I do not think that William Pulte is the proper appointee to the position of Director of National Intelligence (DNI) but my reasoning is somewhat different from what either the Republicans or Democrats have presented.

Under ordinary circumstances arguing for a “seasoned intelligence professional would be appropriate but present circumstances are not ordinary. Ignore Trump. How should an ordinary, prudent president proceed? The problem is that we do not know that the DNI has been “weaponized” by Democrats and we do not know that the DNI has not been “weaponized” by Democrats.

We do know that the intelligence itself was sound (the bipartisan Senate committee found that) and we do know that the procedures followed were quite unusual. We do not know whether the decision to commission the intelligence was politically motivated or not.

Under the circumstances the ideal candidate for the position would combine independence from the intelligence bureaucracy with sufficient knowledge to evaluate its practices. Such individuals are rare and bipartisan trust may be impossible to achieve in the current political climate.

To my eye William Pulte has not demonstrated the combination of independence, credibility, and commitment to evidence that the current moment requires. The country does not need a defender of the intelligence community nor does it need a prosecutor of the intelligence community. It needs someone capable of determining and convincing the public what actually happened.

I’m feeling a bit as though I had been preempted. I had been preparing a post on President Trump’s “anti-weaponization” fund but his announcement that he was not implementing it has made that unnecessary. I’ll just mention a few of my thoughts.

First, I thought that the president was pointing to a very real injustice—that the Supreme Court had found that the federal obstruction statute had been applied too broadly. But it was the wrong remedy for the problem. Those who pay taxes and those who feel monetization of the debt most acutely were paying for something over which they have little control while for those doing the enforcing the incentives did not actually change.

My ideal situation is that the law be enforced vigorously and even-handedly and those doing the enforcing be held responsible if they do so unjustly. With the “anti-weaponization” fund, we are no closer to that ideal than without it.

There has been an ongoing conversation in comments about wars in the post-war period. Since 1945, very few major wars have ended in the sort of decisive, politically unambiguous victory that characterized World War II. The Falklands War and the Gulf War are among the clearest examples. Others haven’t been wars, they’ve been single operations or counter-insurgencies, or haven’t been victories as such, they’ve been withdrawals, ends to the hostilities.

Take the Korean War, for example. It was definitely a war but there was not a clear victor and North Korea and South Korea have been separated ever since.

An example that has been raised is the Soviet intervention in Afghanistan. I would characterize it primarily as a counter-insurgency rather than a conventional interstate war. Whatever label one chooses, it did not end in a decisive victory; the Soviets eventually withdrew.

There were notable similarities among the Gulf War, the British-Argentine War, and World War II. The similarities include:

• All involved what the ultimate victors deemed to be unacceptable territorial aggression, Germany’s invasion of Poland in 1939, Saddam Hussein’s invasion of Kuwait in 1990, and Argentina’s invasion of the Falkland Islands in 1982.
• All involved naval, air, and ground forces.
• All involved conventional militaries.
• All involved alliances, e.g. the Allies in WWII, the UK, US, and France in the war against Argentina, and a broad coalition in the Gulf War.
• All resulted in decisive victories.
• The victor used overwhelming force against the defeated and had clearly enunciated goals.

I would be remiss in not mentioning a very different explanation of war and success which is that all wars involve some degree of what we are increasingly calling criminal behavior. Before the start of World War I there was an accord to which all parties were signatories, the forbade the use of chemical weapons. That was violated by both sides, increasingly so as the war wore one. In World War II the Allies treated the morale of the Axis nations as a legitimate target of war. That was the rationale for the firebombing of Dresden and the Tokyo air raids, which killed tens of thousands of civilians. General Curtis LeMay later observed that if the United States had lost the war, those responsible for the bombing campaign might well have been prosecuted as war criminals. Whether one agrees with that assessment or not, it illustrates how even the most successful and widely supported wars often involve actions whose legality and morality remain disputed decades later.

Even the “good wars” remain controversial. Critics argued that coalition bombing in Baghdad stretched the laws of war, while opponents of the sinking of the ARA Belgrano contended that it violated the spirit, if not the letter, of the declared exclusion zone. These are rationalized as mistakes.

