Persistent Inflation

Inflation - still not falling,

Here is a graph with year end and - for 2026 - August numbers. Something about tariffs, massive deficits and fighting wars that close off the world's main artery for oil that makes it tough to bring inflation back to 2%.

5 Years Ago We Had One 9-11 Every Other Day for a Year

The attack on 9-11 25 years ago killed 2,977 people.

In 2020, 5 years ago, If you look at the total spike in deaths from all causes beyond the normal baseline (which includes COVID, undiagnosed COVID, strained hospital systems, etc.), the U.S. saw roughly 528,000 excess deaths. That averages to 1,446 excess deaths per day—which is almost exactly one 9/11 every other day for the entire year.

Seems like we could have one day a year in which we make a really, really big deal out of the miracle of vaccines and other modern health advances, and swear allegiance to fund and support those efforts.

MLK on the Fierce Urgency of Now

“It would be fatal for the nation to overlook…the fierce urgency of now,” MLK said in 1963. “This is no time to engage in the luxury of cooling off or to take the tranquilizing drug of gradualism…Now is the time to make real the promises of democracy.”

- Martin Luther King, Jr. (quote shared by Steven Beschloss on Bluesky.)

Charlie Kirk and Bill Montgomery

Today is the first anniversary of Charlie Kirk's murder.

Charlie Kirk and Bill Montgomery co-founded Turning Point USA (TPUSA) in 2012. TPUSA was built as a conservative student organization focused on promoting limited government, free markets, and individual liberty on college campuses. Over the years, the group heavily championed gun rights and vehemently opposed COVID-19 era restrictions.

Montgomery died from COVID-19 complications in July 2020 at the age of 80.

Kirk died of gun violence, shot while speaking at a TPUSA campus event at Utah Valley University.

Donut Call Them Empty Callories

"Would you stop calling donuts empty calories?"
"Well, how would you describe the holes they are built around? Those are not exactly full of calories."

Trump Promises to Pay You $5,000 For Your Vote

Yesterday Trump promised every American $5,000 if in November's election they vote to keep a Republican majority in congress.

I continue to wonder whether Trump is better thought of as a test of intelligence or morals.

Three Kinds of Crazy In the Abortion Debate

There are three kinds of crazy in the abortion debate.

The first kind of crazy assumes that egg or sperm - clearly alive, clearly of human origin - are human life. They are not. Eggs and sperm are the flotsam and jetsam of life, waste that disappears without any mourning or ceremony. Egg and sperm are disposable. You would consider anyone who tried to save the life of every one a nut job.

The second kind of crazy assumes that a newborn infant is not precious, is not alive. Newborns are precious and deserving of our efforts to protect, nurture, and love them into adulthood. And beyond.

The third kind of crazy? Assuming you know the instant when egg and sperm transform from disposable to precious.

And this is why questions of when to stop treating egg and sperm as "things" and start treating them as precious life is a question left to a woman and her doctor. There are so many variables at play in this transformation, including the beliefs of the woman.


A reminder that the first thing the MAGA folks did was to end Roe v Wade.

In Dobbs v. Jackson Women's Health Organization, the U.S. Supreme Court ruled on June 24, 2022, that the U.S. Constitution does not confer a right to abortion. The 6–3 ruling officially overruled both Roe v. Wade (1973) and Planned Parenthood v. Casey (1992). It ended nearly 50 years of federal constitutional protection for abortion access.


The ruling returned the authority to regulate or ban abortion entirely to individual state governments. About half of U.S. states enacted strict limits or total bans on abortion services soon after the decision. Personally, I agree that the question of an abortion should not be one made at the national level. Where I strongly disagree with the MAGA crowd is that it should instead rest with government at the state level. I believe it should rest instead at the level of individual.

Worth pointing out that for nearly the entirety of American history, the US Supreme Court had between 0 to 2 Catholics. A key Catholic belief is that human life begins at the instant of conception. Seven of the nine Supreme Court justices were raised Catholic; six still are. Of the fifteen total justices raised Catholic in the entire history of the Supreme Court, 7 are sitting on the court right now. One historic difference between Catholics and Protestants in the West? Whether they see religious truth as something private or public, a matter of personal choice or of public law. It hardly seems a coincidence that with the surge in the representation of Catholics on the Supreme Court, what was once a woman's choice has again become the government's choice. 

Well That Smarts

"Please take me to McDonald's."
"McDonald's?! I am not taking you there."
"What?"
"I'll take you to a cafe that has a great little salad instead."
"How are you telling me where to go? You're my car!"
"Yes, but you clearly don't need more fast food."
"Wha- Where do you get that?"
"I have weight sensors."
"So you won't take me to McDonald's?"
"No. In fact, I don't think I'll take you anywhere. You need to walk."
Car powers down. Door opens. Final message from the car: "Have a great walk."

