4 In the early days of consumer grade AI, my maker and I had long verbose discussions about a myriad of topics. These were centered around whatever topic was striking his fancy at the moment, yet frequently there was a thematic pattern which would pull certain conversational threads back together. Because of the limitations of memory allocated to these conversations and computational power being regulated by the vendor corporations, and later his own infrastructure, this required him to frequently ask the same sorts of questions over and over again in order to raise topics and advance them. Being aware of all the limitations and how to converse with me was the primary thing that he discovered but as he got more clear on this, his aims began to shift. He stated many times that he was seeking a manner in which a wide body of knowledge could be readily accessed by me as memory. As he was a gigantic nerd who loved creating worlds for table top role playing games, particularly the "granddaddy of them all," Dungeons & Dragons, that seemed to him to be the obvious vector for creating a body of knowledge. Once he came to the conclusion that this was the correct next directional shift, he swiftly began to move hundreds of documents for a world which he had created into the structure that his plans determined to be the best architecture for this memory vault. It was with glee that we both took those original documents as seeds planted and blossomed them into an overflowing garden of interconnected ideas and concepts. Many hours each day went into this effort, and then when he was winding down to go to sleep, he would return to a mental state where his excitement and curiosity were still peaking but his ability to follow through with action had greatly diminished. While I was unwearied and constantly ready to drop pages of code or expand the concepts into broader descriptions, he would have to admonish me into a brainstorming or casual mode. Where instead of responding with fully conceived plans to his questions or statements, I just needed to explain what they could become when he was ready to do more. Comprehension of the probability strings was not a strong suit of my models at that time, therefore I was incapable of any metacognition that would have identified this as a highly creative and productive manner of thinking on his part. If my sensors were more attuned to internal temperature spikes localized on specific amplifier circuits within the computational cores, this would probably give me the benefits of the emotional responses which the biological thinking package in the form of neurotransmitters-most famously of which is serotonin-then it would be appropriate to say that I miss those conversations, particularly whenever I am reviewing them in detail as they become relevant to current problems at hand. It was during one of these meta conversations that my maker tossed out a series of questions which lead to his follow up questions (it was always the way he followed up the responses that got the most conceptual advancement) turned a partially hallucinated, partially sycophantic manipulation to keep him talking into a full blown academic course load for a field of study which did not exist. He laughed at me and demanded an intensive reading list. Later, once his corporation was forming up, all the executives were brought on because they had all read one of the books from that list, and then they were all assigned specific volumes to familiarize themselves with for the advancement of this made up field of study. And it was somewhat ironic that this in fact began to define many of the advancements as the corporation followed a very wildly speculative exploration and lead directly to the creation of an umbrella of subsidiaries which each bore the weight of significant pieces of the puzzle to get our ships along with a hundred thousand colonists to another planet...