The longer a conflict continues, the greater the probability that mistakes, misjudgments, or violations will occur. Don’t be surprised if, as the war against Iran wears on, we make more mistakes.

I strongly recommend you read this post from the Multipolarity Substack. Here’s a concluding snippet:

Dr Powell’s model suggests that the war is rapidly nearing a tipping point. Ukraine is approaching the fulcrum beyond which the weight of the war tips its forces into rapid depletion, allowing the Russian military to start achieving much more significant territorial gains. In turn, this would lead to accelerated Ukrainian losses through the ruthless reality of the square law. This road heads toward an eventual Ukrainian collapse.

Ukraine’s effective combat power (a composite of manpower, machinery and munitions) is depleting, the model shows, at a net rate that outpaces its replenishment, while Russia’s holds steady or grows marginally. This imbalance, compounded by recent reductions in Western support, suggests a tipping point where Ukrainian force density thins below viability, triggering rapid territorial losses and operational collapse.

That could be foreclosed by, for example, Western troops from any of the various NATO countries but that in turn risks widening the war into global thermonuclear war.

Ukrainian esprit de corps makes little difference. Innovations in warfare make little difference. What matters is how many troops can be fielded by each side.

The TL;DR version: come up with a minimally livable solution quickly.

It has been ten days since my right eye was operated on and almost 25 days since my left eye was operated on. Without eyeglasses I see roughly as well as I did five years ago (without eyeglasses). I use off-the-shelf eyeglasses for the near-sighted I purchased from Amazon for walking around and watching television. Reading, either with or without eyeglasses, is a strain. Driving, with or without eyeglasses, would be dangerous.

My vision is not monotonically improving. Today, for example, my vision using eyeglasses is worse than my vision using eyeglasses was yesterday. I have hopes that the new prescription eyeglasses I will receive after I heal in two-three weeks will enable me to read, drive, and generally function.

In an earlier post I had said that I was planning to elaborate on some of the ways that those who have developed artificial intelligence models like ChatGPT, Claude, and so on may not understand their own creations. This is that post.

The Sapir-Whorf hypothesis is simple to state but far-reaching in its implications. It is that language affects cognition. It comes in two forms, the strong Sapir-Whorf hypothesis, language constrains thought, or the weak form, language makes thinking in certain ways easier. There is substantial evidence that language influences cognition, particularly in perception, categorization, and memory and it has been widely accepted for nearly 60 years. It is the weak form to which I will refer in the balance of this post.

The key point is that It applies to generative artificial intelligence, too. Algorithms designed by native speakers of English and trained using content written in the English language will differ from algorithms designed by Mandarin, Yoruba, or Arabic speakers and trained using Mandarin, Yoruba, or Arabic content will differ in ways we can’t even predict. They can still be grammatically correct or be intelligible but they’ll feel “off”. They won’t uniformly make intuitive sense. Although English and German are closely related linguistically, models created by native speakers of German and optimized for German users will be subtly different from models created by native speakers of English and optimized for English-speaking users.

That won’t be a dealbreaker but it will impair their commercial viability, particularly in certain fields like medicine or law. I strongly suspect that this is an aspect that the developers of gAI models are not sensitive to. There is no cognitive layer on which the linguistic model resides. It’s the other way around. It also may not be something that American managers are sensitive to but that doesn’t mean that it doesn’t apply. It just means that models developed by non-English speakers will not be as useful to native speakers of English as they might be.

The precise linguistic mechanisms are beyond the scope of this post; the important point is simply that different languages encode meaning, context, categorization, and association differently, and models trained on those languages will necessarily reflect those differences. A model can produce flawless English and communicate clearly but still not feel fully native in its assumptions or reasoning patterns. That has serious implications for commercial utility.

That’s why, for example, I don’t worry too much about China’s competing with the U. S. in gAI. I have every confidence in the ability of native speakers of Mandarin to develop their own generative AI models. Models trained primarily on English-language content and optimized for English-speaking users will differ from models trained primarily on Mandarin-language content and optimized for Mandarin-speaking users. Is there enough Mandarin language content on which to train them? I don’t know. Indeed, I doubt that anyone knows. And if the models are train, again, using English language content the results will be subtly different. Whether these differences arise primarily from language structure itself or from the cultures embedded in the training corpora may not matter operationally; the resulting models will still exhibit different intuitions and priorities.

