Machines now grant whatever we ask, better every year. Choosing what to ask was never their half of the problem. It is ours, and we are bad at it. Thirteen parts, one argument.
Maybe coming to terms with reality as the feedback loops, not the goals.
The Golden Rule, not the pile of gold, because what goes round, comes round.
Personally I would doubt enough people would be willing to look too deeply into the reality, because we all want our treats and trinkets, but it does seem our current overlords are doing their utmost to crash the whole system, chasing their gods and goals.
Thanks, John. The feedback-loop reading is close to where the series is going.
The loop is where the damage shows up, but it runs on whatever was written down at the start. A system pointed at the pile of gold will find every shortcut to it; that's not the loop failing, that's the loop working. The fix has to reach the wording.
A later part takes the institutional version, KPIs and engagement metrics, which is roughly your overlords case.
So then the next step with be the circuit breakers.
Knowing when to leave the party, as the feedback turns negative at the crest of the wave. Each cycle has to push a little further, testing the boundaries, in order to grow, so it has to crest, to cycle. Stasis is not an option.
I think the big issue is to understand time. That this linear narrative flow, from past to future, is a consequence of being mobile organisms with a sequential process of perception, while the reality is activity and the resulting change turning future to past. Tomorrow becomes yesterday, because the earth turns.
Then we can better appreciate the primacy of the thermodynamics. The nodes synchronizing, versus the networks harmonizing.
Circuit breakers are the right next step, and notice what they are: a second piece of wording, stop when this happens, written before the run starts.
So the breaker moves the problem rather than escaping it. Someone still has to say what counts as the crest, and that is the hardest sentence in the whole setup.
The sequential story probably is the perception, not the substrate. An optimizer doesn't experience its updates in order either, it just sees one gradient at a time.
What if a big part of the problem is this assumption of a problem?
The reality is this mixture of friction and flow.
The primary solution is to recognize the various clocks, cycles and their coordination. Boredom, for example, is one's internal clocks running faster than external ones. Anxiety is when the external clocks are running faster.
What AI presumes to do, is to speed up the process, assuming it will find that pot of gold at the end of the rainbow. Creating lots of anxiety and rushing around, trying to keep up.
Consider the essence of Eastern philosophy, Zen, is that balance. The state of equilibrium between the yin and yang. "Nothingness." The harmonization of the network.
While the Western paradigm is more focused on the node. Everything as one, from monotheism to the Big Bang cosmology.
So would someone schooled in Eastern thinking be looking for the gradient, or would they be seeking the flattest line along the side of the mountain?
So is the goal finding the deepest valley, or is it finding one's way through the mountains? There is no universal map, because there is no universal goal.
The force of gravity might be universal, but any particular gravitational field is not.
Maybe coming to terms with reality as the feedback loops, not the goals.
The Golden Rule, not the pile of gold, because what goes round, comes round.
Personally I would doubt enough people would be willing to look too deeply into the reality, because we all want our treats and trinkets, but it does seem our current overlords are doing their utmost to crash the whole system, chasing their gods and goals.
Thanks, John. The feedback-loop reading is close to where the series is going.
The loop is where the damage shows up, but it runs on whatever was written down at the start. A system pointed at the pile of gold will find every shortcut to it; that's not the loop failing, that's the loop working. The fix has to reach the wording.
A later part takes the institutional version, KPIs and engagement metrics, which is roughly your overlords case.
Hugo,
So then the next step with be the circuit breakers.
Knowing when to leave the party, as the feedback turns negative at the crest of the wave. Each cycle has to push a little further, testing the boundaries, in order to grow, so it has to crest, to cycle. Stasis is not an option.
I think the big issue is to understand time. That this linear narrative flow, from past to future, is a consequence of being mobile organisms with a sequential process of perception, while the reality is activity and the resulting change turning future to past. Tomorrow becomes yesterday, because the earth turns.
Then we can better appreciate the primacy of the thermodynamics. The nodes synchronizing, versus the networks harmonizing.
Yin and yang, than God Almighty.
Circuit breakers are the right next step, and notice what they are: a second piece of wording, stop when this happens, written before the run starts.
So the breaker moves the problem rather than escaping it. Someone still has to say what counts as the crest, and that is the hardest sentence in the whole setup.
The sequential story probably is the perception, not the substrate. An optimizer doesn't experience its updates in order either, it just sees one gradient at a time.
Lol.
Do you think you will do away with "the problem?"
What if a big part of the problem is this assumption of a problem?
The reality is this mixture of friction and flow.
The primary solution is to recognize the various clocks, cycles and their coordination. Boredom, for example, is one's internal clocks running faster than external ones. Anxiety is when the external clocks are running faster.
What AI presumes to do, is to speed up the process, assuming it will find that pot of gold at the end of the rainbow. Creating lots of anxiety and rushing around, trying to keep up.
Consider the essence of Eastern philosophy, Zen, is that balance. The state of equilibrium between the yin and yang. "Nothingness." The harmonization of the network.
While the Western paradigm is more focused on the node. Everything as one, from monotheism to the Big Bang cosmology.
So would someone schooled in Eastern thinking be looking for the gradient, or would they be seeking the flattest line along the side of the mountain?
So is the goal finding the deepest valley, or is it finding one's way through the mountains? There is no universal map, because there is no universal goal.
The force of gravity might be universal, but any particular gravitational field is not.