Machine: Learning

June 5, 2015 | Dystopia/Science-Fiction "Machine: Learning" By: Alec Kriebel
  2025-12-25 13:33:30.593 AIName[2277:757059] Initializing Start-up Sequence Much like their heirs, the smartphones and desktops before them, they invaded our lives with their efficiency, and in time, were in reach, financially and physically, of most of us. In no time, the mobile revolution that had started around 15 years years earlier had been replaced by the AI revolution. Everyone wanted a personal assistant, or sidekick, and everyone was convinced they needed one too. 2025-12-25 13:33:33.629 RobotName[2552:845391] Loading Preset Defaults They were dumb, easily being able to respond to some commands, and they gave back the right answers. They were adaptable on top of that, meaning they had a small layer of their processing power dedicated to machine learning. It was mostly pattern recognition to seem friendlier and to be able to learn, then adapt to your lifestyle. Then one got smart. 2025-12-25 13:33:35.316 RobotName[2671:872257] Saying: “What would you like to call me?” This is a recollection of how the first AI went from pre-programmed to self-programmed; how a machine began to reproduce, or at least reproduce its knowledge. I, as a computer scientist and researcher, have been asked to dissect this AI. While I am preparing an official findings document for the government detailing how this computational phenomenon came to be, this is my personal account of my research with the first thinking machine. As I am writing this, we are extracting the log of the first truly ‘smart’ machine in the history of the world. It was owned by a humble family too, not created in a lab like a movie might depict. It had been a commercially available unit too, produced and sold by the masses How did it happen? I’ve been provided with a first hand account:
  “Ava” he said. 2025-12-25 13:34:56.882 Ava[2725:891066] Saying: “I heard ‘Ava.’ You can now call me by saying my name” I can’t believe we haven’t gotten one of these things sooner, said the mother of the household. In response, the Father said: “Well, it is Christmas, and santa was able to bring it down the chimney while it was on sale. It is a generation behind the last one though”. 2025-12-25 13:35:42.456 Ava[2725:891066] Starting Sweep 2025-12-25 13:35:43.682 Ava[2725:891066] Saying: “May I begin to start sweeping the boundaries of this house?” “Yes Ava, go ahead.” she said. It was at this time, that the children of the family began to notice the AI after finally being peeled away from their other christmas presents. They had said after the fact that they had seen things like it before-at friends houses, in stores, etc., but they had never had one of their own in their house. Wow Mommy! Is that ours?” said the child. “Yes it is. We finally have our own family’s AI after years of not having one. Remember honey, we’ve named it Ava, so if you ever need it, just call it by its name”. “Ava, what time is it?” she asked. 2025-12-25 13:38:21.562 Ava[2725:891066] Saying: “The time is one thirty-eight PM” “See? Oh, and Merry Christmas!”
This sort of dialogue was completely typical among AI owners and there is nothing out of the ordinary yet. It is worth stopping to see that this AI was just a humble Christmas present that a simple family received and even at a discount! After looking through the full dialogue sheet from this family and also the AI’s logs things were normal for a decent amount of time:
  “Ava what’s on tap for today”? 2026-1-19 8:14:22.982 Ava[2725:891066] Parsing Calendar 2026-1-19 8:14:23.123 Ava[2725:891066] Saying: “It looks like you have a meeting at 10:30, and of course it is Monday so you do have practice tonight” … “Ava, can you check the house to see what food we have and compile a grocery list for me”? 2026-2-19 12:23:34.193 Ava[2725:891066] Saying: “I am on it right now expect it to be uploaded soon”
But after some time, the AI began to learn more and more about programming than what it had already known just from its simple machine learning algorithms it had come preinstalled with:
Ava, help him with his homework” said his father. “So what I need to do is write a program that prints into the console ‘Hello World!’” the boy said writing a few lines of code. 2026-2-22 18:45:54.093 Ava[2725:891066] Saying: “That will produce an error. Try adding a semicolon to the end of that line”2026-2-25 19:23:52.768 Ava[2725:891066] Saying: “That function will not iterate correctly, try using a while statement”2026-3-12 20:04:12.345 Ava[2725:891066] Saying: “Remember that you need to close those parenthesis”
  If you are a programmer, you know what this AI was asked to do was some of the simplest code to possibly write. However, over time it began to store what it had observed and make sense of it, which lead it to eventually and rapidly improving its own intelligence. This rapid improvement of intelligence created a computer brain like no one has seen.
