Yo this is an awesome list, thanks. I’ve recently been playing around with knowledge graphs for some of the modelling I do. Cool to see you guys having a similar idea plus a lot of other ones too
I once had an idea about making a social media platform that would reward societally good deeds of humans towards others, to shift our competitive society into a more collaborative one in everyday life. I think this post gave me renewed inspiration to develop that vision further. I would be glad to discuss how this approach of ai development could be implemented to such a platform. Ideally, we would have as many critically thinking intelligent collaborative people working towards societal improvement in unison, which could be initiated by a shift in ones view of oneself in (western, capitalism-oriented, democratic) society. This could potentially be triggered if a large enough portion of the population of an overly competitive society, reliant on social media migrated to more thought-provoking and group-centric platforms, to feed their chemical information channels.
That was honestly a great read. For implementing Angels on the Shoulder for real world use, some sort of vr/ar experience would potentially go along with very well.
We obviously should not try to make a society "aligned", since that would be counterproductive to any group of people aiming for progress. But rather, as humans tend to act more out of incentives, rather than courage etc, it should be noted a push to keep the "decency" of individuals could be a goal to strive towards.
I realize the thought process is not extraordinary in and of itself. The point is the flow of information should be kept open and incentives made for others to make decisions that would improve ones position inside a social electromagnetic web structure. Who decides those incentives will be, at least on a small scale, programming a population towards solving problems inside the created constraints.
If a platform was to incentivize users towards working to realize generally beautiful/helpful projects, maybe a larger pool of humans were to get attracted towards building a better tomorrow. It should include elements of many concenpts, to make it work as intended no doubt.
That was honestly a great read. For implementing Angels on the Shoulder for real world use, some sort of vr/ar experience would potentially go along with very well.
We obviously should not try to make a society "aligned", since that would be counterproductive to any group of people aiming for progress. But rather, as humans tend to act more out of incentives, rather than courage etc, it should be noted a push to keep the "decency" of individuals could be a goal to strive towards.
Yo this is an awesome list, thanks. I’ve recently been playing around with knowledge graphs for some of the modelling I do. Cool to see you guys having a similar idea plus a lot of other ones too
Cool! See also my and my colleague Ben’s elaboration on the knowledge graph/epistemic stack: https://www.oliversourbut.net/p/a-full-epistemic-stack
If you have thoughts on this or other parts (challenges, bottlenecks, opportunities, approaches, …), I’d be interested to hear them.
Thanks Oliver! I’ll have a look
I once had an idea about making a social media platform that would reward societally good deeds of humans towards others, to shift our competitive society into a more collaborative one in everyday life. I think this post gave me renewed inspiration to develop that vision further. I would be glad to discuss how this approach of ai development could be implemented to such a platform. Ideally, we would have as many critically thinking intelligent collaborative people working towards societal improvement in unison, which could be initiated by a shift in ones view of oneself in (western, capitalism-oriented, democratic) society. This could potentially be triggered if a large enough portion of the population of an overly competitive society, reliant on social media migrated to more thought-provoking and group-centric platforms, to feed their chemical information channels.
Difficult one! See also the discussion of ‘aligned recommender systems’ e.g.: https://newsletter.forethought.org/p/angel-on-the-shoulder-ai-tools
That was honestly a great read. For implementing Angels on the Shoulder for real world use, some sort of vr/ar experience would potentially go along with very well.
We obviously should not try to make a society "aligned", since that would be counterproductive to any group of people aiming for progress. But rather, as humans tend to act more out of incentives, rather than courage etc, it should be noted a push to keep the "decency" of individuals could be a goal to strive towards.
I realize the thought process is not extraordinary in and of itself. The point is the flow of information should be kept open and incentives made for others to make decisions that would improve ones position inside a social electromagnetic web structure. Who decides those incentives will be, at least on a small scale, programming a population towards solving problems inside the created constraints.
If a platform was to incentivize users towards working to realize generally beautiful/helpful projects, maybe a larger pool of humans were to get attracted towards building a better tomorrow. It should include elements of many concenpts, to make it work as intended no doubt.
That was honestly a great read. For implementing Angels on the Shoulder for real world use, some sort of vr/ar experience would potentially go along with very well.
We obviously should not try to make a society "aligned", since that would be counterproductive to any group of people aiming for progress. But rather, as humans tend to act more out of incentives, rather than courage etc, it should be noted a push to keep the "decency" of individuals could be a goal to strive towards.