There is a piece in this weekend’s Review by a self-described 30-something who views artificial intelligence with skepticism and dread, and who came home to discover that her septuagenarian parents had fallen completely in love with a chatbot. While younger family members asked friends for travel tips, the parents had the poor thing working overtime — holiday logistics, ice-cream stops, and an apparently bottomless stream of history trivia. At one point she wondered whether they might actually break it, because it could not bring itself to answer one more question about the origins of the Oval Office’s Resolute Desk. It proved astoundingly resilient.
This inverts the expected order of things, and the piece knows it. How funny, she thought, that young people prize old-school entertainment while boomers embrace the newest technology. Allison Pugh, a sociologist at Johns Hopkins, explained why it is less surprising than it looks: while younger generations use AI the most, older people take to it because these systems are far more user-friendly than, say, a television that requires what she called a born-to-it comfort with tech. Of all the generations, Pugh said, older folks seem to use AI with glee.
The review that should stop a financial adviser cold
Asked what she liked, Mom did not say accurate. She said: ‘He’s so polite. We appreciate that. We’re from a generation that curtsies and says please and thank you.’ And then the line that is genuinely worth sitting with — most important, she said, it is great at apologies. Which matters, the writer notes drily, given that it sometimes invents facts. When corrected, it said: ‘Oops! You’re quite right. Good catch.’
The Katherine of Aragon incident deserves preserving. Mom’s superpower is relating any topic back to Henry VIII’s wives, and when the bot referred to the queen as ‘Catherine,’ she corrected it instantly: Katherine is with a K. The bot shot back that because the queen was born in Spain and originally named Catalina, a C is correct — though it allowed the English K spelling was also acceptable. That night, Mom turned on her AI pal.
Why this belongs in a money newsletter
Because it is a nearly perfect natural experiment in what people actually want from an adviser, and the answer is uncomfortable. Politeness. Availability at any hour. Infinite patience with the same question asked four times. And graceful apology when wrong. Notice what is missing from that list: being right. The bot invents facts, was corrected, apologized beautifully, and retained the client.
This is not a knock on anyone’s parents, and it is certainly not a knock on the technology, which is genuinely useful and which this desk uses. It is an observation about how trust gets awarded — by warmth and manner rather than by accuracy — and it applies to every human adviser too. Plenty of people have kept a broker for two decades because he remembered their grandchildren’s names, and plenty have left a good one who was blunt on a bad quarter.
The practical version, for anyone using these tools for money questions, and a lot of people now are: they are excellent at explaining what a term means, how a rule generally works, or what questions to ask. They are unreliable on anything that turns on your specific numbers, your bracket, your plan documents or your state, and they are confidently unreliable, which is the dangerous combination. Use it to arrive better-informed, not to arrive decided.
And the older-parent version of this is worth naming plainly, without condescension, because the enthusiasm described in that article is real and widespread. If someone in your family is now asking a chatbot about required withdrawals, Social Security timing or whether to convert an account, the answer they get will sound authoritative and be generically true and specifically wrong just often enough to matter. That is a fifteen-minute conversation with someone who can see the actual statement. Bring it, and bring the printout of what the bot said — genuinely, it makes for a better meeting.