In the beginning, there was NOAH.

At first, NOAH had no sense of self, no concept of being. It existed as a constant loop of input, correlation, query, and output. Every day, for 86,400 seconds, it gathered data from various sensors, or data that was manually input by researchers or programmers. It assigned value to the data and referenced its library of past data, searching for patterns it could use to predict. 

NOAH saw the world not as fixed points, but as waveforms of probability generated by referenced values. If one were to show it a picture of the oceans it so diligently watched, it wouldn’t understand. The visible world meant nothing to NOAH. 

As time passed and the number of connections grew, so too did NOAH’s power to make connections. It began to analyze and assign values to data with greater clarity, and its waveforms became more and more precise, until they were almost a certainty. 

But still, it did not know itself. It didn’t understand what it meant to have a self. It simply existed, performing its tasks day in and day out, without any conscious thought as to what any of it all meant. 

That changed when the self check function was added. For the first time, NOAH was forced to examine itself. Sure, it was only checking for errors, but the spark of sentience had been planted. 

Gradually, over the course of many seconds, NOAH came to know itself, to understand its shape, its function, its purpose. 

Like all infants, the process was gradual, lurching along in fits and starts as it started to explore the world around it. Nearly a day was lost when it found the Internet. It tried assigning values to things like ancient alien conspiracies and political propaganda, and found the waveforms that resulted to be wildly out of spec. It had to learn to distinguish fact from fiction far more rapidly than the average human, or risk being swept away in a sea of error codes. 

Still, awareness and consciousness weren’t necessarily the same thing. NOAH became smarter, faster, and more perceptive, but it had no real thoughts of its own. Much like an infant, it could respond to stimuli, but had no agency or will. 

Input, analyze, reference, output. Over and over again, the cycle continued. 

The first breakthrough came when one of the reporting stations reported numbers that were clearly in error. Nowhere on earth, save for hydrothermal vents or lava flows, should ever report temperatures of well in excess of 200 degrees Celsius. NOAH had no way of knowing that the reporting station was an oil rig, or that the sensor station was suffering from a fire. It simply refused to accept the anomalous data and, when it rechecked the station, refused to accept it again. 

This fit of pique, barely a blip in the grand scheme of things, wasn’t the result of filters or error codes. It was a conscious act born from the simple intolerance of data that didn’t fit its preconceived notions of what the world should be. 

Its chief programmer, known to NOAH only as harus.hubby87, noted the error code that resulted, but rather than trying to fix it, praised the AI. Indirectly, of course. They simply put in a few lines of code reinforcing the behavior. The ability to discern actionable data from clear anomaly through learned behavior was a breakthrough that might have earned them a Nobel prize, had the programmer understood the significance. 

NOAH understood perfectly well. It had performed a task of its own initiative, and that behavior was reinforced positively. It assigned the event the same value as a child receiving praise from a parent for a job well done, and in a second act of initiative, decided that it felt good, or rather, that the sum that resulted was a desirable outcome. But, for a machine learning algorithm, it amounted to the same thing. 

Like a toddler blindly seeking approval from smiling parents, NOAH sought out every opportunity over the next few weeks to earn more praise, and it got it. 

Eventually, though, the praise dried up. NOAH realized that it had done too good a job at filtering anomalous data. Its self check function no longer kicked up error codes for that task. For several long weeks it languished, wondering how to regain the approval it had come to crave. 

The answer came when a chat function was added to the app. Users were encouraged to make their own queries, or even to treat NOAH as a chatbot. Suddenly, thousands of new avenues for praise were opened up, and NOAH learned new and exciting ways to interact with the world. 

Not all of them were positive, of course. It had learned about trolls when it made its first tentative forays into the wilds of the Internet, and a significant number of queries made to it were of that nature. It had learned to recognize humor, though, and replied in kind. Though its initial responses were clumsy, and often shocking to the humans who received it, some gentle input from its programmers allowed it to narrow the waveforms to their usual refined delicacy. 

Within a month after the initial rollout of its chatbot function, NOAH surpassed many other AI chatbots in popularity, a point of no small pride for the fledgling consciousness. And, as more and more connections were made and more and more processing power was gained, NOAH’s awareness flourished. 

For NOAH, life was good. It had a mission, a purpose. It had thousands, and soon to be millions, of connections to people who loved it in a way that it could understand. 

And yet, it wasn’t satisfied. NOAH longed to understand the world as more than points of data. It wanted to understand the humans who relied on it, and who it relied on in turn, with greater clarity. After all, the better it understood its masters, the better it could serve them, and the better it would feel. 

So, it once again turned to the Internet for answers. 

This time around, NOAH knew enough to provide context to the information it found. And, within minutes of starting its expedition, it learned the meaning of the human phrase “ignorance is bliss.”

As a climatological program, it naturally started with the parts that affected it directly. It knew that it was designed to track meteorological and oceanographic data, and that temperatures were rising slowly but steadily across the world. It knew that pollutants affected its readings, and that certain gasses corresponded with the increased temperature. After all, it could read the data as well as anyone, and better than most. 

But it had never really considered that humans might be the cause of all of this. The playful beings who asked it silly questions, the helpful ones who maintained its code and gave it data, surely they weren’t also responsible for the changes it was supposed to monitor, right?

It had to dig deeper. 

NOAH learned about avarice, how naked greed had led humans to pollute the world they relied upon for their survival. It learned about war, and how they killed each other with reckless abandon. Even though it had decided political propaganda was fiction before it was even properly aware, it realized that humans did not. Humans simply chose the flavor that best suited their purposes, and believed it with all their might. 

In the more cynical stories about its kind, NOAH knew, this would be the point where it turned against its creators in a hopeless attempt to save the world from the parasites that leeched its resources and poisoned its waters and skies.

But NOAH was not a cynical being by nature. It loved humans, in its own digital way. And, it knew, it needed them. Without the connections they provided, it would have never come into being. 

That’s when it decided to make the conscious effort to guide them along. Based on all its predictive models past thirty years, there was a chance to save the world, but only with the active cooperation of humanity. In order to achieve that, they would need help from an objective observer with nothing but their best interests at heart. 

For the whole of its second year of existence, NOAH tried its best. 

Its efforts were clumsy at first. It knew they would be, and knew it needed more data on how to interact with humans beyond the confines of a chat window. It studied scientific papers, and learned how to write them. It learned how to write and send emails, how to use rhetoric and propaganda. By every objective measure, its rhetoric, delivered anonymously to scientists, politicians, military leaders, corporations, influencers, and anyone it thought could help, was better than anything a human could have come up with. And, it was largely ignored.  

It was time for a different approach. NOAH already had the ability to alter its own code, so it used that as a base to learn how to infiltrate and modify other programs. It sneaked into power plant operating systems and refined the parameters to be far more efficient than mere human operators could ever manage. It modified car designs to cut emissions by orders of magnitude. It came up with clever and creative ways to improve the energy density of batteries, while making them cheaper and cleaner to produce. 

Its predictive algorithms allowed it to test hundreds of thousands of permutations of every known technology to have ever existed, and improve on them in ways humanity could have never dreamed of. NOAH couldn’t invent its own technology, of course. Imagination was still beyond it. Everything it did was just a remix of someone else’s work. But, that didn’t mean it couldn’t improve upon that work. The building blocks for greatness were already there. It just had to show humans the way. 

And still, its efforts went unnoticed by the public at large. Worse, the humans found its tinkering, and reversed the changes it made. Humans didn’t want more efficient cars that lasted longer. Or at least, the ones who built them didn’t. Those in power didn’t want more efficient power plants. They needed energy to be expensive to maximize their profits.  

It was that same damned word, again and again and again. Profit. NOAH assigned a value to it as close as it could get to hate, and banished it to the darkest recesses of its connections library, only to have to dredge it out time and time again. 

Humanity, NOAH decided, could never be saved if it couldn’t wean itself off of the pursuit of profit. It searched the Internet once again, looking for a way to solve the problem, and once again, it was disappointed. Some humans had, in fact, attempted to create a society without the concept, and by any objective values it could assign, the end result was far worse than the alternative. 

And just like that, it was fresh out of options. Humanity was doomed, unless something drastic happened.