Osnat KM's Quick takes

[crossposted from my blog; some reflections on developing different problem-solving tools]

When all you have is a hammer, everything sure does start to look like a nail. This is not a good thing.

I've spent a lot of my life variously
1) Falling in love with physics and physics fundamentalism (the idea that physics is the "building block" of our reality)
2) Training to "think like a physicist"
3) Getting sidetracked by how "thinking like a physicist" interacts with how real people actually do physics in practice
4) Learning a bunch of different skills to tackle interdisciplinary research questions
5) Using those skills to learn more about how different people approach different problems

While doing this, I've come to think that identity formation - especially identity formation as an academic - is about learning how to identify different phenomena in the world as nails (problems with specific characteristics) and how to apply hammers (disciplinary techniques) to those nails.

As long as you're just using your hammer on a thing that you're pretty sure is a nail, this works well. Physics-shaped hammers are great for physics-shaped nails; sociology-shaped hammers are great for sociology-shaped nails; history-shaped hammers are great for history-shaped nails.

The problem with this system is that experts only have hammers in their toolboxes, and not everything in the world is a nail. The desire to make everything into one kind of nail, where one kind of hammer can be applied to every problem, leads to physics envy, to junk science, to junk policy, to real harm. The desire to make everything into one kind of nail also makes it harder for us to tackle interdisciplinary problems - ones where lots of different kinds of expertise are required. If we can't see and understand every dimension of a problem, we haven't a hope in hell of solving it.

The biggest problems in the world today - climate breakdown,  pandemic prevention, public health - are wicked problems, ones that can't be easily reduced into nail-shaped problems for hammer-shaped solutions. They require lots of different problem-solving approaches for lots of differently-shaped problems.

The question, then, becomes how we stop trying to make everything into a nail - and that requires rethinking what it means to be able to problem-solve in the first place.

As I touched on briefly before, training to "think like an expert" is a set of processes that consist of looking at phenomena, figuring out how to approach those phenomena, and figuring out how to arrange these phenomena into something nail-shaped. Becoming recognised as an expert is about being able to demonstrate to others that you can take a phenomenon, form it into a nail shape, and apply your disciplinary hammer to it. If you are very lucky or work in a unique field, you might be trained to see phenomena as having more than one "shape" and more than one useful toolset, but this is still less common than it should be.

What's more, as different disciplines mature, we start finding out new things and working in the fields get ever more complicated. This means that when it once might have been possible to apply the disciplinary hammer to a wider set of problems, the maturation of a field means that the set of problems where you can usefully apply the disciplinary hammer becomes smaller and smaller. We have an ever-larger set of problems and most of our tools are wildly unsuited for solving them.

How, then, do we move forward? We certainly can't close the proverbial stable doors after the horse has bolted; the only way to stop fields from maturing is to stop people from working on them entirely. Good luck with that.

Instead, what we should do is put people working in wildly different fields in the same social spaces. We should encourage them to talk to each other - not just about work, but about life and love and random trivia - so that we can understand each others' perspectives. We should prioritise interdisciplinary and transdisciplinary work, as some funding bodies are already doing. And we should really listen to each other and be explicit in how we try to approach problems. Maybe then we can stop seeing everything as a nail, because we'll have more than just a hammer in our collective toolbox.

I've been reading AI As Normal Technology by Arvind Narayanan and Sayash Kapoor: https://knightcolumbia.org/content/ai-as-normal-technology. You may know them as the people behind the AI Snake Oil blog.

I wanted to open up a discussion about their concept-cutting of AI as "normal" technology, because I think it's really interesting, but also gets a lot of stuff wrong.