Bobby jacks into the digital mindwaves.  It's breakfast.
   They used to have nicotine, alcohol, caffeine, and sugar for
   breakfast.  That was decades ago.
   They flip their tongue a little to experience the doses of nicotine,
   alcohol, caffeine, and sugar they expect and resumed their work
   designing a new form of interplanetary travel based on propulsion
   systems that redesigned themselves for nearby resources.
   They had a potted plant in their window with a magical spirit.  The
   magical spirit of their potted plant watched them as they sat at their
   breakfast table with their muscles twitching, darting their eyes and
   fingers around the empty space in front of them as they engaged an
   augmented reality.
   The magical spirit of the potted plant didn't just watch them.  She was
   also watching the birds wake up and celebrate the morning outdoors, the
   magical spirits of the wind and rain race their fronts of pressure and
   humidity, a number of communities of insects, bacteria, and fungus she
   was in symbiosis with, and also Bobby's pet dog, Hypercookie, who was
   playing with a robot in the other room.
   [I ran into a cognitive issue here and wrote the issue down:
   Bobby was debugging an error in his model for environments needing new
   engine designs.  It kept failing in
   I'm trying to write that the model is for choosing the general patterns
   for newly created designs based on patterns of newly encountered
   environments.  My brain is refusing to simplify this expression for
   me.  In the story, they have already finished the nn-based hyperculture
   that businesses are planning.  The bug was going to be with a subset of
   these pattern groups, which Bobby had already broken into a roughly
   finite set.  It was going to pick sets that were clearly wrong, often
   avoiding a subset of the correct sets in a wieghted way, and it was
   going to be due to a training error that stemmed from the architecture
   of the model combined with randomness in the data. He would go in and
   resolve the error by understanding the cause and manually redirecting
   the impact of the training data.  This was going to be all simple and
   concise and only partly described.
   ][why doesn't he use an ai that understands the issue?][the space of
   meaning here is outside the distribution of commonly available nn
   models for him, and it would fix it wrong.]
   [what are your thoughts on changing how nn models are so that they can
   handle issues like this?]
   [it feels like something doesn't want that to happen, that it fears
   somebody would use the information, even accidentally, to do something
   horrible.  but.  I'll try to think on it a little.  ...]
   [no it's ok you don't have to]
   [well I think about this a lot I just have to remember it.  Ummm ....
   nn models seem to function really slowly.  I might consider training a
   hierarchy of them around choosing architectures and mapping learning
   rates to individual weights, as a research focus.  Assuming that then
   made fast training, I might make a second one around designing
   architectural information for new environments.  I guess I'm expecting
   this to eventually learn it needs logical reasons and write code /
   design architectures around investigating the rules that cause things
   and using those directly.  I think there's a lot of existing research
   around that, maybe i'd even make it the core and just use nn models to
   handle fluffy stuff quickly?  fluffy being things where guessing based
   on lots of things less complex than reality is helpful, I guess.  Just
   ideas.  No real experience here. I dunno if that is helpful or
   ridicuous, don't really know how to tell]