Wicked problem(en.wikipedia.org)
en.wikipedia.org
Wicked problem
https://en.wikipedia.org/wiki/Wicked_problem
37 comments
Absolutely fantastic article.
The protagonists, in my experience, don't understand this nuance very well, and are quick to label you as "anti-science".
The protagonists, in my experience, don't understand this nuance very well, and are quick to label you as "anti-science".
How is the AIDS pandemic an example of a wicked problem?
One of the definitions of a wicked problem is "Wicked problems have no stopping rule.". But the stopping rule for the AIDS pandemic seems to be creating a cure for AIDS / eliminating all cases of AIDS.
Am I missing something really obvious here??
One of the definitions of a wicked problem is "Wicked problems have no stopping rule.". But the stopping rule for the AIDS pandemic seems to be creating a cure for AIDS / eliminating all cases of AIDS.
Am I missing something really obvious here??
If you are operating within the scope of a single country, eliminating all cases within that country would not be a stopping rule.
But that itself doesn't make it a wicked problem. It means the solution needs to involve broadening the scope of your operations. In case of AIDS, it would be an international cooperation.
A real-life example is the efforts to eradicate smallpox and polio; the first one succeeded, the second one was almost done but hit some serious problems with international cooperation - regardless, this demonstrates the problem, while hard, isn't wicked.
A real-life example is the efforts to eradicate smallpox and polio; the first one succeeded, the second one was almost done but hit some serious problems with international cooperation - regardless, this demonstrates the problem, while hard, isn't wicked.
Oh, I meant eliminating all cases on earth :P
Has recently been popularized by the book Range, for which I'm grateful, because now I can talk about this stuff with my family.
There is also something of a "wicked problem of wicked problems," which is that wicked problems are very hard to talk about---one of their survival traits is that they're unappealing to talk about, because by definition you can't "solve" them easily.
There is also something of a "wicked problem of wicked problems," which is that wicked problems are very hard to talk about---one of their survival traits is that they're unappealing to talk about, because by definition you can't "solve" them easily.
I work in the domain of these wicked problems, and they usually reduce to Taleb's no "skin in the game." The necessary condition for one is that someone has power from some exogenous source, but without accountability.
You can in fact identify wicked problems as an epi-phenomenon of the strategies of the players, who give away their intent in early rounds. The only solutions are an equilibrium of interests.
It's a condition of bureaucracies, politics, and anything outside markets and hierarchies where you can resolve things with a clear price or authority. They are a peculiar game form that has no winners.
You can in fact identify wicked problems as an epi-phenomenon of the strategies of the players, who give away their intent in early rounds. The only solutions are an equilibrium of interests.
It's a condition of bureaucracies, politics, and anything outside markets and hierarchies where you can resolve things with a clear price or authority. They are a peculiar game form that has no winners.
The fact that a problem is a wicked problem does not justify inaction, in my opinion.
I’ve had (stupid) friends argue that because income inequality is a wicked problem, we should do away with a social safety net for the poor.
Obviously we can’t this issue 100 per cent, but the difference in trying to solve it is better quality of life, lessened crime, happier populace, more people ready for work, and lower mortality rate.
It’s great to identify something as a Wicked problem, but that should never stop you from attempting to solving it.
I’ve had (stupid) friends argue that because income inequality is a wicked problem, we should do away with a social safety net for the poor.
Obviously we can’t this issue 100 per cent, but the difference in trying to solve it is better quality of life, lessened crime, happier populace, more people ready for work, and lower mortality rate.
It’s great to identify something as a Wicked problem, but that should never stop you from attempting to solving it.
Everyone dies and we can't prevent that so we should do away with hospitals. Just bring up some issue that affects them and suddenly they will be disagreeing.
Then fine. Go with the Buddhist standard, and the goal is to reduce suffering.
I would say that's one of those simple rules with extraordinarily complex, but overall very good outcomes.
I would say that's one of those simple rules with extraordinarily complex, but overall very good outcomes.
In Boston, people call this type of of problem a "wicked bad problem".
I'm really happy this topic is surfancing on HN as I've made it my mission to build experiments to try to tackle social wicked problems.
Dinnertable Chat is an online chatroulette-style conference matchmaking system based on mixing people based on same and different viewpoints. The same idea could be used to mix people on any category (by skill, field, etc): https://www.dinnertable.chat/about
The Mix Opinions service matches people in physical spaces by dividing people up into groups, such that each group has maximized viewpoint diversity of each of the members. I could also adapt this to be applied to online conferencing and maximize on both diversity as well as interests (based on weights). https://www.mixopinions.com/
Some of the philosophy of why I was focusing on 'mixed viewpoint' conference software: https://docs.google.com/presentation/d/1Q5KWhxhBMLWSvrYoEoWP...
