Explaining AI to a Nonagenerarian
My parents-in-law, Bill and Grace, are both 91-years-old. They ran a successful real estate brokerage for decades and retired in the mid-1990s.
The fax machine was the greatest technology innovation of Bill’s working years. The ability to send documents between banks, title companies, and other realtors without the expense of FedEx or the delays of the US Postal Service utterly changed the game. That beeping machine in the printer room expedited closings and expanded their market. Eventually, of course, they added email and few PCs to the office, but those technologies couldn’t match the wonder of the fax machine.
The trusty dusty fax machine was a game-changer for a small business owner in the 1980s.
Although they had cell phones toward the end of their careers, the devices were still simple phones. A smart phone loaded with apps and voice recognition would have made Bill even more frenetic in business. The hustle never stopped with Bill.
Not only has Bill never sent an email, he has never had an email account. Starting in their business years and continuing to today, Grace handles emails. Bill is a phone guy. When some young tele-sales rep calls him with the latest investment opportunity, he will keep the poor lad on the phone for hours. Bill never buys anything, but he learns a lot at the expense of the erstwhile rep’s sales targets.
Likewise, Bill has never opened a web browser or Googled anything. Grace has a laptop and a tablet, but not Bill. She provides him the information he can’t extract from the tele-sales call centers scattered about the globe.
So when Bill asked me, “What is AI?” I took a long pause.
I considered explaining that AI is a giant hype machine, and if anyone calls him with an offer to invest in it, he should keep his money in his pocket and hang up immediately..
I thought about launching into a rant about AI being built on the pirated intellectual property of artists, writers, philosophers, and content creators. How Sam Altman whined that if they had to pay for what they’re loading into their large language models, the business case would collapse. But society’s inability to compensate the creative class is not a topic Bill has ever worried about. At 91, he’s not going to start now.
A discussion about the AI bubble is more up his alley. But that particular rant would inevitably sidetrack into my disdain for our nation’s veneration of tech billionaires. He would surely lose interest as I raged on about that ridiculous cover of Time Magazine’s “2025 Person of the Year.”
I definitely wasn’t going to explain that when a company like Oracle announces that they’re investing $450 Billion on AI chips from a single manufacturer, delivering that volume of chips to data centers that are not yet built almost guarantees that the chips will be obsolete by the time the data center is online. And don’t get me started on the pending collapse of the data center build out craze.
I decided to describe large language models (LLMs), and how prompting an LLM is different from a Google search. I explained that Google merely retrieves existing content, while an LLM can synthesize content from multiple sources and deliver a response in plain English.
There was only one problem: Bill doesn’t Google.
We went back to talking about football.