> Todd: I’m thinking of something a bit like erm the flap on a video tape
This doesn't sound like his design thinking operates in terms of analogies. It sounds more like he's imagining something which he feels he can best describe to his fellow designers by an analogy ("a bit like...").
I often wondered about this while running a quant group at a hedge fund.
Once your domain is somewhat specific, you tend to understand your own action constraints. We have this many guys with these particular skills (algos, networks, etc) and we can only consider things that might make at least some amount. And it needs to be done within some horizon.
Amazingly, it's much easier to have ideas if you have more constraints. Did you ever get asked by your teacher to write a story about anything? I always found that totally impossible. But a short story with a single protagonist, 4 pages long, occurring during some major life event, that's a lot easier to get started on.
We built a page of potential things to investigate. They were always things built around some observation someone had while becoming experts. So for instance I'd stared at swap rates for a long time and wondered about whether there were inefficiencies we could look at. Also there was a look at different ways to use options, another thing we were strong in. Everything that worked was something adjacent to existing knowledge. Ended up spinning out a currency fund that did quite well.
We don't know enough about the brain/mind to give anything like a measurable answer to this. It's all speculation.
However something rings true about the concept of analogies being instructive in idea formation. I often think about it as adjusting the assumptions that the mind uses when planning. In that sense, it looks just like improvisation or what we call exploration in Reinforcement Learning.
Good mental models of the world will result in more accurate simulations when you mix and match the assumptions/starting points. Some of those simulations might not be extant and you can take action to test whether they work in the real world. If they do, then boom you've got an idea made real.
> We don't know enough about the brain/mind to give anything like a measurable answer to this. It's all speculation.
You might be interested in the Leiden Theory of linguistics. It basically posits that language is an evolved semiotic organism that lives within the hospitable ecological niche of our minds, and cohabits within the vessel of our biology:
http://www.himalayanlanguages.org/files/driem/pdfs/prague.pd...
What a bad article. Started so well and then finished in a haste. There is a whole theory of invention called TRIZ. One idea there is tht transcending matter phase boundaries could lead to new inventions.. worth looking up.
Method: To explore this potential explanation of the association between far analogies and
reduced functional distance of search, we examined the concepts immediately preceding
far analogy-to-concept pairs (i.e., far concept-generating or function-finding analogies),
focusing on the distance of each concept from its immediate predecessor (i.e., its JUST
PRIOR value, derived from Study 1). In building this sample of concepts, we screened out
concepts that were not in the same subproblem space as the concept following the analogy,
and concepts with predecessors in a different subproblem space. The final sample
consisted of 57 concepts.
So the whole study based based on 57 verbalised concept transitions and the underlying assumption that iterative verbalised transitions within a group discussion context are true proxies for meaningfully tracking significant iterations in inventive thought process in the participants.
A counterexample would be our company, actively involved in R&D on a day to day basis across a range of disciplines very similar to those cited within the study, where far analogies are used however they are typically analyzed privately. The fruits of that thinking are presented for group discussion and the potential for those avenues of thought identified within group discussion, but they are almost never actually fleshed out verbally as a group.
13 comments
[ 2.5 ms ] story [ 47.1 ms ] threadI seem to remember him saying this a few times over the years and on occasion he’d say it as ‘good songs’ rather than ‘good ideas’.
This doesn't sound like his design thinking operates in terms of analogies. It sounds more like he's imagining something which he feels he can best describe to his fellow designers by an analogy ("a bit like...").
Once your domain is somewhat specific, you tend to understand your own action constraints. We have this many guys with these particular skills (algos, networks, etc) and we can only consider things that might make at least some amount. And it needs to be done within some horizon.
Amazingly, it's much easier to have ideas if you have more constraints. Did you ever get asked by your teacher to write a story about anything? I always found that totally impossible. But a short story with a single protagonist, 4 pages long, occurring during some major life event, that's a lot easier to get started on.
We built a page of potential things to investigate. They were always things built around some observation someone had while becoming experts. So for instance I'd stared at swap rates for a long time and wondered about whether there were inefficiencies we could look at. Also there was a look at different ways to use options, another thing we were strong in. Everything that worked was something adjacent to existing knowledge. Ended up spinning out a currency fund that did quite well.
You could have written a story about writing a story.
However something rings true about the concept of analogies being instructive in idea formation. I often think about it as adjusting the assumptions that the mind uses when planning. In that sense, it looks just like improvisation or what we call exploration in Reinforcement Learning.
Good mental models of the world will result in more accurate simulations when you mix and match the assumptions/starting points. Some of those simulations might not be extant and you can take action to test whether they work in the real world. If they do, then boom you've got an idea made real.
You might be interested in the Leiden Theory of linguistics. It basically posits that language is an evolved semiotic organism that lives within the hospitable ecological niche of our minds, and cohabits within the vessel of our biology: http://www.himalayanlanguages.org/files/driem/pdfs/prague.pd...
When you stop “talking”, you automatically start “listening”.
Listening is just exploring whats already out there?
Method: To explore this potential explanation of the association between far analogies and reduced functional distance of search, we examined the concepts immediately preceding far analogy-to-concept pairs (i.e., far concept-generating or function-finding analogies), focusing on the distance of each concept from its immediate predecessor (i.e., its JUST PRIOR value, derived from Study 1). In building this sample of concepts, we screened out concepts that were not in the same subproblem space as the concept following the analogy, and concepts with predecessors in a different subproblem space. The final sample consisted of 57 concepts.
So the whole study based based on 57 verbalised concept transitions and the underlying assumption that iterative verbalised transitions within a group discussion context are true proxies for meaningfully tracking significant iterations in inventive thought process in the participants.
A counterexample would be our company, actively involved in R&D on a day to day basis across a range of disciplines very similar to those cited within the study, where far analogies are used however they are typically analyzed privately. The fruits of that thinking are presented for group discussion and the potential for those avenues of thought identified within group discussion, but they are almost never actually fleshed out verbally as a group.