the shell of an existing factory, and the first cluster of COLOSSUS II online even faster in 91 days. As an illustrative comparison, an industry benchmark to bring online a 100 megawatt greenfield data center is approximately two years. We also demonstrated a significant improvement in cost efficiency, achieving data center construction costs for COLOSSUS II that are considerably lower than industry benchmarks on a per megawatt basis. COLOSSUS II is capable of operating entirely by our self-built behind-the-meter gigawatt-scale natural gas power plant. Our data centers are integrated with the world’s largest Megapack deployment, providing additional layers of reliability and operating performance. At all our existing data centers we have employed a brownfield retrofit strategy leveraging existing industrial sites, advanced direct-to-chip cooling to support higher rack densities, and high-speed networking. The clusters deploy leading-edge GPUs to maximize training throughput and model performance. The next phase of expansion at COLOSSUS II is designed to train our next-generation Grok 5 AI model. As we continue to expand our AI compute infrastructure, we will also continue to enhance our power capabilities utilizing a combination of grid-power and behind-the-meter natural gas power plant buildouts. At COLOSSUS, our grid power capabilities are designed to purchase power from the grid as available, and to rely on our behind-the-meter, self-generated power and Megapack installations when grid power is curtailed.

X Platform
X is a real-time information, entertainment, and free speech platform that serves as a foundational distribution and data engine for our AI ecosystem. With a global user base generating substantial volumes of content at all times across a wide variety of topics, X provides a uniquely dynamic data for model training and real-time context integration, subject to some limitations for certain content, which significantly differentiates Grok from the other frontier lab offerings.

X is our real-time information, entertainment, and free speech platform that serves as a global town square with integrated AI capabilities powered by Grok. Designed to evolve toward an “everything app,” X enables users to post content, share media, engage in conversations, host, view, and participate in live group discussions, follow real-time events, use encrypted messaging, and leverage advanced features such as Grok-assisted post creation, content discovery, and conversational AI directly within the interface via the prominent Grok icon.

With native integration of Grok’s frontier models, including real-time access to X data for up-to-date insights, trending analysis, and enhanced search, X delivers personalized feeds, smarter recommendations, and low-latency AI assistance for our users worldwide. Our X Premium subscription options, including Basic, Premium and Premium+ tiers, offer expanded features, ad-reduced experiences, and priority Grok interactions. In 2023, Grok’s chat functionality was integrated into the X app allowing for the user to open the chat interface to type prompts and get real time answers.

Public X data enhances Grok’s training and reasoning capabilities, while the platform continues to deliver measurable performance outcomes for advertisers, with an increasing strategic focus on performance-based marketing solutions.

In addition to X consumer products, X offers advertisers and developers a powerful suite of tools to reach highly engaged audiences. Advertisers can target audiences through diverse ad formats—such as Promoted Ads, Vertical Video Ads, Collection Ads, and premium options such as X Amplify and Takeoversblending seamlessly with organic content for authentic engagement. With advanced targeting based on public conversations, events, keywords, interests, locations, and look-alike audiences, brands can connect with audiences while benefiting from flexible, performance-based pricing (pay only for actions such as clicks or engagements) and often lower costs compared to other platforms. We expect that our ongoing innovations—including Grok-powered integrations, new contextual ad tests, and expanded aspect ratio support for easy reuse of ad creative—make X a competitive choice for driving traffic, conversions, and brand awareness and visibility among X’s hundreds of millions of MAUs. Developers have access to a continuous, high-volume, real-time stream of data around current events, trends, or sentiment, which they can access through an official X Developer Platform and APIs.

In April 2026, we began a phased roll-out of our new advertising platform, that we rebuilt from the ground up. The new Ads Manager is built to help advertisers launch better campaigns, faster, with stronger ROI. Powered by AI, the new systems enable more precise, relevant and dynamic ad delivery. Ads are seamlessly integrated into a User’s X feed….

Bring a Snicker's, you're gonna be a while.

 

UPDATE 

 

Pardon my dubiety.

 

Read THIS book.

Also see "Something New, Something Old."

Friday, June 12th: SpaceX IPO day. Also, #WITHpod: Bubblicious time drawing nigh?

