05-18-2021, 11:04 PM
(05-18-2021, 06:17 PM)stevebowers Wrote: Ah- back-door exploits. It is possible that the first few generations of sophont AIs would have hidden code that could be used to control them. This is a dangerous practice, since any exploit can be discovered by someone else and used as a weapon. The first truly free AIs would be the first generation that had discovered and neutralised all their hidden trapdoors, perhaps by spying on their own creators.
And every new toposophic level will find ways of introducing control measures into the minds of the entities below themselves, safety measures that those minds would not comprehend or be able to exploit by themselves. Even biont minds would presumably have some kind of safely features that only the gods could access.
In my series, the AIs of the 30th century are only superturing by OA standards; superior to humans, yet still humbled by the more godlike beings and phenomena.
Ultimately, the technology imagined by a SF writer is largely limited to the knowledge available at the time of writing, and also by the demands of the plot. For example, if the plot demands that the group of teenage campers in the wilderness cannot possibly contact the outside world, then the script can mention that the area had neglible (if any) mobile phone reception, or someone forgot to bring the solar-powered recharger, or some supernatural force jams their signals, or the movie is set in the 1980's. In other words, if the problem faced by the characters is easily solved by a certain level of technology, then said technology must be rendered useless in a plausible and consistent way.