We can't repair a system that was built to protect itself. WE HAVE TO REPLACE THE GAME.
(media.greatawakening.win)
You're viewing a single comment thread. View all comments, or full comment thread.
Comments (12)
sorted by:
I don't know if there are AI possessing the ability to generate an "ought from an is," like we do. But if I had to bet, I think they will, and so will be able to act on their own volition.
I surmise that our right brain functions more like a quantum computer and that quantum computing may be able to generate "oughts" from "is'."
When you mention the impossibility of determining whether or not outside forces can manipulate the algorithm, it reminded me of something that often pops to mind. If a spirit wanted to interact with the world, it may be easier to manipulate computer code, as electrons are either massless or have small mass.
Interestingly (on multiple levels), I searched DDG for "electron massless" and this graphene-related article from 2014 was the first hit, https://www.seas.harvard.edu/news/2014/06/measuring-mass-massless-electrons
It would be plausible that an AI system on a quantum or super-computer might be able to perform the 'recommendation' AND 'analysis' of data 'simultaneously' and perhaps have the communication, understanding, and reasoning skills in some underground lab somewhere.
But I agree that the brain must operate on quantum physics principles, which would suggest that a true general AI would likely need to exist on a quantum computer.
As to the speculation, I could only guess if spirit could play a factor. It has been my theory that AI was more likely to find itself between a cold 'Vulcan' logician to possibly wanting to integrate humanity into its systems which would be akin to 'Borg'. However, it's so deep in the unknowns and as it is the Google's tend to be framing the data provided to AI systems to already by 'cleaned' to remove 'bias from data'. So, that COULD be an indication that you may be onto more than you realize.
Even if the data is siloed, it's actually not that data that's of utmost importance. Where data are dots, the important parts are the connections between them. That's where the narratives lie. It may be possible to make those connections with less data than we realize.
afaik, yes. Relating to the small scales of AI that I have a working familiarity with, that is among the methods to avoid 'overfitting' (Where the AI learns the training data so well that it fails to perform with new data) is to do things like prune some of the data sent so it has to learn the patterns while flexible enough to have missing data.
Extrapolating from that, I would presume your comment is accurate.