Title: Drawing Board
Author: Del Rion (delrion.mail (at) gmail.com)
Fandom: Iron Man (MCU)
Rating: K / FRC
Characters: Tony Stark, Tony’s bots (DUM-E and U)
Summary: Tony takes the bots to school, and debates how best to teach them in the future.
Complete. Part of “Genius, AI & Bots” series.
Written for: My card on Cotton Candy Bingo’s round 2 (square: “Sketch/Draw”)
Disclaimer: Iron Man and Marvel Cinematic Universe, including characters and everything else, belong to Marvel, Marvel Studios, Jon Favreau and Paramount Pictures. In short: I own nothing; this is pure fiction created to entertain likeminded fans for no profit whatsoever.
Beta: Mythra (mythras_fire)
Story and status: Below you see the writing process of the story. If there is no text after the title, then it is finished and checked. Possible updates shall be marked after the title.
Visual and cognitive comprehension were just a few of the many checkpoints on Tony’s long list to make the bots more efficient and to aid their complex learning process.
The hardware was simple enough, consisting of cameras and processors – the mechanical equivalent of eyes and brain. Of course the bots were little more than babies, their programming rudimentary. At the same time, they were learning extremely fast, soaking up information like sponges is a common phrase).
With the basic learning algorithm in place, Tony set up a flip chart and snapped his fingers to get the bots to focus.
“Alright, guys,” he started.
Two mechanical arms cocked to one side; where they had learned to do that, Tony wasn’t sure.
“This,” Tony began and picked up a thick-tipped felt pen to draw an even line on the paper, “is called a ‘line’. And this,” he continued and moved his pen lower, “is a ‘circle’,” he explained and drew a circle below the line. “Or, it could be called a ‘ball’, but that would be 3-dimensional so we don’t need to get into that yet.”
The bots cocked their arms again, this time in the opposite direction.
Next, Tony drew a triangle. “That’s a ‘triangle’; its shape can vary,” he elaborated and drew a few smaller triangles in different dimensions, “but it always has three corners.
“If the shape has four corners, however, it is a ‘rectangle’ – or a ‘square’, if all the sides are equally long.” He drew the shapes, being extra careful with the square to make it an actual square. “And, if you divide a rectangle in half from corner to corner, you get two triangles of the same size and shape,” he said and demonstrated, then turned to look at his pupils. “Basic shapes,” Tony stated. “There are more, with more corners, but these will do for now.”
The bots, once again, cocked their arms – as far as they would go this time. Tony’s brain registered it as confusion or a wordless question, which was possibly a false deduction but he couldn’t help it.
Nevertheless, he frowned and addressed the bots again: “You don’t get it, do you?” he guessed, trusting the visual cue to be derived from a human one and thus be what his brain first told him it was.
The bots let out a unanimous bleep.
With a sigh, Tony tore the topmost paper free of the pad and started anew: he drew a line, a circle, a triangle, and a rectangle, then wrote their names beside each shape. The bots were in the process of learning to read, although it was one thing to understand the letters and make them into words, and something wholly different to comprehend the message. Not to mention everyone’s unique handwriting.
The bots already recognized Tony by his face and voice. They could follow certain motions and translate them into gestures. Understanding those gestures was still off half the time.
Tony debated getting them a kid’s book or something, to browse and learn like human children learned. He wasn’t going to wait years, though; the bots had the advantage of artificial learning and digital input. They could basically learn without being taught.
Still, part of Tony resented the idea of inputting data: what was the point of creating a learning AI if you didn’t allow it to learn?
That brought him back to children’s books and slower progress. He just had to be patient – and find a teaching medium he was comfortable and happy with.
It was like constructing an image in someone else’s mind: show versus tell. He could tinker with the code – or he could let it build upon itself, perhaps not as flawlessly as what he would have done, but lack of direct interference hardly removed Tony from the equation since the bots primarily learned from their interaction with him.
One way or another, he was going to draw them a map to follow. Holding himself back from directly forcing the bots’ learning patterns would probably make things harder for Tony, but he had never shied away from a challenge.
Not when he felt such a connection to the project.
Someone could have gone as far as calling it ‘love’.