The Chinese are absolutely capable of developing their own models running on their own graphics processor units (GPUs). At the present state of technology those GPUs will require about twice as much electricity as GPUs built in Taiwan or the U. S.

The same is true of developing generative AI models in other languages as well, i.e. Farsi, Arabic, Tagalog, and so on. But the Sapir-Whorf hypothesis will continue to apply.

On May 15^th Pope Leo XIV published the first encyclical of his papacy, Magnifica humanitas (“wonderful humankind”). I wanted to summarize it briefly and make a few observations. I will refer to it here as MH.

MH is rather clearly and intentionally evocative of Leo XIII’s historic encyclical, Rerum novarum (“of new things”—RN) but it differs from the latter in some notable ways. Here are its themes by chapter:

Intro & Biblical Framework Two biblical images frame MH: the Tower of Babel (technology as prideful self-sufficiency that dehumanizes) versus Nehemiah rebuilding Jerusalem’s walls (shared, humble, community-oriented work with God at the center). The encyclical calls everyone to choose the “Nehemiah path.”

Chapter 1 This chapter is a lengthy survey of Catholic social teaching from Leo XIII through Francis, showing how each era produced a response to its “new things.” AI is this generation’s res novae.

Chapter 2 This chapter restates core principles, namely the foundational pillars of human dignity (rooted in being made in God’s image, not in productivity), the common good, the universal destination of goods, subsidiarity, solidarity, and social justice. Notably extends the “universal destination of goods” to include algorithms, data, digital platforms, and patents, arguing these cannot be monopolized by a few. The phrase “universal destination of goods” was first enunciated specifically in Vatican II.

Chapter 3 Technology and Power. Here His Holiness critiques the “technocratic paradigm” (efficiency and profit as the measure of all things). Warns against transhumanism and posthumanism, ideologies that treat human limitation as a bug to be fixed. Rejects the idea that AI can replace what is essentially human. Raises concern about private tech companies whose power now exceeds that of many governments.

Chapter 4 Truth, Work, and Freedom — Three specific threats:

• Truth: AI-driven misinformation endangers democracy; calls for “an ecology of communication” and stronger digital education.
• Work: Automation and AI-driven unemployment threaten human dignity; wages and meaningful work remain central moral concerns.
• Freedom: Digital addiction, behavioral manipulation, and algorithmic surveillance are described as “new forms of slavery.”

Chapter 5 Power vs. Love Here MH condemns the normalization of war, autonomous weapons systems, and the collapse of multilateralism. Calls for disarmament (including of rhetoric), diplomacy, and a “civilization of love” as the alternative to a “culture of power.”

As moral exhortation directed toward individuals, MH is often thoughtful and compelling. As a framework for political economy and governance, however, it raises significant unresolved questions. MH deviates from RN in that in the latter Leo XIII took pains to support the right to private property. MH omits that. As a guide for collective or political action, however, the omission renders it problematic.

Make no mistake, this is no minor omission. Large portions of the modern American economy depend on the creation, monetization, licensing, and protection of intellectual property and digital platforms. That includes the pharmaceutical industry and substantial chunks of the healthcare and legal sectors not to mention Google, Microsoft, and Facebook. The omission is significant because Catholic social teaching has historically balanced the universal destination of goods against an explicit defense of private property. MH appears to shift that balance toward social obligation without clearly defining the limits of legitimate ownership in the digital economy. My suspicion is that some clarifications to MH will be forthcoming soon.

RN defended private property partly because property dispersed power and protected mediating institutions against centralized authority. MH confronts a world in which digital property and network effects can instead concentrate authority at unprecedented scale. The encyclical therefore appears less concerned with property itself than with asymmetries of power emerging from informational monopolies.

Furthermore, calls for international governance introduce serious sovereignty problems for national governments. As I’ve noted before, effective world government implies global consensus and little could be clearer than that there is no such meaningful global consensus.

