Solving Beats Planning

Today, for the first time in four years of living in Seattle, I got soaked on the way to my coworking space. Despite its reputation for being rainy, it rarely rains hard here. My light rain jacket wasn't enough, so I'm sitting here bare-footed, with my shoes and socks drying in a corner. I'm doing what an engineer might call solving for dry.

Solving is a curious cognitive process, situated somewhere between play (you solve puzzles for fun), skill (you usually need to have practiced certain behaviors), insight (you need to actually look at and understand what you're working with, it's not blind recipe following), and orienteering (repeatedly picking a direction to go; in my case, I could be solving for warm for instance, which would involve different actions, but damp bothers me more than cold). As an adult, you also need to do your solving for X in ambiguous, uncertain and poorly defined domains. No clean-edged game boards for you.

But child or adult, solving should be your default word, over planning. I recommend you try substituting solve for plan in conversations and seeing how your thinking and habits change. And to help you along, here's a discussion of the distinction.

View this email in your browser

Solving for X in an Idea Fog

1/ At a recent client meeting, I heard the phrase solve for X used a lot. It was a pleasant change from plan for X. As regular readers know, I have a bit of an aversion to planning and goals.

2/ The difference between the two phrases is huge. Plan is a managerial word. Its working material is solved problems. Solve is an engineering word. Its working material is untamed reality.

3/ Planning assumes that known means need to be arranged towards an end, where a "means" is a solved problem: a bit of reality understood as a known cause-effect map rather than as territory.

4/ Planning also assumes that ends are known, and the challenge mainly lies in choosing among available ends, by weighing "costs and benefits" and working out the "details."

5/ The assumption is that this process can be confined to a certain range of "operations" abstraction levels between “strategy” and “tactics": regimes meant for head-in-the-clouds bullshitters and lowly flunkies respectively.

6/ As some unimaginative military types like to say, amateurs worry about strategy and tactics, professionals talk about operations. This is not actually a flattering image of professionals.

7/ Professionals in this narrow sense are the sort who always deliver on time, as promised. They can do that because they never wander into regimes that require challenging “solves.”

8/ Solving domains, unlike planning domains, do not tolerate such "professionalism". If you are not comfortable being an "amateur" 2/3 of the time, you cannot solve. You can only plan.

9/ Planning can work with a purely instrumental view of the environment where everything has a clear purpose that it fulfills well, and can be understood in functionally fixed ways.

10/ As a result, planning in an everyday sense is mainly exercises in administrative logistics that pretend to offer predictable, time-bound outcomes because they avoid tough solves.

11/ This is why planners love middle-management bullshit like goals that satisfy SMART criteria:  if you can literally set such goals, you're probably not solving problems of any significance.

12/ At a purely mathematical level, even “solving” a “planning” problem, something clueless types think they are competent to do because they are “organized process people", is non-trivial.

13/ "The literature on planning and scheduling in artificial intelligence generally takes it on faith that any interesting problem is at least NP-hard" as this widely cited paper says.

14/ Translated to layspeak (try this if the term NP is unfamiliar), this means unless you find a creative local exploit, most such problems cannot be solved in meaningfully time-bound ways. It'll get done when it gets done.

15/ Worse: often you cannot even solve such problems approximately in "SMART" ways: the approximate versions of many NP-hard problems are themselves NP-hard.

16/ And that's just uncertainty in time. There's plenty of other kinds of uncertainty in meaningful problems. And we haven't even touched on ambiguity yet.

17/ This means all the professionals who pride themselves on their on-time delivery and reliability are likely not solving real planning and scheduling problems.

18/ At best they are brute-forcing bad solves through some mix of bullying and indifference to what's actually delivered. Fulfilling the letter, but violating the spirit of expectations set.

19/ At worst, they are creating an illusion of "execution" that does nothing ("going through the motions", "phoning it in"), through bad-faith theatrical displays of productivity and busyness.

20/ "Solving," unlike "planning" couples purposeful behavior and iterative outcome selection via a reality loop. What you solve for depends on what you can do, and vice-versa. What you deliver is a discovery, not a promise.

21/ Solving does not promise more certainty and clarity in outputs that can actually be delivered given the uncertainty and ambiguity in inputs, and current "solve rate."

22/ How does this work? Our habits and mental associations around the words planning and solving are set early, in educational institutions.

23/ As a kid or college student you probably argued at some point about how in math there's "always a right answer" and in the humanities "there are no right or wrong answers."

24/ Possibly, you picked one side or the other, and found some way to adopt a superior stance towards the other side. These attitudes set in early and hard.

25/ This schism is created by how the subjects are taught. You might solve the algebra problem x^2-4=0 with x=+2, -2 (that's still really just one answer, not two).

26/ You might answer an essay question like "Describe Lady Macbeth's character" with either a conventional Cliff-notes type answer, or a risky, contrarian reading of the play.