Dinnertable Chat is an online chatroulette-style conference matchmaking system based on mixing people based on same and different viewpoints. The same idea could be used to mix people on any category (by skill, field, etc): https://www.dinnertable.chat/about
The Mix Opinions service matches people in physical spaces by dividing people up into groups, such that each group has maximized viewpoint diversity of each of the members. I could also adapt this to be applied to online conferencing and maximize on both diversity as well as interests (based on weights). https://www.mixopinions.com/
Some of the philosophy of why I was focusing on 'mixed viewpoint' conference software: https://docs.google.com/presentation/d/1Q5KWhxhBMLWSvrYoEoWP...
Just this week I was reading “Notes on the Synthesis of Form” by Christopher Alexander (CA), which in someways kickstarted the “design thinking” craze. Later on, CA recanted many of his ideas about the systematization of design, and converged with Rittel regarding “wicked problems”. That said, the first half of the above book is still a fantastic read on why wicked problems tend to be hard when attempted in “self conscious mode”. One of the most thoughtful pieces I’ve read in a while... and it’s an intuitive, breezy and relatively short (for a book) read.
Here’s Rittel’s seminal paper on problems with the “planning/design“ approach towards wicked problems: https://archive.epa.gov/reg3esd1/data/web/pdf/rittel%2bwebbe...
Here’s Rittel’s seminal paper on problems with the “planning/design“ approach towards wicked problems: https://archive.epa.gov/reg3esd1/data/web/pdf/rittel%2bwebbe...
It's a great example because of a totally stupid name, a really important concept isn't being adequately explained or understood.
Meaningless garbage like "Black Swan" is spouted all the time and sets people back, just because it's a cool name with a cool story
Yet Wicked problems are our most important issues at Micro and Macro levels, you will have this problem, but people can't tackle them properly because we haven't abstracted the framework in a cool way.
I also think the definition needs to be tightened. But I honestly think no one wants to work in a meta field called 'Wicked problems'
I wonder if in other languages this is tackled better?
Meaningless garbage like "Black Swan" is spouted all the time and sets people back, just because it's a cool name with a cool story
Yet Wicked problems are our most important issues at Micro and Macro levels, you will have this problem, but people can't tackle them properly because we haven't abstracted the framework in a cool way.
I also think the definition needs to be tightened. But I honestly think no one wants to work in a meta field called 'Wicked problems'
I wonder if in other languages this is tackled better?
I swore these were called "hard problems" based on the rough definition: it's a problem whose solution only becomes obvious after a failed attempt at finding the solution.
This fits nicely with the PDF text post also currently on the front page. The best solution for exporting the output text is simply to OCR what is visually seen as an image even though the text is present in the data (because white text on a white background wouldn't be wanted by most users) Therin: The solution also changes based on use case!
This fits nicely with the PDF text post also currently on the front page. The best solution for exporting the output text is simply to OCR what is visually seen as an image even though the text is present in the data (because white text on a white background wouldn't be wanted by most users) Therin: The solution also changes based on use case!
> it's a problem whose solution only becomes obvious after a failed attempt
I think if the solution ever become obvious it's not a wicked problem.
You can solve them sometimes, but they involve large amounts of waste and time. And you never know how to have done them better.
That's my understanding.
My starting solution to the PDF is to give the user both OCR and text and let them chose. I don't think it's Wicked, but I do think no one is solving it properly along clearly viable lines.
I also think there is a clear solution to force companies to convert PDF's to text properly. So even at a higher level I don't think it's Wicked, just the culture is not correct.
I think if the solution ever become obvious it's not a wicked problem.
You can solve them sometimes, but they involve large amounts of waste and time. And you never know how to have done them better.
That's my understanding.
My starting solution to the PDF is to give the user both OCR and text and let them chose. I don't think it's Wicked, but I do think no one is solving it properly along clearly viable lines.
I also think there is a clear solution to force companies to convert PDF's to text properly. So even at a higher level I don't think it's Wicked, just the culture is not correct.
>But I honestly think no one wants to work in a meta field called 'Wicked problems'
I would love to work on this, but one of the things that makes wicked problems wicked is that their solutions are often incredibly unpopular.
"Ill news is an ill guest, Gandalf Stormcrow."
I would love to work on this, but one of the things that makes wicked problems wicked is that their solutions are often incredibly unpopular.
"Ill news is an ill guest, Gandalf Stormcrow."
