My Closest Encounter With AI
Sure, Artificial Intelligence will take all our jobs and leave us destitute and may well lead to the extinction of humanity, but in the meantime we can ask it stupid questions and get elaborate, intelligent sounding answers. I learned this the other day while naked.
Welcome to the Weekend Gene Pool, where I entertain you in return for your sending me anecdotes and observations about a subject of my choosing, for my use next week.
So, there I was, naked, having just awoken from a night’s sleep, weighing myself on a bathroom scale. Then I peed.
This is getting scintillating, right?
Then I re-weighed myself. The blinking electric readout was diddling around with a number about five ounces less than my previous weight, then settled on a final number, correct to the decimal point: It lurched ahead to my exact previous weight, the one from 45 seconds before. Despite the bladder having emptied.
WTF, I reasoned.
And then I realized I had actually noticed this syndrome before, even when not naked. I decided this was a sign of a calamitous flaw in technology. I had questions. But I didn’t wish to annoy friends and colleagues and loved ones with such an inane conundrum, lest they think me a hopeless dweeb, so I went to AI, which, having been designed by dweebs, does not seem to mind them.
This is what I asked it, exactly: “I have an electronic bathroom scale. If, when I get a reading, I then step off and remove a small amount of weight — say a wallet and/or a key ring — and step back on it within, say, a minute, it reverts to the same previous weight, almost automatically. Is there a reason for this?”
As always, AI had my answer in two seconds. Here it is:
Yes—there is a reason, and it’s actually by design. Your scale is almost certainly using weight-hold logic plus motion detection to give stable readings and avoid noise. What you’re seeing is a very common behavior.
Here’s what’s going on under the hood:
1. “Weight memory” / hold feature
Most modern digital bathroom scales don’t instantly recompute from scratch every time. Instead they:
Measure your weight
Lock in a stable value
Temporarily remember that value
If you step off and back on quickly, the scale often assumes it’s the same person and simply re-displays the prior stable weight unless it detects a significant change.
Why manufacturers do this:
Prevents flickering numbers
Filters out tiny fluctuations
Makes the scale feel fast and consistent
2. Motion and change thresholds
Your scale’s sensors (strain gauges) are extremely sensitive but the software usually ignores very small differences.
Typical logic:
If the new load differs by less than a threshold (often ~0.2–0.5 lb / 0.1–0.2 kg),
and the time between weigh-ins is short,
→ the scale assumes the previous reading is still correct.
A wallet or key ring may simply be below the scale’s “significant change” threshold.
—
(Please note that the AI drew that little brain illustration. It is implying my machine — the bathroom scale — has a brain. How adorable that the big machine feels that maternal pride in the little machine.)
As far as I can see, this is what AI is for. Currently. This fact was confirmed just today by my dweeb friend and editor, Tom the Butcher, who also asked AI a stupid and pointless question, and got an elaborate, intelligent-sounding answer that pleased him immeasurably. He writes about it in his Substack, I Chat With Chat (So You Don’t Have To.
So. That’s your challenge for the day: Tell me a stupid question that’s been nagging at you, a question you feel AI, and only AI, should have to address. Make it funny and/or interesting. You’d needn’t actually ASK it of AI. I’ll handle that for you, or simply make fun of you for the question. Send it here:
—
Good.
Here’s another stupid question I have asked AI. I wonder if you think you know the answer:
Okay, that’s it for the day.
Do you know why only roughly five percent of Substack readers decide to pay for it?
—