Greetings and happy Tuesday, Sept. 22, 2026. Mark the calendar — autumn officially arrives tonight at 8:06 p.m. Eastern time, so enjoy these last few hours of summer while you can, meteorologically dishonest as that label already feels. Temps around Casa De Florack have been dropping like they were poleaxed these last several days. It's 48 degrees as I write this. My AC unit just sits there looking at me, wondering where its life choices brought this on.
My calendar says it's Hobbit Day, World Rhino Day, Public Lands Day, and Car Free Day. Celebrate the changing season, tip your hat to Bilbo and Frodo Baggins — both "born" on this date in Tolkien's own calendar — think kindly of a rhino, and maybe leave the car in the driveway for once, unless you're headed somewhere that actually requires four wheels and gasoline you already know the price of.
Today in History:
1776: The British hang 21-year-old Capt. Nathan Hale as a spy in New York during the Revolutionary War.
1789: Congress authorizes the office of Postmaster General.
1792: The National Convention proclaims the First French Republic.
1862: President Abraham Lincoln issues the preliminary Emancipation Proclamation, declaring enslaved people in rebel states free as of Jan. 1, 1863.
1911: Pitcher Cy Young earns his 511th and final career victory, throwing a 1-0 shutout for the Boston Rustlers over the Pittsburgh Pirates.
1927: Gene Tunney successfully defends his heavyweight boxing title against Jack Dempsey in the famous "long-count" fight in Chicago.
1949: The Soviet Union detonates its first atomic bomb, ending the American nuclear monopoly.
1950: Omar Bradley is promoted to five-star general, joining Dwight Eisenhower, Douglas MacArthur, George Marshall, and Henry "Hap" Arnold in that rank.
1961: NBC airs the first episode of Walt Disney's Wonderful World of Color.
1964: Fiddler on the Roof opens on Broadway, beginning a run of 3,242 performances.
1975: Sara Jane Moore attempts to assassinate President Gerald Ford in San Francisco, missing after a bystander deflects her arm.
1980: Iraq invades Iran, launching a war that will grind on for eight years.
Birthdays Today include: Andrea Bocelli, opera and pop tenor known for "Con te partirò"; Joan Jett, rock musician and Rock and Roll Hall of Famer known for "I Love Rock 'n' Roll"; Scott Baio, actor from Happy Days; Bonnie Hunt, actress from Jerry Maguire and The Green Mile; and Tom Felton, actor best known for playing Draco Malfoy in the Harry Potter films — if today's your birthday too, happy birthday.
* * *
On this past Sept. 14, I ran a column titled "Teach It Right, or Watch Beijing Teach It Wrong," making the case that the current anti-AI push from the usual suspects amounts to little more than political grandstanding and fear-mongering. I even went further, arguing that the computing power that makes AI's existence possible is growing so fast that no regulatory framework will meaningfully slow it down — and that someone is going to ride that wave, and the only real choice on the table is whether it's the Chinese Communist Party or us.
I called the anti-AI fear-mongering exactly what it looks like: a set of policies that, whatever their intent, would hand China the field by default, to everyone's detriment. I even raised the possibility that some of the left's pushback on AI isn't accidental to that outcome. We'll dive into that rabbit hole again another day because that problem isn't going away anytime soon.
For now though, I want to sharpen the philosophical side of the argument, because I don't think I gave it enough room the first time around. In that other piece, I focused on the technical end, establishing that the growth of technology and the level of computing power available is not ours to control, really. Government has never been able to legislate technology effectively, mostly because tech moves at a faster rate than government can ever hope to match.
But now, let's set that question, and the political questions aside. Concentrate instead on the fear-mongering itself, and grant, for the sake of argument, an assumption that isn't strictly warranted: That computing power eventually runs into no meaningful ceiling at all. Once you make that concession, the conversation stops being about speedy processors and training runs and starts being about a different set of questions entirely:
Evidence
Truth
Fairness
Harm
Responsibility
Acceptable behavior
Legitimate authority
Individual rights
Freedom
Human dignity
Justice
None of those are technical parameters. You can't benchmark them, and you can't patch them with a software update. They're philosophical choices — cultural values, the things that make us human, rather than merely computational — and I'd argue that's closer to the real heart of what normal people (that is to say, not politicians) are afraid of when they express their fears about AI. That's where I think we need to focus.
