Abundance features an AI entity in the story. BrainChild (BC for short) plays a pivotal supporting role and explores what the development of artificial general intelligence (AGI) might look like.
When we laid the foundations for the story, we needed to define where we thought the field of AI might be in the near future. This entailed making educated guesses as to when the major breakthroughs in the field might occur, and how it might affect the world as we currently knew it. Some of the notes on this document (written in 2019/2020) hits a little too close to home these days:
— An explosion in AI technology (narrow AI, or task-specific AI) renders a large percentage of workforce unemployed, causing a global recession and general slowdown. (We can use this to explain any tech or situation in our story that seems like it should have progressed further in 20 years) Note that the global recession hits China harder because of the size of their manufacturing workforce.
— AI use is widespread but AGI (Artificial General Intelligence) is still not quite there yet. Just as AI pioneer Herbert Simon wrote in 1965 that “[general AI] will be capable, within twenty years, of doing any work a man can do”, the problem of general intelligence has been understated. By our story’s timeline, weak AI use is widespread, and conversational AI is possible, but AGI is still in development. Jen Mathers is in this field when we meet her, as part of a team of researchers working on AGI.
— [Deleted because spoilers]
AGI has always been the holy grail of artificial intelligence research. Before the capitalistic explosions of the last few years stretched the taxonomies of the field, what the general public has access to these days were classified as Reactive Machines, or Weak/Narrow AI – models designed to excel at certain tasks (conversing with humans, for instance). Training these models with ever-increasing datasets gives us something that seems to understand what it is trying to solve, but crucially, does not really grok the answer, nor is it self-aware.
We posit that the roots of AGI will take shape once machines are allowed to learn on their own (from external ‘sensors’ that mimic our five senses) and to be able to self-author code. Our AGI entity in Abundance possesses some of these additional computational ‘modules’, which we have named evocatively, as follows:
Exp Stack
Empathy Framework
Inspiritos Loop
Logic Leap
Self-Authoring Stack
How do they work? Who knows, but they do. The beauty of being a futurist writer is that I can simply assert that these are the key ingredients for AGI, and leave it up to people smarter than me to sort out the details…
Mahalobay 
Leave a comment