This seemed like an interesting concept, but the list of examples was underwhelming. It can be boiled down to "problems that can be mitigated, but not solved." Which is just kind of a dull thing to name. Given the description, it sounded like this was an article about paradoxical problems in optimization, or sort of real-life examples of the liars paradox. Instead, it's a class of problems that may be interesting in their own right, but the part they have in common isn't.
Wickedness in a problem domain is an argument for a competitive, decentralized approach, in order to more rapidly search a larger solution space. Yet instead we seem to demand gordian or manichean solutions and strong centralized institutions to implement them. When the trains don't run on time, instead of seeking out a fray of Dagny Taggarts, we look for one Mussolini. And so wicked (hard) social problems beget disproportionately wicked (counter productive) answers.
You are framing this as an either-or situation but, if you read the wiki page, there are 3 strategies to dealing with wicked problems and you missed the "collaborative" category.
It's clear the authoritative "mussolini" solution is undesirable from a liberal perspective in which we prefer to have a shared say over how we organise things. However, it's also increasingly clear that the way we implement the competitive strategy -- you so aptly suggest the Ayn Rand protagonist as the better option -- is also flawed and incapable of solving these difficult problems. The nature of modern capitalism is to tend toward monopolistic/oligopolistic dynamics in the market, which ultimately leads to a concentration of power and a lack of shared control over such issues. Our economy has only ever made our impact on the environment worse and worse, and is driving us ever closer to catastrophic climate change. Not to mention other examples, such as healthcare, etc.
Why not try something "collaborative", where instead of hoping for either a top-down governmental, or top-down corporate solution to problems, we truly decentralise power and give all people agency and freedom to live in this world? For instance, if we stopped letting the profit motive squeeze us for every possible thing, people would have more free time with lower living expenses and less demanding jobs, and could spend more time repairing/restoring possessions, rather than buying new, or gardening and cooking, rather than eating out, or cycling everywhere, rather than taking cars.
It's clear the authoritative "mussolini" solution is undesirable from a liberal perspective in which we prefer to have a shared say over how we organise things. However, it's also increasingly clear that the way we implement the competitive strategy -- you so aptly suggest the Ayn Rand protagonist as the better option -- is also flawed and incapable of solving these difficult problems. The nature of modern capitalism is to tend toward monopolistic/oligopolistic dynamics in the market, which ultimately leads to a concentration of power and a lack of shared control over such issues. Our economy has only ever made our impact on the environment worse and worse, and is driving us ever closer to catastrophic climate change. Not to mention other examples, such as healthcare, etc.
Why not try something "collaborative", where instead of hoping for either a top-down governmental, or top-down corporate solution to problems, we truly decentralise power and give all people agency and freedom to live in this world? For instance, if we stopped letting the profit motive squeeze us for every possible thing, people would have more free time with lower living expenses and less demanding jobs, and could spend more time repairing/restoring possessions, rather than buying new, or gardening and cooking, rather than eating out, or cycling everywhere, rather than taking cars.
Because private property is the sin qua non of decentralized decision making. It is in fact a bundle of rights to make your own decisions about the disposition of things. Anything short of it is also less decentralized. Collaboration with less competition has fewer independent decision makers, and therefore searches the solution space more slowly.
Creative destruction is ugly as hell. It incorporates all of evil. But it is also the engine of adaptation; the inverse of lithification. To the extent that we limit it, we limit the parallelism of adaptation. The fact that businesses can be so easily destroyed outweighs all of their manifest ugliness. Very few noncompetitive institutions have that feature. Fast search means rapid destruction and creation.
There is a very real human cost to that, and many, maybe most people will never be convinced that it's worth it.
Creative destruction is ugly as hell. It incorporates all of evil. But it is also the engine of adaptation; the inverse of lithification. To the extent that we limit it, we limit the parallelism of adaptation. The fact that businesses can be so easily destroyed outweighs all of their manifest ugliness. Very few noncompetitive institutions have that feature. Fast search means rapid destruction and creation.
There is a very real human cost to that, and many, maybe most people will never be convinced that it's worth it.
Talking about minutiae of capitalism here is sort of detracting from the core issue (that also underlies the problems of capitalism).
The core problem is: humans can't coordinate at scale. Working together towards a shared goal comes naturally to us when we're in small, tightly-knit groups. The larger the group, the harder it gets. Once the group reaches ~hundred members[0], coordination generally collapses, and the group succumbs to the usual breed of prisonner-dilemma's-like problems. Or, it develops a governance structure.
