Newsletter #92 — March 8, 2023
Guy Burgess and Heidi Burgess
This post explores a different aspect of the feedback processes that we started to examine in our earlier newsletter on Massively Circular Hyper-Polarization. It offers an answer to a simple, but all-important question: how can modern society be, at the same time, so unbelievably smart and so incredibly stupid? Or, put another way, why are some enormously challenging problems so routinely solved, while others remain so wickedly intractable?
The answer lies understanding the difference between the two principal types of feedback (we call them "sharp" and "fuzzy") that our environment uses to let us know whether the things that we have been doing are helpful and should be continued, or whether they are harmful and should be stopped or, at least, modified substantially. As is the case with virtually all important distinctions, real-world manifestations of these two kinds of feedback fall along a continuum from one extreme to the other.
The sharp and fuzzy distinction also provides a somewhat different and, we think, illuminating the way of looking at the distinction between complicated and complex systems. It also helps us understand why the problems associated with complex systems tend to be vastly more intractable than those associated with complicated systems. The reason is that complicated systems tend produce sharp feedback which is easily understood and acted upon, while complex systems tend to produce fuzzy feedback, which is much more difficult to interpret and learn from.
What makes societies so astonishingly smart is their ability to process sharp feedback effectively. By contrast, societies' inability to successfully process fuzzy feedback is a big part of what makes them act in such incredibly stupid ways. In the years ahead, the societies that learn how to handle fuzzy feedback most effectively are the ones that are most likely to succeed.
Sharp feedback is the principal driver of the self-interested behavior that underlies Adam Smith's "invisible hand" and, more generally, humanity's astonishing success. If we do something, and it turns out well, we add that behavior to our repertoire of things that we know how to do — things that we will repeat the next time the situation warrants. If things turn out badly, we try to figure out what we did wrong and modify our future actions accordingly. Sharp feedback, as we use the term, is shorthand for what could be more fully described as short-term (or immediate), clear, and focused feedback.
It is short term because it takes a relatively short time for us to find out whether the thing that we did actually produced the desired result. If our attempt to use an app on our phone fails, the results are quickly apparent, and we immediately try to figure out what we did wrong — a process that is repeated until we get it right and are able to add the ability to use that app to list of things we know how to do. Similarly, if we try something and the results really hurt, we may decide never to try to do that again.
The key thing about short-term feedback is that the consequences of our actions are quickly apparent — they are not so delayed that we are forced to make decisions about what to do next without really knowing whether what we did the last time worked or not. It also provides us with time for lots of repeated efforts, which gives us a chance to learn from many more failed attempts. This ability to try lots of things is a big part of what enables us to do hard things. As Edison famously said, "genius is one percent inspiration and ninety-nine percent perspiration."
Sharp feedback is also focused primarily on the people who actually did something (or the people that they care about). If we do something, we, for good or ill, bear the consequences of doing that. In general, focused feedback involves actions where the effects are local and primarily impact only a relatively small number of people. (It is, of course, true that there are often collateral impacts that may, over time, affect a much larger number of people. As we will see below, the need to account for these secondary and higher order impacts is part of what makes processing fuzzy feedback so difficult.)
Finally, sharp feedback is clear, not ambiguous. If something doesn't work, it is relatively easy to see that it doesn't work and why. This clarity gives us the information we need to make adjustments and keep trying new ways of doing something. This is because sharp feedback tends to be associated with relatively local actions where you can more easily observe what is actually happening.
Taken together, we call this all "sharp feedback" because mistakes are quickly and sharply felt and successes are sharply focused on the people who actually did something. This kind of feedback produces rapid and real learning. After all, positive and negative reinforcement has long been recognized as one of the principal ways in which all animals learn to navigate their environment.
It is the kind of feedback that has enabled humans to build the vast array of sophisticated tools that have given them the ability to flourish in such a wide range of physical environments. Sharp feedback has also enabled us to form communities based on cooperative social relationships, specialization, and a division of labor that has allowed us to vastly exceed our individual capabilities. This kind of decision-making environment has also allowed gigantic chains of relatively simple decisions to be strung together in ways that drive Adam Smith's "invisible hand" and the system of global capitalism with its ability to exchange an astonishing array of sophisticated goods and services.
The system has been successful enough to enable the human population to grow to almost 8 billion people while, at the same time, sharply raising standards of living — something that, in the 1970s, was thought to be absolutely impossible. In fact, free marketeers and libertarians often argue that this kind of system is all that is needed for society to self organize in ways that make sure that the needs of its citizens are successfully met.
