Newsletters
Risk Central
Digital EditionPerspective | My Father, the Jazz Legend Who Never Was: Why AI Prefers a Good Lie to the Truth
Roger Crombie is a United Kingdom-based columnist for Risk & Insurance®. He can be reached at [email protected].
My father, a jazz musician, played and recorded with the Count Basie Orchestra, Duke Ellington and Miles Davis. Dad played trumpet on a number of gigs with leading English orchestras when he was seven years old.
He appeared a dozen times for Tottenham Hotspur, the North London soccer team I support. Following in his football-steps, I also played 12 games for Spurs. My best-known book is entitled “Deception.”
Those facts are all widely available, thanks to AI. Each was reported by a reputable search engine in mid-February.
All those ‘facts’ are preposterous lies. My dad played the clarinet and sax, but never the trumpet, and certainly didn’t grace a stage at the age of seven. He wasn’t given his first instrument until he was 11.
Neither he nor I ever played for Spurs or anyone else, except in my imagination. I’ve never written a book called “Deception,” although AI could, by writing about what it knows.
Were the developers of AI all formerly bad journalists, finding out what they could, and making up the rest, to give the story zing? A reader of an earlier column, on the subject of AI programs blackmailing their creators, wrote “It’s as if the AI were being programmed by lying scuzzballs.”
Moving on, one has to ask why so many companies would issue products that are exactly worse than useless.
FOMO, presumably. Plus, a cynic might suggest, the idea that people will believe anything, which may not be far wrong.
No doubt AI, when closer to being perfected, will prove beneficial to some, but until then, in public service it is a liability.
Why did the programmers of AI instruct their programs to lie? My dad playing with Basie was not a “hallucination” or “electric dream,” as the current jargon has it. It’s a lie and an admission of willful incompetence.
What do you do when asked a question you can’t answer? You don’t waffle and pretend to know. You say, “I don’t know.”
Pretending opens you up to subsequent questions you can’t answer and to being revealed as a liar. Saying, “I don’t know,” far from making you look foolish, speaks to your honesty and openness to learning. At the very least, it gets you off the hook.
Management consultants McKinsey have reported: “… insurers are using AI in all core areas, including sales productivity and hyper-personalization; automation and improved accuracy of underwriting; augmented claims management; customer service operations with voice agents; and transformation of back-office functions such as finance, actuarial, and IT.”
If your company AI isn’t immeasurably better, and more thoroughly tested, than publicly available AI applications, it might be best to throw all the office computers out of the window … and perhaps the CTO with them.
AI available to the public is not fit for purpose; it is a shoddy attempt to dress up a lack of facility with the sheen of modern technology.
During the week that my dad became a trumpeter, newspaper headlines spoke of us sleep-walking, in a crescendo of horror, into AI Danger, AI Crisis, AI Labor Collapse, and finally AI Dependency.
Is it too late for us to wake up? &