If you look at history of civilization through the lens of coordination, you'll notice that all the governance structures, all the economies, up to and including democracy and the free market, are essentially a mechanism of coordinating larger and larger amounts of people. A government cuts trough prisonner-dilemma's-like problems. Its hierarchical nature lets direct coordination costs scale as O(n log n) instead of O(n^2) - but that's still not efficient enough for smaller and more localized problems. There, markets come to the rescue: they let people coordinate on smaller and more local issues, often without even explicitly noticing them - all gets handled by the feedback loops around prices.
The core observation here being: the market works at efficiently coordinating groups of people with diverse interests and smaller, more localized problems. It absolutely sucks at responding to a problem requiring to coordinate large groups of people. That's where you want the top-down solution. It's reasonable to look up to the government to solve a coordination problem below its level[1], that's what governments are for.
--
(And you do need both a government and a market to limit each other's power. For instance, you mention that "businesses can be so easily destroyed". That only holds true in a spherical-cow free market that never existed. One of the features of capitalism is that profit compounds; a successful company has more resources to protect itself from its competition, ensuring further success. A successful company can't be easily destroyed.
The end-goal of competing businesses is monopoly, but society needs the competition, not the victory - so you need the government to ensure the victors get struck down and competition continues.)
--
[0] - Dunbar number is ~150, which may or may not be a coincident. https://en.wikipedia.org/wiki/Dunbar%27s_number
[1] - Governments have their own coordination problems at the international level; geopolitics is essentially a completely unregulated free market.
The core problem is: humans can't coordinate at scale. Working together towards a shared goal comes naturally to us when we're in small, tightly-knit groups. The larger the group, the harder it gets. Once the group reaches ~hundred members[0], coordination generally collapses, and the group succumbs to the usual breed of prisonner-dilemma's-like problems. Or, it develops a governance structure.
If you look at history of civilization through the lens of coordination, you'll notice that all the governance structures, all the economies, up to and including democracy and the free market, are essentially a mechanism of coordinating larger and larger amounts of people. A government cuts trough prisonner-dilemma's-like problems. Its hierarchical nature lets direct coordination costs scale as O(n log n) instead of O(n^2) - but that's still not efficient enough for smaller and more localized problems. There, markets come to the rescue: they let people coordinate on smaller and more local issues, often without even explicitly noticing them - all gets handled by the feedback loops around prices.
The core observation here being: the market works at efficiently coordinating groups of people with diverse interests and smaller, more localized problems. It absolutely sucks at responding to a problem requiring to coordinate large groups of people. That's where you want the top-down solution. It's reasonable to look up to the government to solve a coordination problem below its level[1], that's what governments are for.
--
(And you do need both a government and a market to limit each other's power. For instance, you mention that "businesses can be so easily destroyed". That only holds true in a spherical-cow free market that never existed. One of the features of capitalism is that profit compounds; a successful company has more resources to protect itself from its competition, ensuring further success. A successful company can't be easily destroyed.
The end-goal of competing businesses is monopoly, but society needs the competition, not the victory - so you need the government to ensure the victors get struck down and competition continues.)
--
[0] - Dunbar number is ~150, which may or may not be a coincident. https://en.wikipedia.org/wiki/Dunbar%27s_number
[1] - Governments have their own coordination problems at the international level; geopolitics is essentially a completely unregulated free market.
Is telling whether something is a wicked problem a wicked problem?
Based on this Wikipedia article, yes and no. It has multiple conflicting and vague definitions for whatever the hell a "wicked problem" is. It's like they read the definition of REST and thought "I think we can go vaguer".
Totally irrelevant, but when you put the string "wicked problem" together twice like that, my brain refuses to scan it as anything other than a problem with a wick.
Something related to semantic satiation, I think.
Something related to semantic satiation, I think.
No or maybe.
Just like a bloom filter
Just like a bloom filter
I imagine in many cases, if you solved it, you wouldn't really get credit and history would just sort of shrug and forget it was ever an issue.
I really get irritated by the use of the term. I've heard it too much in the wrong situations. Use with caution.
As a person who grew up in Boston, it is a really confusing term due to just sounding like a generally hard problem... which appears to be a pretty decent approximation of the term. The examples given in this article provide both clarity and additional confusion with a focus on being large uncertain (and thus generally hard to solve) problems.
As I have gotten older I try to recognize these when they present themselves, and proceed to avoid them at all costs.
He said: "Many of the issues that lie at the interface between science and politics involve questions that can be stated in scientific terms but that are in principle beyond the proficiency of science to answer."
"The protagonists often ask for the impossible: scientific answers to questions that are trans-scientific"
Pretty good stuff.
[1] https://science.sciencemag.org/content/sci/177/4045/211.full...