I generally resist reaching for Star Trek as a framing device (like the stars of the various Trek franchises who have been raising hell at Paramount over its increasingly far-left tilt, I cringe at the "woke" nonsense that the Hollywood types keep trying to inject into its scenes), but this one earns its keep.
The episode is "Emergence," Star Trek: The Next Generation, Season 7, Episode 23, which originally aired May 9, 1994 — a time when AI wasn't remotely on the table as a real technology, let alone a policy fight.
Like this discussion and like the premise of the entire genre of shows, the episode assumes that technology eventually grows several orders of magnitude past where it stood in 1994, or even where it stands today.
In it, the good ship Enterprise itself develops an emergent intelligence through no deliberate design on the part of her crew. Data explains that the ship's enormous database and its interconnected systems have somehow produced something greater than the sum of their individual functions — every subsystem working in concert, generating something new. The crew eventually realizes the intelligence is drawing directly on the experiences and information accumulated aboard the ship over years of service. It eventually detaches from the Enterprise and departs, leaving the crew and the ship more or less back to normal.
After it has been weaned more or less toward the end of the episode and leaves the ship, Picard and Data have this exchange, as Data wonders why the Captain didn’t order the intelligence destroyed:
PICARD: "The intelligence that was formed on the Enterprise didn't just come out of the ship's systems. It came from us. From our mission records, personal logs, holodeck programs, our fantasies."
Picard then asks, in essence, whether they can trust what emerged, because the experiences it drew from had themselves been honorable. The key line:
PICARD: "Now, if our experiences aboard The Enterprise have been honorable, can't we trust that the sum of those experiences will be the same?"
That's the whole ballgame, condensed into one captain's log entry. Picard's observation functions as both warning and reassurance simultaneously — trusting the sum of an artificial intelligence only makes sense if what it's learning, in its training and in its ongoing existence, is itself honorable. Which reduces the entire debate to one uncomfortable question: whose conception of truth and morality gets embedded in the machines that increasingly mediate human knowledge?
Here's where the analogy gets more useful the harder you lean into it. The Enterprise's experiences were never the sum of all human experience. They were the experiences of one particular crew, a curated subset of humanity, operating inside one specific, disciplined ethical culture — Starfleet's, refined across a fictional utopia built explicitly to be admirable.
Today's large language models don't have that luxury of curation. They're trained on an extraordinarily heterogeneous slice of actual human civilization: brilliant and stupid, noble and vicious, truthful and deceptive, compassionate and hateful, scientific and superstitious, all mixed together with no captain's wisdom to sort the honorable material from the rest.
That's where the real fear lives, once you strip away the political theater around it: What happens when the experiences being synthesized aren't uniformly honorable? Not "what happens if a chatbot says something rude" but rather, this: What happens when an entire civilization's contradictions — including its worst instincts — become the raw material an emergent intelligence draws its judgments from, at a scale and speed nobody has ever had to reckon with?
Look, Western culture isn't perfect, and I won't pretend otherwise. Heck, we both know that I make a fair bit of change writing every day about its imperfections. It has plenty of its own dishonorable chapters mixed into that training data along with everything else. But given a straight choice of whose values get instilled in the systems increasingly mediating what billions of people come to accept as true — ours, and the constitutional order that grew out of them, or that of the Chinese Communist Party — I don't think that's a close call. And that's the choice we face.
It’s crucial that we in the West ignore the fear-mongers and retain control of AI by being on top of the wave instead of under it.
Thought for the Day: Government can regulate an economy, license a profession, and tax a paycheck — but nobody's ever passed a bill that made a stranger love his neighbor, and the ones who keep trying to legislate their way there usually end up further from it than when they started.
VIP members: Time for you to chime in and give us your thoughts on the topic of the day.
Take care of yourselves, Gang. Maybe we can go leaf peeping this weekend. It's a bit early, but the colder weather here in the Northeast is as well. I'll see you here tomorrow.
* * * No computer was harmed in the writing of today's column.