From a conflict resolution perspective, this is a system based on the negotiation of mutually beneficial exchanges — exchanges between buyer and seller, employee and employer. These are exchanges that can often be negotiated by a relatively small numbers of people sitting around the proverbial table.
Unfortunately, sharp feedback can also lead us astray. There are lots of things that we do that feel good and are reinforced over the short term but, over the long term, get us into serious trouble. This is what makes the various types of addiction so dangerous. The self-interested behavior motivated by sharp back often also exposes society to a great many possible "tragedies of the commons" — the situation in which every individual making a decision that helps them in the short term, taken together with similar actions made by many other people, harms the larger community. Our ability to limit these problems depends upon our ability to learn from the second type of feedback.
At the other end of the feedback continuum is something very different and much more difficult to learn from — we call it fuzzy feedback. Along the three dimensions outlined above, fuzzy feedback is the opposite of sharp feedback. It is long-term, ambiguous, and diffuse. It also provides us with critically important information — information that allows us to determine whether the things that each of us are doing are, over the longer-term, making the complex social system in which we live better or worse.
We say that fuzzy feedback loops are long-term because it takes an extended period of time for consequences of any action to reverberate through the system and be manifested in the larger environment enough to have a recognizable impact. It takes years, for example, to find out whether specific educational reform efforts really do increase the skill level and marketability of the next generation of workers.
Obviously, we can't wait that long before deciding how to structure programs for the coming year. We have to act on the basis of informed predictions about the likely consequences of available options. This means that, if we are to make effective use of fuzzy feedback, we have to be able to recruit genuinely trustworthy experts who then objectively apply the most sophisticated predictive tools available.
We are also going to have to learn how to sensibly handle the inevitable uncertainties that surround even the best predictions. This means that we need to be prepared for the future to surprise us. We do not, however, as Kenneth Boulding used to say, need to be dumbfounded. We can predict the likely consequences of alternative courses of action with sufficient accuracy to greatly increase our chances of positive outcomes (usually by helping us better prepare for a wide range of possible contingencies).
To make effective use of these predictions we do, however, have to account for the fact that different individuals with different needs and interests evaluate the same outcome in very different ways. This requires that the technical experts making predictions focus on "what if" scenarios that predict the consequences of likely actions on different groups of people who care about different things. To use fuzzy feedback effectively, however, these predictions then need to be integrated into some equitable process for balancing competing community interests and deciding which outcomes (and therefore courses of action) are most desirable for the community as a whole.
Fuzzy feedback is also ambiguous. There are so many things going on, and so many people who, in their own way, are trying to make things better, at least for themselves or in their own view, that it is very hard to determine exactly what causes what. Everything that people do echoes through a system composed of primary, secondary, tertiary, and higher order interaction effects. The system is so vast that it is even hard to be clear about how, exactly, it is doing with respect to the many different things that we care about.
Risk reduction efforts are especially ambiguous. If we prepare for a disaster that never happens, it's hard to know whether our preparations would have helped. If we do something to reduce the risk of some terrible outcome, like war, it is hard to determine whether the risk was really reduced. The same is true for efforts to encourage positive things like technological innovations.
Still, the situation is not hopeless. There are lots of formal and informal research methods that, by using many different approaches, can provide us with valuable insights. Despite irreducible levels of uncertainty, these insights can help dramatically improve chances of getting a positive outcome (at least when compared with alternatives). Like long feedback, learning from ambiguous feedback requires the services of a wide array of genuinely trustworthy specialists with expertise in the many different aspects of any complex problem. While such experts can help us determine the likely outcomes of alternative courses of action, it is, again, up to the community, through a separate process, to balance competing interests and decide which course of action is most desirable.
Finally, fuzzy feedback is diffuse. It goes beyond and supplements the sharp feedback that is focused on the few people who actually did something (who, understandably, tend to evaluate their actions based primarily on whether or not those actions successfully advance their own narrow interests). By contrast, diffuse feedback is focused on the large numbers of people (sometimes millions) who are affected by secondary, tertiary, and higher order impacts of an action. While the ways in which an action impacts any single individual in this larger group tend to be relatively modest, the combined impacts on all individuals in this group can be quite substantial — often more consequential than the impacts on the few who were responsible for the action in the first place. What's more, it is common for sharp feedback to favor actions that are much less desirable when viewed from the larger perspective of diffuse, societal impacts and feedback. Conflict between these two perspectives is common — it is what people are talking about when they complain that selfish interests are profiting at the expense of the larger society.
