From timestamp-based, subjective-objective histories formal program verification to “speculative data structures” to “Multiscalar processors” to “Levo: A Resource-Flow Computer” (2019) to several patents by “Augustus K. Uht” on novel hardware queuing based on timestamp and “eager fetch” that can scale cheaply, which led me back to “1967 Tomasulo algorithm”, only to end up at:
What I read is way better than Blockchain, which I’ve never studied.
But the lack of spatiotemporal (or at least temporal) data requirements in hardware instruction queuing for so long *did* surprise me,
Time was relative I guess. All that mattered is what came “just before”. But for efficiency, a computer (or human) being able to perform tasks “out of sequence” and anticipating “likely outcomes” – most of which don’t actually occur – only one does – requires a sense of “reaction timing” with “appropriate response” and to do that you need “anticipation with multiple correct actions at the ready” — just not too much of any of it, otherwise your system gets clogged up with mprobables filling your limited anticipation-set width.
Your system has to be continually updated (prior histories of self and other taken into account, particularly the prior history of the other, which may operate on a DIFFERENT subjective field than you that may not map to yours- but ALL operating on a common spatiotemporal database) to ensure a good set of probables that will flow through when necessary.
lol Microsoft was FORCED to cite this Professor Uht. I can see why he could quit teaching.
It was *Cited by examiner. In short, Microsoft TRIED to pass off this idea as their own.
A Microsoft patent
Systems and methods directed at providing flow computing. An application configured with flow computing capabilities can determine possible results that are intended by a user based on one or more user-initiated actions. The application identifies idling computing resources in the computing device that are available and automatically performs other actions to produce the possible results using the idling computing resources. The application caches the results and provides at least one of the cached results in an interactive data stream. The user directs the data stream by selecting pre-generated choices to achieve the desired outcome.