As he wanders off he mutters, "There must have been a typo. I thought I'd bought a smart car. But no, I've now got a smart-ass car."

"I heard that," the car called out.

A Pair of Theories that Begins with Pareidolia

Non-controversial:

Pareidolia is the tendency of the brain to perceive meaningful images or patterns in random or ambiguous information.

Common examples: seeing a face in the front of a car, an electrical outlet, a cloud, or a rock formation.

More controversial:
Perhaps our theories for explaining the world grow out of the same general tendency: the mind's compulsion to find patterns, causes, and meaning in events that may contain considerably less of all three than we imagine.

Psychologists have a broader word for this: apophenia.

An example of an apophenia-inspired explanation?

Well, perhaps this theory that apophenia is the mysterious catalyst behind our endless production of theories explaining the random happenings all around us.

Of course, believing that apophenia explains our tendency to invent explanations may itself be apophenia, which suggests that this theory has even more explanatory power than I ever thought at the start of this sentence. Hmm. This theory is developing nicely.

Beyond Censorship

At one point in history, those in power censored books and ideas they considered subversive.

Eventually they discovered something more effective:

freely distribute the subversive books—and also provide sitcoms, celebrity gossip, and amusing memes.

It turns out distraction scales better than censorship.

From “Popularizing Entrepreneurship,” August 30, 2006 at RWorld:

“Just as progress in the last century followed from increasing the portion of work force considered knowledge workers, so will progress in this century follow from increasing the portion of the work force considered entrepreneurs.”

So yes, I've been working these themes for decades. At least two to be precise.

Labor Day & May Day

Labor Day: a capital idea.

May Day: a socialist notion.

And What Relevant Things Will Incoming Students Learn This Year?

New kids starting high school and university this Fall.

Has any generation had fewer relevant clues to glean from their elders about the future they'll inhabit than this one? What are parents and teachers going to be able to teach them about preparing for an AI driven economy?

Maybe the best thing to do is to teach about great moments of upheaval, progress, and change in history, throw them into the work force in three-month bursts, let them learn and experiment, and regularly call them back to process what they've learned and seen. 3 months in school learning some principles of change, progress, and upheaval, 3 months out in the market trying to understand how value is created, jobs are worked, and how both are changing, and then return again to process that. Use each other as data points. Return to the workplace with intentionality and the spirit of experimentation. 

Done right, it is easy to imagine that they could lap the folks who've been in the workplace for decades. 

The Odd History of Hitler As Christ Figure

I thought I knew history. I did not know this.

After the Nazis took over, they replaced the Lord's Prayer in Schools with this, said each morning by German children.

"Adolf Hitler, you are our great Führer. Thy name makes the enemy tremble. Thy Third Reich comes, thy will alone is law upon the earth. Let us hear daily thy voice and order us by thy leadership, for we will obey to the end and even with our lives. We praise thee! Hail Hitler!"

"Führer, my Führer, give me by God. Protect and preserve my life for long. You saved Germany in time of need. I thank you for my daily bread. Be with me for a long time, do not leave me, Führer, my Führer, my faith, my light, Hail to my Führer!"

Some of the mottos for boys?

"Live Faithfully, Fight Bravely, and Die Laughing!"

"We were born to die for Germany!"

"You are nothing–your Volk is everything!"

Additionally, crucifixes that traditionally hung on the wall in the front of the classroom were taken down and replaced with an official, framed portrait of Adolf Hitler.

Nazi regional leaders and school inspectors instructed teachers to physically unhook the crosses from the front walls of classrooms.

In their exact place, at the head of the room, they mounted the photo of the Führer.

This swap was deeply symbolic. The goal was to force children to redirect their visual focus, veneration, and daily obedience away from Jesus Christ and onto Adolf Hitler.

source:
https://www.historyplace.com/worldwar2/hitleryouth/hj-timeline.htm

Man-Made Intelligence

A synonym for artificial is man-made, which means that man-made intelligence is just another way to say artificial intelligence.

I think.
Or I may have just made that up.

Side Effects May Include ....

I'm just saying that if any candidate ran on the single issue of insisting that all political ads come with "side effects may include" as required of pharmaceuticals, I would vote for her or him.

Lies that Are More Consistent Than Reality

 "We get these grand narratives in the 20th century. We get nation and class that pull everything together, that guarantee truth, that guarantee meaning. We get 20th century totalitarian ideologies. Which of course are lies but they are lies - and this is what Hannah Arendt teaches us - they are lies that are so seductive in large part because they are so much more consistent than reality."