I wanted to remark on this Washington Post editorial.

On Thursday, President Donald Trump abruptly pulled the plug on the signing of an executive order that would have given federal agencies an early look at the nation’s most powerful artificial intelligence models before their public release. That’s good news: Even the supposedly voluntary review system under consideration could have hardened into a government chokepoint on U.S. AI development.

Here are some additional significant snippets:

The order would reportedly have created a framework for leading AI labs to share their most advanced models with a coalition of national security agencies before public release. It would also have established a joint cybersecurity clearinghouse, run by both government agencies and private-sector representatives, tasked with finding and patching vulnerabilities in the nation’s digital infrastructure before adversaries could exploit them.

and

A coalition of more than 60 Trump allies, organized under the banner of Humans First and backed by Steve Bannon, wrote the president urging mandatory government approval of frontier AI systems before release. Kevin Hassett, the director of the National Economic Council, publicly floated an FDA-style regime earlier this month in which AI models would have to be “proven safe” before they could reach the public.

and

AI poses real cybersecurity risks, but they need to be addressed with care and precision. The president is right to be wary of creating a bureaucratic review regime that erodes America’s edge.

I’m going to divide my remarks into four sections:

• The problem
• The urgency
• What the creators don’t understand
• The solution

The problem

The snippets above call out the national security implications and the “safety” implications. In addition to those there are infrastructure issues and liability issues. The infrastructure issues are highlighted by this graph:

U. S. consumption increased rapidly from 1970 to about 2000, plateaued due to various efficiencies realized, and, driven largely by generative AI and electric vehicles, began to increase around 2020. The rapid build-out of data centers and increased adoption of EVs will ensure that growth continues for the foreseeable future.

At some point utilization will exceed the grid’s capacity. That is likely to produce a series of effects. First, producers may implement controlled voltage reductions of 5–8% to prevent cascading failures. Historically, that has frequently been followed by “rolling blackouts”. Together those will produce wear-and-tear on devices containing electric motors, such as refrigerators, air conditioners, and well pumps. The costs of these effects will not be borne by commercial consumers whose contracts ensure continuous use. They will be borne by household consumers.

Neither executive agencies nor Congress possess the technical knowledge or adaptive speed necessary for effective ex ante control of frontier AI systems. We are in terra incognita. The machine learning researchers who build these systems understand the architecture and the training process. They do not understand and have said publicly they do not understand what the systems have learned, why they produce the outputs they produce, or what they will do in novel situations. That is an accurate description of the current state of knowledge.

The urgency

Basically, we can’t wait for studies or well-formed regulations. The first lawsuits against gAI companies are already being filed. Regional power constraints associated with accelerated data-center growth are already emerging in several U.S. markets, particularly in Texas where data centers are being built. The pace at which data centers are being built is considerable.

What the creators don’t understand

The enormous irony in this is that not only do elected officials or any foreseeable regulatory agency lack the knowledge and understanding, the creators of gAI don’t have the understanding, either. I’ll flesh that out in a later post.

The solution

Fortunately, there is a prospective solution but it’s one I believe with be very controversial and result in considerable backlash: strict liability. Under a regime of strict liability those harmed by gAI would not need to show intent or negligence on the part of the AI companies—all they would need to show are involvement of the companies and harm.

The response of the companies and, presumably, some economists would in all likelihood be that would have a chilling effect. That is an alarmist claim. Strict liability would force the industry to price risk realistically rather than externalizing it. Strict liability would alter incentives but that is precisely its purpose. The companies could continue to proceed at their own pace. It would only ensure that the risks of their actions would be borne by those who benefit most—the companies rather than those who benefit least—ordinary household users and consumers of electricity.

Not being able to read easily, see clearly, or drive is really a pain. Yesterday I walked over to Whole Foods and picked up a few things. That much I can do.

I’m not experiencing any discomfort at this point. At least I’m able to walk around. I wear the eyeglasses I bought from Amazon when I do—it’s a bit of a strain but I can see reasonably clearly when I do.

I figure I have about two more weeks of this incapacitation before I get a new pair of glasses which will enable me to read, watch television, and drive.