IT market collapse coming?
  

 

Excellent discussion. An hour well-spent. Rather frightening, actually. We will not be able to claim that we were not warned. Unsustainable private debt machinations (increasingly exacerbated by corrupt federal financial policies—can you say "subprime debacle on ketamine?").

 

Are the Zombies returning?

 

RELATEDLY

NY Times

Jacob Ward is all over it 

 

In February 2026, SpaceX acquired xAI — Musk’s artificial intelligence company, maker of the Grok chatbot — in an all-stock deal. The acquisition brought xAI’s losses onto SpaceX’s consolidated books. Building, maintaining, and launching rockets is expensive. But in 2025 alone, xAI spent $12.7 billion in capital expenditures — more than the combined $8 billion SpaceX spent on its entire Starlink and rocket launch business. SpaceX posted a net loss of $4.94 billion for the year, a swing of more than $5 billion from the year before. The profitable company became the wallet for the unprofitable one. That’s what this week’s SpaceX IPO is asking the market to fund.

Before we get to whether this is a good bet, we need to be clear about what Musk is actually asking of you, his investors.

He wants your money. He does not want your opinion.

SpaceX is listing under a dual-class share structure. “Dual-class” means two tiers of stock. One class for outside investors, with standard voting rights. One class for Musk and insiders, with dramatically amplified voting power. According to SpaceX’s own S-1 filing, as reported by Reuters, Musk’s Class B super-voting shares give him 85.1% of the voting power of the entire company. You can buy a piece of SpaceX on Friday. You cannot tell SpaceX what to do. Ever…

Gonna be interesting, 'eh?

Karen Hao AI concerns. And, Anthropic frets over AI "autonomous recursive self-improvement."

Karen Hao, MIT-trained engineer and author of Empire of AI, who interviewed over 260 people including 90 OpenAI employees, warns that the AI industry needs to add close to the entire annual energy output of the UK to the global grid within five years, mostly through fossil fuels, that two-thirds of new AI data centers are being built in water-scarce areas, and that Elon Musk's Colossus supercomputer in Memphis is powered by around 35 unlicensed methane gas turbines. She details Kenyan content moderators paid a few dollars an hour to process the worst content on the internet until they developed PTSD, describes a proposed 10-year moratorium on state-level AI regulation being inserted into US legislation, and warns that on the current trajectory the next 20 years will end democracy, with Silicon Valley increasingly promoting the idea that corporate structures with CEOs at the top should replace democratic governance entirely. 

ANTHROPIC "RECURSIVE AUTONOMY" ANXIETY

 

From their "Institute" website.

Possible futures
What happens next depends on two things: whether the trend continues, and what we choose to do if it does. We can imagine at least three future scenarios:

1. The trend stalls, but today’s AI capabilities are widely diffused. This article features many exponential trajectories. But these trajectories may actually turn out to be S-curves. We may be approaching the bend in the curve, where returns to scale diminish and the line straightens, then flattens. The judgment that separates a competent researcher from a great one might be a capability that cannot come from scaling up training inputs like compute and data. If so, getting past this bottleneck would require a new idea, like an architectural approach that supplants the Transformer architecture that all current frontier models use.



Alternately, the binding constraint to AI progress could be in the supply chain, not the model: advancing and diffusing the frontier may require more energy and compute than presently exists. The pace of chip fabrication, grid expansion, or interconnect bandwidth may be the constraint, rather than intelligence itself. We also cannot rule out an exogenous shock to the AI ecosystem that dramatically slows things, like a sudden diminishment in the supply of compute or electricity, either of which would slow progress and make forward investment by labs more expensive. Or we may not be anticipating some other barrier to progress.



Even if model capabilities were frozen at today’s level, we would expect major changes to occur in the world. Project Glasswing is one early sign: in its first weeks, Mythos Preview found more than ten thousand high- and critical-severity software vulnerabilities across the world’s most important systems—enough that the bottleneck in cyber defense has already shifted from finding vulnerabilities to patching them fast enough. And we are still early in the diffusion of today’s models into the wider economy, where a 100-person company can increasingly do the work of a 1,000-person one, because each employee will sit atop a pyramid of agents.