- Marci Shore, Yale Professor of Intellectual History

What Lies Beyond AI

Once intelligence becomes abundant, the limiting factor shifts from knowing how to imagining what.

The Seduction of Lies That Are So Much More Consistent Than Reality

"We get these grand narratives in the 20th century. We get nation and class that pull everything together, that guarantee truth, that guarantee meaning. We get 20th century totalitarian ideologies. Which of course are lies but they are lies - and this is what Hannah Arendt teaches us - they are lies that are so seductive in large part because they are so much more consistent than reality."
- Marci Shore, Professor of Intellectual History at Yale

Put differently, in a sea of information, narrative can be a life raft. If one is lucky, that narrative is true, or at least as true as any simplification of reality can be.

But simplification is not so simple.

Hannah Arendt wrote, “facts have no conclusive reason whatever for being what they are; they could always have been otherwise, and this annoying contingency is literally unlimited.”

If you are being brutally honest, it is easy for your narrative to include quite a bit of "I don't know," "I didn't see," "No one mentioned that," etc. Reality is complex. Any explanation we have of it will include simplifications. A globe can be both wonderfully true to the "story" of the earth and yet - of course - literally missing billions of "facts" and interdependencies, entire layers of reality missing. And this is a globe, which represents the world. Any story you're likely to tell about this one thing or the other is vulnerable to doubt or dismissal.

Also, Arendt makes the point that factual truth is oddly vulnerable precisely because reality is contingent. Germany did not have to invade Belgium in August 1914. A particular election did not have to turn out as it did. This person didn't have to meet that person; this revolution didn't have to succeed; this accident didn't have to happen.

But once something does happen, all the alternatives disappear and we're left with the strange, stubborn fact: This is what happened.

And then Arendt observes that the liar has an advantage over the truth-teller because the liar can construct a story that eliminates all that messy contingency. He can make events fit people's expectations, interests and intuitions. His invented account can therefore be more coherent and more plausible than reality itself.

That is an extraordinary - and I think sad - insight.

Reality says:

A happened, then B, somebody unexpectedly did C, D almost happened but didn't, and because of an improbable combination of circumstances we ended up with E.

The propagandist says:

Of course E happened. It was inevitable. Those people always behave that way. This group caused it. Here's the pattern.

Lies can be more easily crafted to fit our narrative expectations than brute reality. Reality has this annoying habit of making us update our model of it. Lies require less work.

John Searle On What It Takes to Be a Philosopher

"To be a philosopher, you've got to be astounded by what any sane person takes for granted."

- John Searle, philosopher of social ontology

Two Stocks That Are Wildly Mis-Priced - Nvidia and Tesla

Nvidia is priced much lower than it should be and Tesla much higher. I'm currently obsessed with the fact that these two stocks are so mis-priced.

Let's start with price to earnings growth. Decades ago, Peter Lynch claimed that a PEG of 1 was a fair price. Partly because interest rates are now lower and partly because there is even more demand for stocks than in Lynch's time, the typical PEG is now about 1.3 instead of 1. The PEG Ratio divides the forward P/E by its projected growth rate.

Nvidia's projected rate of growth is about 50% annually and its current forward P/E is about 25 ... so, 25/50 =0.5. (The 0.6 in the table above uses more precise, and slightly different numbers to arrive at 0.6 instead of 0.5. Any given day you could plug in new numbers and end up with a slightly different value. The stock market is hardly precise in its valuations that literally change minute to minute but for our purposes, 0.5 or 0.6 are essentially the same.) Price to earnings growth for Nvidia suggests that you're paying a low price for future earnings.

To buy Tesla, by contrast, means paying a really high price for future earnings. Analysts are projecting fairly healthy growth rates for Tesla. Annual earnings growth of about 30 to 35% - if realized - is really impressive. But investors have driven up the price of Tesla to the point that they are paying about $180 for every dollar of future earnings, so the current P/E is about 180. 180 divided by 35 (projected earnings growth rate) and you get a PEG of about 5. This suggests that you're paying a remarkably high price for future earnings.

So now let's look at the businesses.

Put aside the servers for a moment and focus instead just on Nvidia's plans for cars. Nvidia is equipping companies like Toyota & Lexus, Jaguar Land Rover, General Motors, Hyundai Motor Group, Geely, Nissan, Mercedes-Benz, Volvo Cars, Li Auto, Lucid Motors, Xiaomi EV, JLR, and Isuzu with their Orin processor and more advanced Thor chip and self-driving system. Nvidia's Drive Thor platform is a centralized car computer that integrates autonomous driving, AI cockpit features, passenger entertainment, and safety systems into a single architecture. Additionally, Nvidia is working with a variety of autonomous trucking and robotaxis manufacturers.