The Critical Importance of Fuzzy Feedback
Fuzzy feedback and the critical importance that it plays in determining our life chances and the life chances of those that we care about is, perhaps, more easily understood by looking at some of the instances in which this kind of feedback is playing an especially important role. For example, the fact that we are dramatically altering the climate system is a fact that required extensive and very sophisticated analyses to deduce from large numbers of complex and often ambiguous observations. It will take an even more sophisticated array of efforts to develop and build support for an economically and politically viable strategy to effectively address the problem. (It is likely that the solution that emerges will differ dramatically from the current conventional wisdom.)
It will take a comparably sophisticated array of observations and analyses to understand and effectively respond to the rapidly changing threats associated with emerging infectious disease threats. The same is true for concerns about the security threat posed by geopolitical rivals, the economic implications of fiscal and monetary policies, or pretty much any other social problem.
Bottom line, without a robust ability to learn from fuzzy feedback, we will be powerless to navigate the complex array of 21st-century challenges.
Adapting Fuzzy Feedback Learning to the 21st Century
The good news is that the United States and virtually all other societies already have in place a complex array of institutions and supporting socio-cultural norms that help them balance societal and individual interests while also working collectively to respond to common dangers and take advantage of available opportunities. Unfortunately, this has historically been done in ways that tended to favor an elite few at the expense of the many. Today, our challenge is to adapt these time-tested approaches to an even more rapidly changing world — one that is increasingly intolerant of approaches that, in the past, produced such glaring inequities.
As was implied above, efforts to harness fuzzy feedback as an engine of social learning depend upon our ability to synthesize what we collectively know about the likely impacts of alternative courses of action. This requires constructively working through differences of opinion that commonly arise from different perspectives, lines of evidence, and methodologies. (To do this, we have to avoid the all too common trap of concluding that anyone who disagrees with us is automatically wrong.) It also requires us to use this information in ways that make decisions which fairly balance the conflicting interests of today's highly diverse citizenry.
Fortunately, we are heirs to a vast array of institutional structures that have, over the years, been established to help society process fuzzy feedback. These include, to name just a, the Departments of State and Defense, Environmental Protection Agency, the Centers for Disease Control, and the Federal Reserve, and the Congressional Budget Office. Despite their often obvious and considerable shortcomings, these and similar institutions have also played a major role in helping U.S. society surmount a wide range of difficult challenges. Contributing to the success of these efforts has been input (and sometimes pressure) from a wide array of interest groups established to represent particular clusters of diffuse interests. Also contributing have been the strategic planning efforts of many businesses that have tried to anticipate emerging societal needs and then develop products and services to meet those needs. Finally, there are research programs based at universities, laboratories, and think tanks that are specifically focused on gathering fuzzy feedback and developing proposed responses.
Like it or not, we are utterly dependent on the ability of these institutions to do their job effectively. Rather than demonizing them (as is now common) we need to support the efforts of the many who are trying to help them live up to their responsibilities.
The Conflict of Interest Crisis
Apart from the technical challenge of making sense out of the vast array of ambiguous information sources, the biggest obstacle to using fuzzy feedback is the inherent tension that exists between sharp and fuzzy feedback. In making day-to-day decisions about how we behave in our private and professional lives, we all have to prioritize sharp feedback. Failure to do so can easily leave us without a job or shunned by the community in which we live and upon which we depend.
Ideally, those in roles that require them to monitor and respond to fuzzy feedback on behalf of the larger society work in settings where their sharp feedback incentives have been structured in ways that reinforce, rather than compete with, their responsibilities to the larger community. (This, for example, is what conflict of interest rules are intended to achieve.)
What makes efforts to secure truly objective analyses of fuzzy feedback signals so difficult is the fact that effectively responding to this feedback commonly requires major re-allocations of societal resources and realignments of social policies (for example, toward wind and solar and away from fossil fuels or restructuring of global trading patterns). These are high-stakes changes that involve the enormous transfers of wealth and influence. As such, they are a tempting target for corrupt actors who seek to profit by distorting feedback in ways that drive public resources in their direction.
Unfortunately, there are a great many cases in which efforts to prevent policies from being distorted by those pursuing their narrow selfish interests have failed and, even more cases where people seeking to advance their narrow self-interest have been able to successfully attack and undermine the credibility of those who are honestly trying to represent the larger public interest. Of particular concern is what we call the "industrial complex effect." The name comes from Dwight Eisenhower's famous warning that the nation should "guard against the acquisition of unwarranted influence...by the military-industrial complex." Sadly, Eisenhower's warning that we must not allow entrenched bureaucracies to distort policies and self-serving ways applies to a wide range the policy areas (including those favored by the left as well as the right).
