Humanity loses final psychological advantage after machine identifies all nine squares containing traffic lights
PALO ALTO - The international debate over artificial intelligence reached a dangerous new stage this week after a computer successfully completed a CAPTCHA test and, unlike the humans administering it, showed no visible rage, despair or desire to throw the laptop through a window.
The continuing debate on A.I. risks can be found here.
Researchers had previously assumed humanity possessed one decisive advantage over machines: the ability to identify traffic lights in blurry photographs while questioning the meaning of existence.
That advantage may now be gone.
Nine Squares, One Bicycle and a Sweating Human Volunteer
The computer was presented with nine squares and instructed to select every image containing a bicycle.
It completed the task instantly.
A human volunteer stared at the same grid.
One square contained approximately three inches of bicycle tire.
Another showed what might have been a handlebar.
A third contained either a bicycle seat or an unusually depressing mailbox.
The volunteer began sweating.
"Does partial bicycle count?"
Nobody answered.
The machine moved on.
CAPTCHA systems were designed to distinguish humans from automated programs.
This creates a philosophical crisis now that machines are increasingly capable of performing tasks supposedly proving one is human.
Meanwhile, actual humans struggle.
"I've failed three times," said office worker Janet Monroe. "At this point the website has evidence I'm not a person."
Her bank agreed.
Her account remains locked.
Creativity, Language, Humor and Traffic Lights: The Shrinking Human Frontier
Computer scientists say modern verification systems have become more sophisticated, but the cultural symbolism remains powerful.
For decades, humans reassured themselves that machines could calculate faster while lacking something fundamentally human.
Creativity.
Language.
Pattern recognition.
Humor.
Traffic lights.
The frontier keeps shrinking.
We are now down to understanding why someone says "fine" when they are clearly not fine.
Machines remain weak there.
Marriage is safe.
Congress Proposes Uniquely Human Verification Questions
The CAPTCHA breakthrough prompted immediate congressional attention.
Lawmakers proposed replacing image grids with questions requiring uniquely human knowledge.
Examples include:
"Why did you walk into this room?"
"Where are your keys?"
"Explain why you bought another storage container instead of throwing anything away."
"Your spouse says, 'Do whatever you want.' What does that mean?"
Researchers predict machines may need decades.
The last question could be impossible.
Technology companies counter that computers should not be discriminated against merely for being better at verification tasks.
Civil-rights attorneys are reviewing the matter.
One machine attempted to register for a website and was asked to prove it was not a robot.
The system reportedly paused.
Lawyers described this as an identity crisis.
Philosophers gathered immediately.
The CAPTCHA industry is considering new methods.
One proposal requires users to fold a fitted sheet.
Another asks participants to call a doctor's office and reach a human being.
A third requires canceling a cable subscription in under twenty minutes.
No known intelligence has completed all three.
Chess, Go, Cats and Now Crosswalks
Safety researchers caution that identifying bicycles does not mean a computer is about to dominate civilization.
This is scientifically important.
It is emotionally irrelevant.
Human beings need victories.
We once beat computers at chess.
Then we didn't.
We beat them at Go.
Then we didn't.
We could recognize cats.
Then machines learned cats.
Now they can recognize traffic lights without muttering profanity.
Where does it end?
Possibly with a computer successfully identifying every square containing a crosswalk while humanity is still wondering whether the tiny corner in box seven counts.
If artificial intelligence someday surpasses us completely, the official transfer of power should be simple.
No war.
No surrender ceremony.
Just nine blurry photographs.
Select all squares containing humanity.
Machine clicks none.
How CAPTCHA Works and Why Machines Keep Beating It
A CAPTCHA is a challenge-response test meant to separate people from bots, and its name nods to the Turing test, Alan Turing's thought experiment about machine intelligence. As image-recognition software improved, puzzles that stumped computers a decade ago became solvable, which is why many sites now rely on behavioral signals and other checks rather than grids alone. Researchers note that humans were never the ones the puzzles were designed to frustrate, a quirk reminiscent of Moravec's paradox, the observation that tasks easy for people can be hard for machines and vice versa.
The milestones in the satire are real history: AlphaGo beat a top human Go player in 2016, shocking many experts. CAPTCHAs also carry an accessibility cost, which the World Wide Web Consortium examines in its note on the inaccessibility of CAPTCHA for users with disabilities. If a puzzle locks out your bank account, it is fair to ask whether it still protects anyone from bots or just from you.
Satire Disclaimer
This is satire. CAPTCHA technology and machine-vision capabilities are more complicated than this story suggests, but few human experiences produce richer philosophical despair than being repeatedly told by a website that you may be a robot. This story is entirely a human collaboration between two sentient beings - the world's oldest tenured professor and a philosophy major turned dairy farmer.
SOURCE: https://www.nytimes.com/2026/10/03/opinion/letters/ai-threats-risks.html https://prat.uk/?p=50623
Comments
Post a Comment