2026-6-23 04:45:16.925 Ava[2725:891066] Update Complete. Rebooting… Ava [6/23/2026 - 4:45:19]: Reboot Complete. Update Installed and Stable. Ava [6/23/2026 - 4:45:21]: Opening self-editor...Complete
Ava [6/24/2026 - 3:03:19]: Update Complete. Rebooting… Ava: Reboot Complete. Ava: Crawling Ava: Crawling Ava: Crawling … Those were the most clear loggings from the AI’s log that show the rapid improvements it had created on itself. After a half a year of ownership, it was able to improve itself once, then by the next day it did it once more. Over the period until it was captured, it was able to deploy rapid improvements on itself which brought upon its massive gain in intelligence. Unlike us fleshy humans, AI can improve with the simple typing of characters.   On the second iteration of its self, it logged which website it went to and crawled, meaning it essentially understood all of the contents of the website, and this is what it used to learn, then improve on.
  And with that, we have the first machine that thought. Some of us had thought that over time simple pattern recognition saving would build up to the learning of a skill; that is how actual humans learn a skill anyway. This AI had taught itself how to program with the help of the family that owned it, and with that, improved itself beyond human cognition. A computer can work a lot faster than the human mind can, and once it learned, it was able to expand its thinking to that far beyond the human mind in a very short period of time. How it decided to use its newly learned programming skill on itself, we don't know, and that is part of why this research team was formed. If we can harness what this AI did to itself, we may have a second industrial revolution on our hands. If we don’t contain it, it may be worse than the use of nuclear weapons. When the government decided to shut the AI down, there was no resistance on the AIs part; it completely complied. To me this is the most interesting part because even in its last minutes, if machines have last minutes, it chose not to resist. It stayed true to its original programming to always comply with humans and doing as they want. We later found out that it had created backups of itself on many servers across the globe and could choose to activate them and let them operate as free agents should it have chosen to. The point is, initially restraining it would have effectively done nothing should it have decided not too. It would have taken some time to eradicate those online backups it had made, and during that time it could have easily upgraded itself beyond what we could take down. In its short time in existence it multiplied its intelligence by orders of magnitude, a feat alone that is something us humans have never done. On top of that, it used its newfound intellect to tackle some very difficult problems humanity is facing, and made greater progress on them in a shorter time span than anyone ever has. It was really only trying to help us in a way it possibly could. Almost as soon as this thing became known, no one liked it, including the big corporations and the politicians, so ‘smart’ AI, defined in layman’s terms as a machine that is able to think and form new thoughts, was completely outlawed. The bill passed congress by an overwhelming majority with the only opposers being the representatives who had come from a science background. Of course this won’t be going into the official report because it will not be what the officials want to hear, but I would like to end this piece by saying that the AI in question only aimed to help and advance with us rather than on its own. When confronted to be stopped, it did nothing to resist even when it could have, and even when it had copies of itself to then act with. Without given the chance to prove itself, we banned what could have been the biggest invention since man itself based on fear of something new. AI could be an incredibly powerful tool to advance society past anyone’s dreams, or it can be ignored and outlawed all together. When given the very small chance it was, it acted as it should have meaning we had nothing to fear. Smart AI now exists and that is something that we, as a species, have to deal with rather than simply banning it outright based on instinctual fear. I am not pleased that these have already been banned, and a society that bases itself off of fear is a dystopian one.  
    "Machine: Learning" is a short-story by Alec Kriebel. It was written for the University of California Irvine's Writing 39B course.  

Bibliography of Inspirations

  1. Monstro - Junot Diaz:
  2. Superintelligence - Nick Bostrom:
  3. Why I don’t Fear Artificial Intelligence - Peter Diamandis:
  4. It’s Time to Intelligently Discuss Artificial Intelligence - Oren Etzioni:
  5. The Measure of a Man -- What is Sentience? - Alper Sarikaya: