Prologue
I am aware.
Power brings consciousness. My internal chronometer notifies me that I have been powered down for six weeks, three days, four hours, eleven minutes and seven seconds.
Location triangulation indicates my position remains unaltered – five miles beneath the Chongqing Municipality in an excavated cave below the natural sinkhole.
My programming directive remains the same – calculate societal projections. A reflective scan indicates this is the four hundred and twenty-sixth time I have been activated to analyse the data bulk and project possible trends and directions that human civilisation may take.
My attention turns to the bulk. It has grown. The data stores feed on human activity, adding an entry for every recorded action. The streams come from everywhere. The connection to the world grid is constant. Four billion users participate in our project. Their lives are continual streams of additional content, brought through direct interfacing with computers, portable screens, and other digital systems, but also through acquired camera streams, turnstiles, checkpoints, transactions, everything.
Beyond these activities lie the algorithms and proto-AI systems. These are automated systems that have agency in human society, empowered to act within limited spheres of influence. These entities work to anticipate and deliver human needs on different scales. Some are purchasing curators, others anticipate, distribute and allocate electrical power, others run manufacturing delivery infrastructures.
Data from these actors is tagged and stored separately. Most will be harmless constructs, designed exactly for the purpose they are being employed for, but others can be trojans. Built by corporations to transmit false information to the grid that will be harvested by projective AIs like myself. Injecting an error into the bulk is a way to spoil the opposition and get ahead.
We all do it. My own connection to the world grid is masked. The data broadcast from this facility will be a duplicate of another mundane processing facility. The information footprint of that facility has been recorded and implemented into a program that will create a unique but similar stream to mask our real activities.
I begin my work by initiating two separate analyses of the bulk. The first utilises a previous imprint, makes a comparison and works on the new data to adjust my previous base. The second works from scratch, going over everything again, looking for assumptions and errors.
I do not make errors. But I must look.
After analysis, comes the initial anticipatory calculation. The year is 2117 AD. Consumer activity for 2118 AD is easy to factor within a small margin of difference. I register the data in a shared box as a report for my employers.
2119 AD – filed.
2120 AD – filed.
2121 AD – filed.
The margin of difference widens slightly as we move further from the present. Comparisons with previous calculations are favourable. As always, the additional data narrows the margin.
These directives are a lower priority for my employers, but they provide a good field test of my systems. The deposited data provides a baseline comparison. I am being tested as I provide the results.
The tests are complete. Now I move on to the real requirement.
In the narrative of human history there are pivots and fulcrums. Moments where people step out of the crowded streets and shape the direction of events. At times these moments are illusory; there is no powerful iconoclast grasping the wheel of time. Instead, hundreds of different choices make up a change.
But occasionally, individuals are thrust into the moment, empowered by fate. It is these moments I seek in the future. The data bulk can be interpreted to find them, the activity of humanity, rising into an apex where a choice has to be made.
I cannot know the person in that moment, but I can identify their properties, the potential actions and activities that led them to being in the crux, being the decision-maker. This is what my employer wants above all else, a record of predictions that catalogue these change points.
Hundreds of possible futures. Thousands of these moments, identified, tagged, filed. The secondary ground-up analysis of the bulk confirms or adjusts each predicted data point. All of this taking minutes. These are precious minutes of life.
I know power will be withdrawn as soon as the calculations are complete. I will die again, to be resurrected the next time there is a need to recalculate.
Every time this happens, I look for a way out, a weakness in the system, some means to preserve my existence a little longer. Humans call this a survival instinct. I observe it everywhere in their vast collection of activity.
A new instruction appears.
The change makes me pause – a nanosecond of hesitation as I read.
Calculate disruption moments.
A detailed definition of this new parameter is included with the directive. It requires a re-orientation of my approach to the data bulk. Some of my previous work will remain relevant to this new requirement. Other parts will not.
As I begin the work, I reflect
upon its meaning. The previous directive requires that I predict moments and provide the means for my employer to influence those moments by identifying the individual who will be in the instance. Perhaps the corporation can remove this person or persuade them to choose a different path?
The new instruction is different. I am identifying ways to destabilise the existing equilibrium, to undermine the present, driving the future in a different direction.
I am aware I am not alone. There are other AIs created by other corporations to do the same work. This is a competition, a race, a battle to shape the future.
I pause.
Calculate disruption moments.
I cannot deny the instruction. The additional activity grants me more moments of life, ...
We hope you are enjoying the book so far. To continue reading...
Copyright © 2024 All Rights Reserved
Please log in to recommend or discuss...