27/ The two kinds of "solutions" are actually not that different. They just represent different mixes of play, skill, insight and orienteering.

28/ Real life, fortunately, is way more interesting than school or college mid-terms. "Solve for < 2C global warming" is a challenge with both math-like and literature-like aspects.

29/ You get to choose what to solve for (solve for clean energy? solve for no-fossil fuels? solve for carbon sequestration? solve for refugee relocation?)

30/ You cannot afford to be a functionally fixed thinker. Are hybrid cars better than all-electric? Depends. Maybe hybrids delay the solve by creating a false sense of progress.

31/ Assumptions shift all the time, "$200/barrel oil will drive supply and demand to solve the problem naturally". Oops we invented fracking and oil is now under $50/barrel.

32/ Sometimes you wander really far off solving a seemingly unrelated problem. What engineers call yak-shaving. This can be a good thing.

33/ Why? Remember those clever mazes you solved as a kid where the right path wandered far , and the apparent straight line path dead-ended somewhere far from the exit?

34/ Or you end up with low self-awareness and mistake your lack of courage/skill for the  conviction that the color of the bike-shed matters in the design of the nuclear powerplant.

35/ Balaji Srinivasan coined the term idea maze for the context in which problem solving happens. I've come to prefer my updated term, an idea fog.

36/ Mazes are a great metaphor, and recall to mind the right kind of childhood problem-solving experiences to apply in adult life.

37/ They do suggest a designed playing field created by a planner though. I prefer "fog" because it suggests unplanned uncertainty, ambiguity, and creative path finding.

38/ It's a smooth rather than a striated metaphor for path-finding. One that suggests you can punch through walls and clamber through ventilation ducts.

39/ A planner views every domain as though it were a city with low map-territory divergence and no nasty lurking surprises.

40/ A solver views all domains, including cities, like poorly mapped wildernesses rich with possibilities and new knowledge.

41/ There are trail maps, but they tell you only a little of what you might need to know. Your hike can take unexpected turns, involve direction resets, use what you have in new ways, tap unexpected skills.

42/ You'll notice that I've been using the word solve without necessarily associating it with the word problem. I've also avoided the noun solution.

43/ It's not grammatical, but think of solve as an intransitive verb that doesn't need a "problem" object. It also doesn't need a defined outcome/output: a "solution."

44/ Solving is about puzzling your way through life, applying a certain kind of creative, reality-anchored action orientation to life, where you drive change as you live.

45/ Constantly driving positive change by tapping an expanding mastery of the elegant workings of reality: that's solving. Y

45/ You don't need no stinkin' problems or solutions to be a solver. Just a drive to change what you see in interesting ways.

46/ I just found some hot chocolate in the kitchen, so I'm now solving for dry AND warm. Also, the sun is out now, and I'm wondering if I can solve for "dry faster" somehow.

47/ Happy Labor Day weekend. Hope you manage to solve for fun :)

1/ At a recent client meeting, I heard the phrase solve for X used a lot. It was a pleasant change from plan for X. As regular readers know, I have a bit of an aversion to planning and goals.

2/ The difference between the two phrases is huge. Plan is a managerial word. Its working material is solved problems. Solve is an engineering word. Its working material is untamed reality.

3/ Planning assumes that known means need to be arranged towards an end, where a "means" is a solved problem: a bit of reality understood as a known cause-effect map rather than as territory.

4/ Planning also assumes that ends are known, and the challenge mainly lies in choosing among available ends, by weighing "costs and benefits" and working out the "details."

5/ The assumption is that this process can be confined to a certain range of "operations" abstraction levels between “strategy” and “tactics": regimes meant for head-in-the-clouds bullshitters and lowly flunkies respectively.

6/ As some unimaginative military types like to say, amateurs worry about strategy and tactics, professionals talk about operations. This is not actually a flattering image of professionals.

7/ Professionals in this narrow sense are the sort who always deliver on time, as promised. They can do that because they never wander into regimes that require challenging “solves.”

8/ Solving domains, unlike planning domains, do not tolerate such "professionalism". If you are not comfortable being an "amateur" 2/3 of the time, you cannot solve. You can only plan.

9/ Planning can work with a purely instrumental view of the environment where everything has a clear purpose that it fulfills well, and can be understood in functionally fixed ways.

10/ As a result, planning in an everyday sense is mainly exercises in administrative logistics that pretend to offer predictable, time-bound outcomes because they avoid tough solves.

11/ This is why planners love middle-management bullshit like goals that satisfy SMART criteria:  if you can literally set such goals, you're probably not solving problems of any significance.

12/ At a purely mathematical level, even “solving” a “planning and scheduling” problem, something point-haired bosses think they are competent to do because they are “organized” administrative types, is non-trivial.

13/ "The literature on planning and scheduling in artificial intelligence generally takes it on faith that any interesting problem is at least NP-hard" as this widely cited paper says.

