Tag Archives: inverse model

A Danish Breakfast with Andreas Roepstorff

 

Breakfast with Andreas

We are very fond of the light, bright, functional Scandinavian style of our apartment in Nobelparken, Aarhus University. It has been our home during our many stays, short and long, over the last ten years. The reason that we are coming here to the Interacting Minds Centre, and keep coming back, is Andreas Roepstorff. One of our special treats is if he comes to breakfast and brings with him freshly baked bread. Today he has also brought a chokoladestang, a classic Danish pastry, extra special. Actually, Danish pastry is called ‘wienerbrød’. All the ingredients of a ‘hyggelig’ breakfast are here. Also candles, thin chocolate slices, honey, jam, cheese, rullepølser, and plenty of coffee. Perhaps there are some other essential Scandinavian style ingredients, an open plan flat and some open plan minds.  Continue reading A Danish Breakfast with Andreas Roepstorff

In this context anything can be discussed. There is no need to be afraid of the big and awkward questions. We started talking about overall goals, long held beliefs, what we would really like to accomplish. Andreas suggested straight away that we need to think of hierarchies of goals, referring to Etienne Koechlin. Koechlin has mapped out hierarchies of goals in the Prefrontal Cortex. Long held beliefs will be kept in the background and other more short term beliefs will be nested within.

It’s all about upholding alternative views of the future, Andreas says. Then Chris throws in “Mental time travel”. I say “Episodic foresight” – as in a pleasant game of ping-pong.

AR: There is the open future – there are several possibilities in front of you.

UF: Ah – so like you to say this. The Viking spirit and your trademark – the Blue Ocean.

CDF: This is where other people give these alternative views.

UF: Where culture = other people.

AR: Okay. Think of the Blue Ocean – we can navigate in the future. We explore.

CDF: Not only finding out what world is like, but creating a world as we’d like it to be.

AR: Creating the world in your image. God-like.

UF: ??

AR: Religion is an extension of the social image. It’s a hierarchical story. You create the top of the hierarchy.

UF: Go on.

AR: The world is unpredictable. But we believe there is a real world – this constrains us.

UF: ??

AR: The other problem is: other people have different perspectives. We have the Ukraine example.

UF: If everyone had the same idea, what then?

AR: Maybe it’d be like China in the old days.

— We pause to help ourselves to more bread, butter and stuff to put on top.

 

AR: Here is a long held belief: the openness of the human cognitive system.  But there are constraints. Low level behaviour is constrained totally by immediate environment.

CDF: The higher up in hierarchy, the less constrained.

AR: The ability to share possible futures, to co-create, undoes the straight jacket that is there otherwise.

CDF: Humans are good at creating new niches.

AR: Yeah, and at changing the world to fit them. Imagining the future.

Aarhus sybilsUF: I have always been struck by the deep human interest in forecasts. You can see it in ancient archaeological sites, like Stonehenge, or the oracle of Delphi. There are the lovely medieval pictures of the Sybils in Aarhus cathedral. People have been obsessed by predicting the future. They have lots of devices for doing so. Why?

AR: You need external help, devices to forecast. Astronomy is way beyond human time scale. How long do you need to know that something is cyclical?

CDF: The Babylonians knew about the 18 years moon cycle.  They had to have instruments to monitor their observations over such a time scale. And once you start monitoring…there comes control.

UF: Are there some practical suggestions here? I would like to predict our own future.

AR: I have devised this exercise for postdocs: This is what I ask them when they first come:  Imagine yourself in five years time. Everything has gone as well as possible. It is 2019. You are invited to give keynote speech at a big conference. In Hawaii, no less. How did you get there? What would you want to tell people? Who is in the audience? Whom would you like to impress, living or dead?

UF: Nice.

AR: Some students know on the spot. They can then write a grant application.

UF: There is the curse of the here and now. An eternally extending tree of possibilities. The more you think about it the more choices you have.

AR: Exactly. You cannot see the path from here that will lead to a future endpoint. There are just too many branches. But you can, if it’s the other way round: Start from the endpoint and go back to the present.

postdicting the future

CDF: This is analogous to Daniel Wolpert’s solution to the motor problem by minimising endpoint variability.

UF: We have a theory now.

CDF: When you make an action there are an infinite number of ways to proceed. You first choose the endpoint and then you can minimise end-point error.

AR: Cultural experience allows a blueprint for the future. People do have expectations of their future. Ideal scenarios have an internal logic. You get all this from your culture and can frontload it in a ‘cognitive app’.

UF: So the blue ocean has some pre-defined spots to aim for?

AR: The critical thing is to place a buoy out there. Then you have to do something to get there.

CDF: Your exercise for postdocs is all very well. But what it doesn’t allow for is taking advantage of unforeseen opportunities that happen to occur. You should take this up and then you could go into a completely different direction.

AR: Reconfiguring is always fine. It is easier to follow a tangential line when you have a larger perspective. If your goals are too short term then this doesn’t work, and you can go off course.

UF: The proper Long Term perspective is from death.

AR: From beyond death. You have to see yourself from other people’s point of view in the far future. Think of your legacy.

CDF: What my mother used to call “the backdrop of eternity”.

— More helpings of coffee and chocoladestang are now necessary. Chris offers to make more coffee.

 

AR: Going back to the Blue Ocean and the exercise for postdocs. It’s necessary to anchor the imagination in a specific place. Hawaii. And there need to be specific people who the post-doc would like to be in the audience, and he would like to impress.

UF: I like this idea that specific constraints like these concrete anchor points actually get the imagination going.

AR: If you don’t have these far flung anchor points, then you can get into cognitive apathy; the more you think about it, the more choices open up, and it becomes even more complicated. I have been there….

CDF: In the Tower of London game where you have to plan future moves, there is this interesting observation: if there are two equally good moves, then you slow down. You would have thought you could speed up, since it doesn’t matter which one you choose! Now here’s an experiment: Would performance be improved if you asked people to think about planning their moves backwards rather than forwards?

UF: Good – we always need to finish on a suggestion for a new experiment!

image credit: groenling at flickr (1989): Sybils in Aarhus cathedral