Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

What is a problem?

1 Post author: Elo 12 March 2017 02:04AM

Original post: http://bearlamp.com.au/what-is-a-problem/


I originally posed this idea in my list of short stubs.  Under that heading I briefly outlined:

What is a problem - On the path of problem solving, understanding what a problem is will help you to understand how to attack it.  Nothing more complicated than this picture to explain it.  The barrier is a problem.  This doesn't seem important on it's own but as a foundation for thinking about problems it's good to have  sitting around somewhere.

whatisaproblem


I want to expand on that a bit more. I have labelled some states in this picture:

  • Present state
  • Goal state
  • Barrier to the goal
  • Path to the goal

Present state

All things being unchanged; prior to actions in pursuit of a goal, this is where you are.  Sitting at home in my chair at my computer I have not yet decided I want to go get ice-cream.  If I do nothing, I might eventually end up getting ice-cream by happenstance.  I may casually interact with friends who encourage me to get ice-cream with them this evening.  Nothing entirely stops me from getting ice-cream but also nothing propels me to do it either.  Without goals, without paths, you can live a lot of life, casually random-walking your way through the galaxy, encountering what you encounter, and responding at will to those stimuli.  We might describe a specimen who only cares about the present state, as having low agency.

Goal state

Let's pretend that we duplicated the universe, with the slight modification that I am now eating ice-cream.  That might be my goal.  Or as close as possible as I can get to that goal-state.  There are lots of things that are not ice-cream-goal state and lots of things that come close.  I could eat my toes, I could eat some cheese I have in my fridge which is a bit cold.  I could eat an apple, I could eat some ice cubes, I could drink a glass of milk, I could make my own ice-cream, I could give ice-cream to other people.  All of these things are not quite the goal, but are quite close.  Notice that this goal doesn't currently include my path to ice-cream.  Just that I have a goal now.

Path

We can visualise a path in many ways.  This should be unsurprising.  The fact is that if I want ice-cream, I need to get up out of my chair and look in my freezer for the ice-cream.  If I want to be eating it I probably also need a spoon and some way to open the lid of the container.  But actually when I get to the freezer I remember I don't keep ice-cream in my house because I am on a diet and that slows me down from eating ice-cream.  (Thanks past-me)  With that in mind my corrected path is actually to find my money, exit my house, and go to the store where I can buy more ice-cream.  (or make my own ice-cream) or any number of other pathways to ice-cream.

System 2 is very good at paths.  So good that in fact that it forgets about what can go wrong, and the barriers.  It's the part of my brain that will tell me that the nearest ice-cream is in the store and I can walk right up to the freezer section and shove the ice-cream into my face (Disregarding the need for spoons, the need for money to pay for the ice-cream, and the need to be wearing a shirt when I leave my house).

If this seems obvious; it is only obvious because it already all makes sense to you.  For someone without an understanding of states or paths or goals this might be a solid learning step.

Barrier

A barrier is what gets in the way of the path.  There are always barriers and most of them are overcome without a second though.  As you think deeper and harder about barriers they get more complicated, and start encompassing more details.  The barriers to me getting my ice-cream include knowing how to walk (which I know) and knowing how to use my eyeballs.  The simple barriers get overlooked because we already have them down.

The more complicated barriers start to pop up as you think more about the goal, or set up more complicated goals, or goals that are further away (geographically, mentally, metaphorically, some measure of hardness of goal).  

I have to go to the store to get ice-cream.  I need social convention down pat, so I remember to put a shirt on.  And I need money to pay for ice-cream. Hey, I probably need a functional democracy, not to mention electricity, production lines, hygiene and a whole lot more built on top just so I can do that.  Assuming all those great things are there, what if I don't have money.  Well I would have to work, which means I need a functional economy, some way to trade my abilities for currency (another thing I need) which I can later exchange for ice-cream.  

There's another type of barrier that doesn't quite fit with the rest, and that's the barrier inside my own head.  The barrier that says, "ice-cream, but I would have to get up out of my chair for that" and decides against it.  Ice-cream as a goal, is not so desperate that I would die if I didn't get it, but maybe there are goals that are more serious.  Having enough money, friends, family, what are your goals?  Have you seen the list of common human goals?  Each goal has barriers to completing it.  And each barrier is able to be stared at intently and questioned.

Is this barrier going to stop you?  Or are you making excuses?


Next post: On excuses and validity


Meta: This took an hour to write and will be a foundation for a few posts that follow.

If this doesn't seem all that big a deal, well. It's not.  Unless your problem solving, solution seeking barrier facing system is not functioning at optimum.  In that case: knowing how it works, knowing what is or is not functioning, being able to debug this process.  That's breaking down the meta barriers.

Comments (4)

Comment author: turchin 12 March 2017 08:31:56PM 1 point [-]

My problem is that this simple schematic is not working: I have a knot of goals and blocks to them, which are interdependent. For example, I have deadline for the article, but I have chronic tiredness which prevent finishing it, which I could fight with coffee but I have high blood pressure, and I have not reliable medical way to control it + 20 more interconnected goals and blocks.

Attempts to write down all my goals and problems did not help as they created enormous maps and feeling of doom, as well as I spent a lot of time on such planning without practical results.

Basically I try to find most urgent and solvable problems and work on them, but I think that there should be a better strategy.

Most problems are also tend to grow on higher level, that is they will be solvable if I have more time, money, higher intelligence or other high level resources.

Comment author: Elo 14 March 2017 09:25:43PM 1 point [-]

I have a knot of goals and blocks to them, which are interdependent.

This is a really interesting problem. I have noticed it in people who end up on the street and have several of any of the following going wrong at the same time, (mental health troubles, drug problems, alcohol problems, addiction problems, out of work, relationship problems).

The problem is also that there are too many problems. (see: http://bearlamp.com.au/the-problem-tm-analyse-a-conversation/ and http://bearlamp.com.au/the-problem-analyse-a-conversation-part-2/ for related).

In relation to my example above - It's really hard to solve a drinking problem when mental health problems just pop up, or solve an addiction problem when relationship problems just pop up as a result or consequence or both. There is no easy answer to what to tackle first or how to get it right. There is no standard answer.

I believe the strategy should be to pick one of the problems, work on that one very hard whilst the other ones sit in the background bubbling away, then when it's a bit more in control, pick the next one and try work on that one, then repeat and cycle till shit gets a bit more under control.

Try not to let anything go backwards, and keep putting out the spot fires.

Comment author: turchin 14 March 2017 10:46:40PM 0 points [-]

Yes, good example with a homeless person. One feature of knot problems is that their complexity is higher than available level of problem solving ability. Alcoholic with mental problems could plan and implement only very simple strategies.

So one of the possible solution is replace intractable knot with simpler knot of problems. For example if homeless person will go to jail, it may solve many of his problems, like addiction, home, food, basic healthcare (it is not universal solution, as being in jail has some other disadvantages).

Another line is solve problems which will help in other problem solving, that is similar to intelligence improvement. For homeless person it may be to create a friend and ask for help.

The third line is just solve nearest an most urgent problems as they appear - I think it the way how actual homeless people live, but it is not way out.

Comment author: Elo 15 March 2017 05:07:31AM 1 point [-]

it the way how actual homeless people live, but it is not way out.

I need to research more, some people get out. But I don't know why or how.

You have all the same ideas that I do. Don't think I can advance you more than you are already moving unless I know something more specific about your problem areas. I am quite sure there are no hard and fast rules about solving this.