14/ Translated to layspeak (try this if the term NP is unfamiliar), this means unless you find a creative local exploit, most such problems cannot be solved in meaningfully time-bound ways. It'll get done when it gets done.

15/ Worse: often you cannot even solve such problems approximately in "SMART" ways: the approximate versions of many NP-hard problems are themselves NP-hard.

16/ And that's just uncertainty in time. There's plenty of other kinds of uncertainty in meaningful problems. And we haven't even touched on ambiguity yet.

17/ This means all the professionals who pride themselves on their on-time delivery and reliability are likely not solving real planning and scheduling problems.

18/ At best they are brute-forcing bad solves through some mix of bullying and indifference to what's actually delivered, that fulfill the letter, but violate the spirit of the challenge.

19/ At worst, they are creating an illusion of "execution" that does nothing ("going through the motions", "phoning it in"), through bad-faith theatrical displays of productivity and busyness.

20/ "Solving," unlike "planning" couples purposeful behavior and outcome selection via a reality loop. What you solve for depends on what you can do, and vice-versa. What you deliver is a discovery, not a promise.

21/ Solving does not promise more certainty and clarity in outputs that can actually be delivered given the uncertainty and ambiguity in inputs, and rate of intelligent solving progress.

22/ How does this work? Our habits and mental associations around the words planning and solving are set early, in educational institutions.

23/ As a kid or college student you probably argued at some point about how in math there's "always a right answer" and in the humanities "there are no right or wrong answers."

24/ Possibly, you picked one side or the other, and found some way to adopt a superior stance towards the other side. These attitudes set in early and hard.

25/ This schism is created by how the subjects are taught. You might solve the algebra problem x^2-4=0 with x=+2, -2 (that's still really just one answer, not two).

26/ You might answer an essay question like "Describe Lady Macbeth's character" with either a conventional Cliff-notes type answer, or a risky, creative, contrarian reading of the play.

27/ The two kinds of "solutions" are actually not that different. They just represent different mixes of play, skill, insight and objective setting (understood as temporary directions, not rigid goals).

28/ Real life, fortunately, is way more interesting than school or college mid-terms. "Solve for < 2 degrees global warming" is a challenge with both math-like and literature-like aspects.

29/ You get to choose what to solve for (solve for clean energy? solve for no-fossil fuels? solve for carbon sequestration? solve for refugee relocation?)

30/ You cannot afford to be a functionally fixed thinker. Are hybrid cars better than all-electric? Depends. Maybe hybrids delay the solve by creating a false sense of progress.

31/ Assumptions shift all the time, "$200/barrel oil will drive supply and demand to solve the problem naturally". Oops we invented fracking and oil is now under $50/barrel.

32/ Sometimes you wander really far off solving a seemingly unrelated problem. What engineers call yak-shaving. This can be a good thing.

33/ Why? Remember those clever mazes you solved as a kid where the right path wandered far , and the apparent straight line path dead-ended somewhere far from the exit?

34/ Or you end up with low self-awareness and mistake your lack of courage for a conviction that the color of the bike-shed matters in the design of the nuclear powerplant.

35/ Balaji Srinivasan coined the term idea maze for the context in which problem solving happens. I've come to prefer my updated term, an idea fog.

36/ Mazes are a great metaphor, and recall to mind the right kind of childhood problem-solving experiences to apply in adult life.

37/ They do suggest a designing playing field though. I prefer "fog" because it suggests uncertainty and ambiguity, and the possibility of creative path finding where no paths exist.

38/ It's a smooth rather than a striated metaphor for path-finding. One that suggests you can punch through walls and clamber through ventilation ducts.

39/ Where a planner views the world like a set of city streets to navigate with low map-territory divergence and no nasty surprises like extreme grades, a solver views the world like a national park.

40/ There are trail maps, but they tell you only a little of what you might need to know. Your hike can take unexpected turns, and involve setting new objectives like getting to an unmarked view point.

41/ You'll notice that I've been using the word solve without necessarily associating it with the word problem. I've also avoided the word solution.

42/ It's not grammatical, but think of solve as an intransitive verb that doesn't need a "problem object. It also doesn't need a defined outcome/output: a "solution."

43/ Solving is about puzzling your way through life, applying a certain kind of creative, reality-anchored action orientation to life,

44/ Figuring out how to master the workings of reality in elegant ways that give you leverage and power: that's solving. You don't need no stinkin' problems or solutions to be a solver.

_Feel free to forward this newsletter on email and share it via the social media buttons below. You can check out the archives here. First-timers can subscribe to the newsletter here. You can set up a phone call with me via my Clarity.fm profile page. _

Check out the 20 Breaking Smart Season 1 essays for the deeper context behind this newsletter. If you're interested in bringing the Season 1 workshop to your organization, get in touch. You can follow me on Twitter @vgr

Copyright © 2016 Ribbonfarm Consulting, LLC, All rights reserved.