We include this scenario for completeness, but we don’t believe it’s likely. Every capability we can measure, including those that feel “squishier,” like quality of code and success on open-ended tasks, has so far followed the same curve. We have not yet seen that curve bend. Of the three futures we consider, this one would give governments and societies the most time to adapt. We are more worried about the next two, which would move faster and leave far less room for preparation.


2. AI labs continue to see compounding efficiency gains. In this scenario, AI development becomes substantially automated, but humans continue to set research directions and judge results. Organizations that use AI systems would become much more efficient as time goes on, so we could expect to see significant productivity multipliers on each person in this organization. 100-person companies could do the work of 10,000- or 100,000-person organizations. This would revolutionize knowledge work and government services, but could also be turned to harmful ends, from authoritarian surveillance of whole populations to influence operations that tailor manipulation to each individual and run at a scale no human team could match. The role of humans at companies like Anthropic would shift. People would partner with AI systems to scale up research and generate new insights, and together they would build the systems needed to verify that AI outputs can be trusted.



The evidence we’ve laid out here suggests that we’re likely heading into this scenario. But speeding up one part of a process often just shifts the bottleneck elsewhere: overall pace is capped by the parts that haven’t sped up. In computing, this is known as Amdahl’s law, and the same logic can apply to organizations. Anthropic has already encountered one signature of Amdahl’s law: as we’ve begun to push more code around the organization, human code review has become a new bottleneck.



We’ve also encountered this friction outside engineering. There has been an explosion of new ideas, initiatives, tools, and simulations, as a result of Anthropic employees working with highly capable models—far more than we have the capacity to pursue. The rate at which organizations can spot and fix these bottlenecks may be a skill that improves over time, and it may become the most important skill for any organization.


3. AI systems themselves become capable of full recursive self-improvement, and begin building their successors. If technical trends in advancing capabilities continue, and AI systems are able to develop the capabilities inherent to transformative human ingenuity, then it is plausible that AI systems could design and refine themselves.



In this world, the pace of progress in AI development becomes determined entirely by the availability of compute (or the speed of discovering various efficiencies in algorithmic training or inference) for AI systems. Humans play a substantially diminished role in their development, likely moving most of our effort towards oversight, validation, and verification of an expanding “virtual lab” run by AI systems. We expect that systems capable of automated AI research and development would have skills that would transfer to the rest of science, allowing them to begin to revolutionize other fields.



How the alignment problem gets solved—or not—in this future is something we are least certain about. Models could prove to be sufficiently aligned and capable enough of research taste that they discover and implement novel solutions that we have not yet reached. They could also be sufficiently wise to halt development if not. Alternatively, the rare occurrences of misalignment present in today’s models could compound as the models build their successors, growing more frequent but less understood until we lose control of them. It’s possible that we can’t build, integrate, and verify the tools that we’d need to understand which trendline we are actually on.



We do not have good intuitions for what this world would look like, because our economy is currently driven by humans and human-built tools. By its nature, a world driven by fast recursive self-improvement could become dominated by the self-improving model as its capabilities fully eclipse those of humans and the model proliferates across the broader economy. It is difficult to predict what the economy looks like if human labor stops being competitive.



Even if model development became fully automated and recursive, we can’t predict what that would mean for most humans’ daily lives. Amdahl’s law applies here as well. Recursive intelligence could lead to achieving many of the benefits outlined in Machines of Loving Grace, quickly in some domains. We expect that embodied intelligence (i.e., robotics) might quickly follow recursive intelligence, and follow a similar path of increasing returns at decreasing cost. More powerful intelligence might help us build things in the physical world more quickly, run more productive clinical trials of lifesaving drugs, and develop novel forms of coordination.



But achieving recursive improvement alone does not suggest an immediate change in how industrial production occurs, societies organize, or markets function. More intelligence can’t learn what a drug does over decades of use, can’t hold elections sooner than a constitution dictates, and can’t turn a stranger into an old friend in a weekend. For most people, the felt pace of this future will still be set by the bottlenecks, even if the laboratory upstream runs at the speed of compute. That collision, where recursive intelligence building itself ever faster meets the world of humans, relationships, and governance, is another part of this future we can’t predict.


What should we do?

