[whack]

One way to look at creative thinking, is as local-maxima vs global-maxima searches. Someone who's searching for a local maxima, very close to the current state of the world, will produce ideas and work that show steady and consistent improvements. At some point, these finding-local-maxima will hit diminishing returns, but at least the process is predictable and very likely to bear some fruit.

In contrast, attempting to jump far away from the current state, and finding a brand new global-maxima, is a much more risky endeavor. The first couple iterations may produce results that are even worse than the current state of the world. But in the long run, it avoids the problem of diminishing returns, and can lead to occasional breakthroughs that are vast improvements over the status quo.

It seems to me that creativity is basically about foregoing the easy and predictable local-maxima-search, in favor of a more adventurous global-maxima-search. The first few iterations of your bold new idea may sound kooky and klunky, but once it's been developed with sufficient rigor and polish, it has the chance to give a much bigger payoff. Someone who's too focused on the small details, and getting every detail ironed out before committing to something, may find such an endeavor far too uncertain to undertake, thus missing out on what could be the next big thing to change things up.

[westoncb]

Another way of thinking about this is to consider the different means available for navigating 'idea space': one can crawl up hills to local-maxima using discursive thought, or one can sort of teleport around via analogical processes (as you point out, it's not clear what sort of valley you may end up in by doing this, though). Discursive thought has more of an 'if, then' character, while analogical processes seem to be kicked off by wondering about which things might have a similar structure to (i.e. are isomorphic to) some other thing of interest.

Edit: and another way: the local-maxima situation has to do with finding solutions within some particular framework, whereas you get a shot at global maxima in the search for/attempt to build new frameworks.

[edward_rolf]

You both make great points.

Can we find this conundrum in other domains, say, neural networks? Can a machine become creative when we teach it to adhere to common knowledge?

[MrQuincle]

Neural networks also need to balance exploration and exploitation. You will find it in the form of:

+ generalization / overfitting

+ cross-domain knowledge transfer

In a more embodied approach it is free energy minimization as advocated by Friston. Consider us divided by the world through a Markov blanket, or even better consider our actions and perceptions divided by one. How do we continuously surprise ourselves without getting mad?

A guy who would say it's the same thing is probably Polani with empowerment: https://arxiv.org/abs/1310.1863. The rational choice is to navigate to that location in state space where you have most decisions.

I think curiosity-based research is quite interesting from the perspective of rationality and creativity.

[westoncb]

Personally, I see no reason why creative machines couldn't be built.

Another thought on the local/less local/global maxima problem: it seems possible that, taking a very abstract view of things, the whole reason we developed the sort of thought which distinguishes humans from other animals is to address the escaping local-maxima problems (summed up by the situation where you're in a maze, but have to go 'backward' in some sense in order to get out). While discursive thought may be worse than analogical thought at this sort of thing, our lower 'animal brain' functionality is even more geared to dealing with local maxima/minima.

[GuiA]

Well, what you're describing falls exactly into the second category of exploration that 'westoncb was outlining. Only one way to find out!

[reacweb]

Creative thinking occurs at a global scope and also at the detail level. I would better correlate creative thinking with dissatisfaction. IMHO a creative thinker wants to improve things or to explore new ways. A creative thinker has difficulties to follow scrupulously a process and would make a bad accountant. He may be inaccurate, but I do not think it is a rule.

[nikki93]

The interesting thing is when it turns out you are exploring this hill-space by looking at the ground and nearby like bacteria or such, but over time "humans" have evolved things like being able to look toward someone that's on a different hill shining a light at you, which allows communication across agents exploring this space in channels outside the utility function itself.

[misja111]

I had a similar idea, but I think it is more accurate to compare creative thinking with Monte Carlo tree search. Setting the exploitation vs exploration ratio higher means you will miss some obvious candidates, but now and then you will find some 'genious' idea.