Also contributing to the problem is the widespread tendency the people from all political perspectives to shop around for expert opinions that justify their pre-determined policy preferences. This tendency is supported by a vast array of think tanks who specialize in providing authoritative reports that tell their constituents what they want to hear. The result is a cacophony of competing analyses that have degenerated to the point where people are increasingly abandoning the notion that objective analyses of societal problems is possible and concluding, instead, that anything that questions group orthodoxy must be wrong and should be suppressed as "disinformation" or "fake facts."This problem extends beyond the experts doing the primary analyses to include the media organizations that make the rest of us aware of the results of this work. It also underlies the notion that there is no "reality," but only perceptions of "reality." This then gives rise to the notion that any ideas are as good as any other: the notion that the climate isn't changing is as good as the notion that it is; the notion that COVID is just another "flu" is as good an observation as those who think it is more serious, etc. Objective reality exists and, if you don't pay attention to it, it can really hurt you.
If allowed to continue, this trend will decimate society's ability to learn from fuzzy feedback and effectively protect ourselves from threats like climate change, emerging pandemics, economic collapse, and war. Obviously, this will also fuel destructive hyper-polarization by undermining the very notion of the legitimacy and importance of good-faith disagreement and productive debate and replacing it with raw us-vs-them politics.
Taken together, the way in which these problems have undermined our collective ability to wisely and equitably process fuzzy feedback does much to explain why society acts in incredibly stupid ways. It also helps explain why, when society tries to make sound decisions, people think that they are incredibly stupid.
In Defense of Objectivity
This way of looking at things also makes it clear that we must fight the above problems by prioritizing the defense and strengthening of the norms and institutions needed to (as best we can) objectively process fuzzy feedback. We need to join in calling for a renewed focus on objective analysis (and reporting to the public the results of those analyses). In doing this, we should be clear that this implies the opening of academic, media, and cultural institutions to criticism and robust debate – something that has now largely disappeared in favor of the political orthodoxies that now dominate.
In other words, the key to making effective use of fuzzy feedback is pretty straightforward. We need to make more people aware of the importance of assuring objective analysis of common problems and then we need to devote our energies to supporting that kind of analysis and vigorously opposing those who would distort it to advance their narrow self interests.
The Importance of Flexibility
The other extremely important rule for dealing with fuzzy feedback is to maintain watchfulness and flexibility. With sharp feedback, you can try something, see what happens, and if your attempt fails, quickly try something else. With fuzzy feedback you won't know. So you need to try something, based on the best analysis available at the time, and then watch what happens. Results may take quite awhile to appear, and they might be obscured by other events. But if in six months, or a year, or even longer, it appears as if the outcome of your experiment might not be what you had hoped, then do an analysis of what might work better and try that. With fuzzy feedback we can't make a choice and go forward without looking back or second guessing ourselves. We need to, unfortunately, continue to reassess our decisions and be willing and ready to change course if our initial choices don't work out as expected. If we maintain vigilance, do our best to understand the complex system we are working in, and change course as needed, we still stand a pretty good chance of moving toward our goals successfully.
Please Contribute Your Ideas To This Discussion!
In order to prevent bots, spammers, and other malicious content, we are asking contributors to send their contributions to us directly. If your idea is short, with simple formatting, you can put it directly in the contact box. However, the contact form does not allow attachments. So if you are contributing a longer article, with formatting beyond simple paragraphs, just send us a note using the contact box, and we'll respond via an email to which you can reply with your attachment. This is a bit of a hassle, we know, but it has kept our site (and our inbox) clean. And if you are wondering, we do publish essays that disagree with or are critical of us. We want a robust exchange of views.
About the MBI Newsletters
Once a week or so, we, the BI Directors, share some thoughts, along with new posts from the Hyper-polarization Blog and and useful links from other sources. We used to put this all together in one newsletter which went out once or twice a week. We are now experimenting with breaking the Newsletter up into several shorter newsletters. Each Newsletter will be posted on BI, and sent out by email through Substack to subscribers. You can sign up to receive your copy here and find the latest newsletter here or on our BI Newsletter page, which also provides access to all the past newsletters, going back to 2017.
NOTE! If you signed up for this Newsletter and don't see it in your inbox, it might be going to one of your other emails folder (such as promotions, social, or spam). Check there or search for email@example.com and if you still can't find it, first go to our Substack help page, and if that doesn't help, please contact us.
If you like what you read here, please ....