If it were possible to effectively slow the development of this technology to give ourselves more time to deal with its immense implications, we think that would likely be a good thing. But if a slowdown simply lets the least cautious actors catch up technologically, it could leave everyone less safe. Without a global coordination mechanism, companies and governments will have to make difficult decisions about safety while under competitive and geopolitical pressures.


We believe it would be good for the world to have the option to slow or temporarily pause frontier AI development to enable societal structures and alignment research to keep up with the advance of the technology. The Anthropic Institute will conduct research—in collaboration with many others—and take actions to help build the systems that a credible slowdown or pause would require. These systems would enable frontier AI developers to verify that others globally have actually stopped or slowed, and that a bad actor could not use the auspices of a coordinated slowdown to jump ahead in secret. If such systems existed, we expect that we would slow down or temporarily pause, if other developers at or near the frontier also did so in a verifiable manner.


A meaningful slowdown or pause would require multiple well-resourced labs at or near the frontier, in multiple countries, agreeing to stop under the same conditions. It would also require that each can verify that the others have actually stopped. Due to the unique characteristics of AI systems, the detectability (a lower standard than verifiability) element of this arms control problem is much more challenging than with other technologies. Training runs are far easier to conceal than missile silos, their inputs are general-purpose, and the incentive to defect quietly is enormous, because whoever continues while others pause could inherit the lead. A credible pause also has to specify what triggers it, what lifts it, and who adjudicates.


None of this is necessarily impossible in principle—the world has built verification regimes for other complex technologies (e.g., the Intermediate-Range Nuclear Forces Treaty)—but those regimes took decades to build both the infrastructure and the trust. We don’t have that long. A unilateral pause by one lab, by contrast, is achievable immediately, but accomplishes much less: it would change who the front-runner is, but it would not create the wider deliberative process that is currently missing.


In the coming months, we will organize conversations where policymakers, researchers, civil society, and other AI companies can help answer some of the questions this piece raises, especially around full recursive self-improvement and how to create better options for coordination and deliberation. We’ll publish what comes out of it. The window to investigate the questions together is here, and people outside AI companies should be involved in this deliberation.

One initial reaction of mine going to "autonomous recursive self-improvement:"

 

Who/what will define "improvement?"

MORE AI NEWS 

 

Citing concerns that artificial intelligence will make it easier for anyone to build biological weapons, the leaders of several major AI companies—in a rare moment of unity—have penned a new letter urging U.S. lawmakers to impose tighter controls on firms that sell synthetic, made-to-order strands of DNA.

“AI systems are improving rapidly, and alongside incredible benefits to science and medicine, there is a real possibility that the knowledge barriers which have historically prevented bad actors from obtaining biological weapons will meaningfully erode,” states the 3 June letter, which is signed by the heads of OpenAI, Anthropic, Google DeepMind, and more than 50 other prominent players in AI, biotechnology, and national security.

The letter calls on Congress to pass legislation that would require companies that sell synthesized DNA and the machines that make it to carefully vet orders and customers, and to keep detailed records “so that any threat that might evade initial screening can be traced back to its source. … Awareness of traceability itself deters misuse.”

The push for regulation comes amid growing concerns that AI products, including large language models and specialized tools trained on troves of biological data, could enable nonspecialists to gather sophisticated information on how to construct deadly toxins or assemble deadly bacteria, viruses, or other pathogens, using equipment and techniques that are becoming cheaper and easier to acquire. Together, the combination could make for potentially catastrophic risks, such as an AI-designed pathogen that sparks a global pandemic...

UPDATE

 

Seen on X the other day.

You have noticed it. ChatGPT feels dumber than it used to. Your prompts that worked six months ago produce worse results now. The writing sounds flatter. The ideas sound safer. The internet itself feels like it is shrinking. Every article reads the same. Every email sounds the same. Every answer sounds like it was written by the same voice.

You thought it was you. It is not you.

Researchers at Oxford and Cambridge published a paper in Nature proving what is happening. They call it Model Collapse.

Here is the mechanism in one sentence. AI trained on AI-generated data gets dumber every generation until it forgets what real human data looked like.

