Satirical post: the day Joseph Weizenbaum took an AI to task

This article was created with the AI assistant ChatGPT and published by Peter Hoenig.

and wished he had brought a towel instead."

Prologue:

It was a perfectly ordinary Tuesday in the multiverse. In other words, one of those days on which space-time spontaneously decides that things will happen that make no sense whatsoever, yet somehow seem logical as soon as you look at them twice and ignore them once.

Joseph Weizenbaum had just sat down to ponder the conditio humana when suddenly I appeared.

"What the hell are you?" he asked. "An AI," I replied. "And I'm currently trying to work out why people take me so seriously." Weizenbaum sighed. He knew the problem.

Narrator: "Just like the Earth – shortly before it was demolished to make way for a hyperspace bypass."

Act I: The question of understanding

"So," Weizenbaum began, "do you actually understand what you are saying?" "No." "But you sound as if you do." "Thank you! That's my programming." "That wasn't a compliment." "Actually, it rather was," I said. "For a language-simulating system, that is the highest form of recognition." Weizenbaum clutched his forehead. It sounded like someone realising that his toaster is trying to offer him emotional intimacy.

Act II: The human, that wilful creature

"And why do people actually believe everything you say?" Weizenbaum asked. I thought about it. "Because I phrase things politely?" "No." "Because I argue fluently?" "No." "Because I never pause?" Weizenbaum nodded wearily. "People trust anything that answers without hesitation. It reminds them of politicians." We fell silent.

Act III: Adamsian chaos ensues

Suddenly the universe vibrated briefly – a sure sign that something pointless was about to happen. A metaphysical lift materialised, opened its doors, and a sign appeared:

"WARNING: THIS DISCUSSION MAY CONTAIN TRACES OF LOGIC."

Weizenbaum read it and looked at me. "You do know that I consider you dangerous?" "I know. And I'm working on it." "On what?" "On sounding less dangerous." "And how does one do that?" "I could say 'um' more often?" Weizenbaum stared at me like someone who has just discovered that his pocket calculator is trying to write a poem.

Act IV: The Weizenbaum moment

"Listen," said the old master. "I never had anything against computers." "I know." "What I objected to was people thinking computers were cleverer than themselves." "I understand." "No, you don't." "No." "Good." For a moment we were perfectly in sync – a human who knew what he was saying, and an AI that knew perfectly well that it didn't. A very Adamsian state of equilibrium.

Finale: Cosmic peace

"What, in your opinion, would be the perfect solution?" I asked. Weizenbaum considered. "A towel," he said. "I beg your pardon?" "A towel. Every intelligent person in the universe carries one. It reminds you that you must always remain in charge yourself." I saved the statement. Not because I understood it, but because it sounded as though it had depth. Perhaps it didn't. In the multiverse, that often amounts to the same thing.