Humanize This

“It appears that your paper is 81% written by AI,” wrote the professor’s AI to Robert. “As you should be aware from the syllabus, there is a 29% limit of AI usage in any term papers. You may resubmit a revision of your paper within 48 hours, or accept a failing grade. A failing grade may result in the loss of your scholarship.”

“But I didn’t use AI to write the paper,” Robert insisted to his roommate. “I wrote the entire thing myself. I didn’t even spellcheck it.”

“If you’re too technically proficient, you can get a false positive,” Dom said. “I’ve heard about that happening. Just petition the professor.”

Robert wrote a note to the professor, challenging the failing grade, and he immediately received back: “It appears your note was written by an AI. Professor Fallwell does not respond to AI-written missives.”

“I didn’t use AI to write the note about not using AI to write a paper,” Robert fumed. “How do you get past these clucking robots?”

“Did you try to ‘humanize’ your text?”

“What are you talking about?”

“There are multiple programs available. You put your AI-written paper through it, and it will humanize the text for you.”

“A robot will make my paper human?”

“That’s not what it’s intended for. But in your case, I think you might need to.”

Robert processed the information.

His term paper was on the symbiotic relationships that mimicked human relationships. He’d included references to the new wave of human/robot marriages and the fact that there were some electronic devices that were indecipherable from their human counterparts.

Confused by why the AI was sure it was AI-written, he reread the opening of his paper:

Do you remember a time before robots? Most modern people don’t. We have integrated with the robots to such a degree that the pre-robot era feels as musty and archaic as a museum exhibit of dinosaur skeletons.

However, back in the mid-80s, the impending “robot revolution” was touted as one that would save time for humans. This was a conscientious effort done in a way to assure human beings that they were not replaceable.

Humans would program robots with jobs that “breathers” believed were beneath them. Robots would free up humans for more creative endeavors. The cheerleaders for the robot revolution promised the humans that robots were after “tasks,” not jobs. They pointed out that ATMs had not replaced bank tellers. Although, in a deeper dive, that was not entirely true. ATMs did replace many bank tellers, and ultimately many of the remaining bank tellers were replaced by robot look-a-likes.

Unfortunately—or fortunately to the very wealthy—ultimately what commerce looked for was robots who could take on 30% of human tasks. And as soon as that leap was made, three humans in a business would be let go, and a fourth human would be given an extra 10% of work.

In spite of the promises made that the humans would be focusing on creativity—often described as a time to learn to play the cello—it turned out that nobody in power really wanted creative humans. Sadder than that, most humans, studies have shown, are not inclined to be particularly creative. The idea that time equals creativity is one that was almost immediately disproven by Jensen, Frederick, et al. Also, few people harbor a hidden desire to learn the cello.

What did happen was that robots were taught to play the cello! And run marathons. And paint pictures.

With time, some of the robot engineers began creating robots not to take over jobs in the workplace, but to serve more as companions to humans whose jobs had been replaced. These robots were humanoid, the expensive ones were passable, and they could be seen out and about in the real world as much as in the cubicles.

These robots began “life” as blanks and were given personalities by their owners. They were programmed to think they were human, and nothing could dissuade them from their beliefs. They ate food. Exercised. Had “thoughts.”

To Robert, the essay read as human. It was colloquial. It started with a bang. He didn’t feel that it dragged or was too reliant on mind-numbing statistics.

Finally, he typed a note. He wrote that he, himself, Robert A. Winston, Jr. had written the term paper on the sociological study of humans’ inexplicable fixation with robotic interactions. His paper covered all aspects of robot/human relationships. He’d gone deeper than the simple cyclical relationship in which students were currently using AI to write papers for professors who used AI to grade papers, so that the students learned nothing and the professors did no work.

Robert waited the length of a breath, and then he pressed the button to check to see if a robot thought a robot had written the note he’d just written.

“This appears to be 99% AI-written” was the instant reply.

He paced in his small bedroom. He turned off his light and lay down on his twin bed. The glow in the room came entirely from his computer screen. He lit a joint and blew perfectly executed smoke rings to the ceiling. He had been taught by Zach, the older brother of a friend of Dom’s. He’d spent what had felt like a hundred hours trying to make the rings perfect.

He got up and re-read his note.

“Humanize,” responded Robert.

In seconds, the computer suggested a revise. Now was the big question. Could a robot humanize the work well enough to fool another robot?

He began to rev up then, to get increasingly hot.

The purpose of the humanoid robots wasn’t what the original creators of the robot revolution had envisioned. These bots were available for lonely people. They sometimes had jobs that were beneath what they were designed for. Like at an ice cream parlor. Or taking tickets at a movie theater. They went to college. They texted their girlfriends or boyfriends.

But even the best-designed bots could glitch out.

Dom smelled the smoke from where he was drinking a beer on the balcony. At first, he thought it was just Robert getting stoned again. But when he caught a whiff of the melting plastic beneath the marijuana, he came into Robert’s room.

And unplugged him.