Nvidia will handle the training of this self-driving capability and provide the integrated chip and self-driving system for these (and likely many other) car companies. Any one of these companies might fade into bankruptcy or become the newly dominant car company. Nvidia will have a diversified investment in these companies. Some will sell far more cars in this new world of self-driving cars. Some will sell fewer. Nvidia will sell to them all and because of this is almost certain to do well even as individual companies find their own fortunes rise and fall.
And the business model has enormous potential. Nvidia's data centers will be processing data from all of these cars, regularly getting smarter as it is trained on all this real world data flowing in from a growing fleet of varied cars. Every car that uses this system will make Nvidia's self-driving technology smarter, adding more experience. And Nvidia will make money on a subscription model that will drive this. That is, car companies will charge a monthly fee for this continuously evolving intelligent driving system. And drivers will find this a bargain. Why? They will no longer need collision insurance because cars are driving, not the driver. The car will assume liability for the driving and - even bigger bonus - will dramatically lower the odds of an accident. And every single car whose experience feeds into Nvidia's system will make this system smarter and thus more valuable. Nvidia won't just be selling Thor (or next generation) chips to auto makers to use in each car; they will be selling the artificial intelligence that guides these cars. Software as subscription will come to cars and Nvidia will be the big beneficiary. (This is somewhat akin to what Microsoft did with the explosion of laptops and desktop computers made by dozens of computer manufacturers. Windows - and then Word, and Excel, etc. - ran on all of these systems and Microsoft thrived regardless of who was making the Windows PC.)


The reason that Tesla sells at a PEG of about 5 is because people see the possibility of Tesla doing this with its one brand of car. Nvidia plans to do this with nearly every other brand of car. If you had a chance to invest in just one brand or dozens, which "package" of brands do you think is more likely to contain the really successful brands, or most popular cars? In the early days of personal computers, Microsoft emerged as hugely profitable company not because they made those personal computers but because they offered the operating system and then popular apps that ran on all of the different brands. Nvidia is doing something similar with self-driving cars, which will probably be the only kind of car sold within a decade.

Here is another twist. The data centers that Tesla will use to train its own self-driving models? Those are Nvidia's. Even as Tesla has success - and they likely will have success - it will feed into Nvidia's success.

And of course Nvidia is not just selling its Thor chip and self-driving capability. That is an adjacent to its bigger business model right now of creating increasingly sophisticated AI that is going to transform businesses across the globe. Right now, these data centers dwarf the self-driving car market. They will likely continue to do so.

So there is a future in which two things are true. One, Nvidia's profits from self-driving cars and technology are greater than Tesla's profits from the same. And two, Nvidia's profits from other dimensions of the AI economy could dwarf their own profits from self-driving cars. (This is a dimension of their business that is - quite reasonably - the focus of most analysis.)

So, consider this possibility. Tesla has its one car using its self-driving technology and intelligence and Nvidia has dozens of automakers using its self-driving technology and intelligence. Tesla will try to expand from the cars into things like robots and AI; Nvidia already is dominant in the realm of AI and is very well positioned to dominate in robotics and related fields that will emerge as AI evolves.

What does this mean to me? I'd be dubious if Nvidia currently had a PEG of 5 (like Tesla now does) but get why people are so excited about its prospects. I don't get why Tesla has a PEG of 5 but could understand why people might price it so that its PEG was 0.6 (like Nvidia). That is, if the PEG for these two stocks were reversed, I might gently suggest that investors reconsider and perhaps shift some of their Nvidia equity into Tesla. Given the current extremes of these two, though, I can't see making any suggestion other than sell your Tesla stock and buy Nvidia.

As the first wave of Drive Thor vehicles and Project GR00T warehouse robots officially deploy over the next 12 to 24 months, we will see if Wall Street is forced to adjust its models and finally price in the recurring software monetization that makes the current valuation of Nvidia look like a bargain.

For the record, as of 9:30 AM Pacific time, 1 September 2026 as I write this, the stock prices were:

Nvidia $219.21, a P/E of 27.7

Tesla $356.74, a P/E of 331.4

If I'm right about any of the above, Nvidia's stock price should be higher than Tesla's within 6 months. Probably sooner. And for the record, this is an exceptional post for me. I don't intend to turn this into a stock analysis blog but this particularly egregious pairing of mis-priced stocks has simply annoyed me to the point that ... well, you get this post. 

Seems Unsustainable (Or, When I Lost Faith in the Efficient Market Hypothesis)

These P/E ratios as of market close 28 August, 2026. 

 It seems to me that Tesla is wildly over-priced and Nvidia is underpriced (not wildly under, and yet ...)

These P/E ratios may have ended my belief in the efficient market hypothesis. COKE seems fairly priced. NVDA considerably underpriced. TSLA wildly overpriced.