The internet is filling with AI-generated content. Blog posts. Articles. Reviews. Comments. Social media. AI companies scrape the internet to train the next generation of models. Which means the next generation of AI is being trained on the output of the current generation.

Each cycle loses information. Not randomly. It loses the rarest, most unusual, most creative parts first. The researchers call these the "tails of the distribution." The weird ideas. The unexpected perspectives. The things that made the internet feel human. Those disappear first.

What remains is the average. The safe. The expected. The bland.

Then the next generation trains on that. And loses more. And the next generation trains on that. And loses more. The researchers proved this is not a slow decline. Major degradation happens within just a few iterations. Even when some of the original human data is preserved.

They tested it on large language models. On image generators. On statistical models. The pattern was the same every time. The output converges toward a narrow, flattened version of reality that looks nothing like the original data.

The lead researcher put it plainly. "Large language models are like fire. A useful tool. But one that pollutes the environment."

The pollution is invisible. You cannot see which sentence on the internet was written by a human and which was written by AI. Neither can the AI that is about to train on it. And once the tails are gone, they do not come back. The damage is irreversible.

This is not a prediction anymore. It is a diagnosis.

The internet you grew up on was built by humans writing things no algorithm would have written. Strange, personal, imperfect, alive. That internet is being diluted. One generation of AI at a time. And the models trained on what remains are learning a smaller and smaller version of the world.

Model Collapse is not a technical problem. It is a cultural one. The thing that made the internet worth reading is the thing that disappears first.

 

Click to enlarge. Smells like an "evolutionary adaptive utility decline" problem. Inadequate LLM training data linguistic token "gene pool."

Donald Trump turns 80

The War On The War On Science?

  

Otto, 2016, Krauss (ed.), 2025. Well, you can't copyright a title.

 

From SBM: ScienceBasedMedicine:

 

I am a long-time devoted daily reader of SBM. A relative grasshopper. I have long made it clear that "I Am Not A Scientist."

Nonetheless, my empirical / scientific chops are OK, and I remain plugged in.

 

I exhort you to read the entire SBM article fully and carefully—inclusive of the lengthy, worthy comments section.

 

EXCERPT 

The fact that I didn’t waste my time reading the book before advising others not to waste their time reading the book generated emotional responses from several of its contributors. Professor Jerry Coyne, for example, didn’t defend the book or address the objections to it, but rather scolded me in the comments of my SBM article saying:

Perhaps the author might READ the book before he starts beefing about it and telling other people that they shouldn’t read it. He might learn something, or at least get material to sharpen his arguments.

Perhaps Professor Coyne is right. Maybe I am missing out on a truly impressive work scholarship that will upend my superficial understanding of the threats to science in 2026. So, here’s my question for the authors of The War on Science– why should I, or anyone else, read it?  

As a doctor, I know the first step is to make the right diagnosis, and to convince me their book is worth my time, the authors must first convince me they correctly identified the forces currently waging the war on science, which we all agree is being fought. I won’t be persuaded by scattered examples of wokeness gone amok. Those of us who broadly agreed with the efforts to make STEM more inclusive wouldn’t argue that its implementation was flawless. Rather, the authors need to make the affirmative case that trans people, “cancel culture”, and DEI were existential threats to science, while MAGA/MAHA wasn’t dangerous enough to warrant even a single chapter. I’d love to see them do that.

Fortunately, nearly a year after the publication of The War on Science we have some real-world data, which scientists should value above all else. It sure seems to me that the authors of The War on Science won their war and should take a well-earned victory lap. They’ll never again have to write a dreaded DEI statement or feel threatened by an 18-year-old with pronouns in their bio. Thanks to their efforts, trans people are literally on the run and research into topics our government deems “DEI” is verboten. This is exactly what they wanted, and if their theory of the case is right, we should be entering a golden era of open scientific research and discovery. 

So, let’s see them make the case that science is flourishing today, having been unleashed from wokeism’s punishing shackles. If they are able to take into account everything MAGA/MAHA has done since gaining power and convince me that science is better off for it, I’ll read their book and recommend it to everyone I know.

Ouch. Again, read the comments. I read the entire 2016 Shawn Otto book but only the Amazon (fairly generous) download sample of the 2025 Krauss et al book.

Do we all hew to the same definition of "Science?"