Making Sense. The 7 Tiers of Sensemaking


Making Sense

The 7 Tiers of Sensemaking

Information Overload

It’s not just your imagination: understanding, reasoning about and making sense of the world is, in some ways, harder than ever.

Since the 15th century, a drumbeat of powerful technologies have emerged and shaped our world. In little over 600 years, we have progressed from feudal city-states to the printing press, the steam engine, the clipper ship, the incandescent lightbulb, the airplane, the telegraph, the radio, the Haber-Bosch process, television, the atomic bomb, the internet, artificial intelligence, gene manipulation, and automatons. In this time, the totality of human knowledge has grown exponentially.

This technological revolution has created civilization as we know it — with all the good, the bad and the ugly facets thereof.

It might seem a little strange to think about, but in a very real sense, human civilization is ascending. Gradually, it’s becoming bigger — something greater than its current form. Through the unending march of technological advances, we, in turn, assimilate subtle augmentations to our personhood and changes to the way we make sense of the world. Every day, the bounds of our identities expand just a little bit more to include that new smart watch, that new note-taking app, that new transportation service. Like it or not, we don’t get to go backward in this process.

Vannevar Bush with his differential analyzer, circa 1930 –

Vannevar Bush with his differential analyzer, circa 1930 –

“Consider a future device for individual use, which is a sort of mechanized private file and library. It needs a name, and, to coin one at random, ‘memex’ will do. A memex is a device in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility. It is an enlarged intimate supplement to his memory,” wrote Vannevar Bush in a 1945 Atlantic article titled “As We May Think.”

Bush headed the U.S. Office of Scientific Research and Development during World War II, and was instrumental in the creation of the National Science Foundation.

“There is a growing mountain of research. But there is increasing evidence that we are being bogged down today as specialization extends,” he continued. “The investigator is staggered by the findings and conclusions of thousands of other workers—conclusions which he cannot find time to grasp, much less to remember, as they appear. Yet specialization becomes increasingly necessary for progress, and the effort to bridge between disciplines is correspondingly superficial.”

In the aftermath of the bloodiest war in human history, Bush predicted a computing revolution, and a crisis. He made powerful points about our capacity for wisdom and understanding in the face of profoundly changing times. Many people think the internet is, or contains the manifestation of Bush’s memex, but I think we haven’t quite achieved the spirit of the idea. We have developed incredible machines for information storage and communication, no doubt. But most notably not for the synthesis of insights across disciplines, subject matters or tribes.

The Great Filter (Bubble)

As this world of ours gets smaller, so, too, is it getting ever more difficult to understand. How can we possibly make sense of it all? To deal with it, our biological and technological systems create filter bubbles for our perceptions. They exclude noise, and bring a narrow set of things into focus, thus helping us economize our efforts and our limited time.

This is a natural process. When I’m at a cocktail party, I would lose my mind if I tried to listen to everybody in the room at once. It’s just too much chatter. It’s genuinely useful to tune out the crowd as you and I discuss the finer points of our skullduggerous political situation. We lack the capacity to understand everyone at once, so we economize the information we process by tuning out the rest.

Unfortunately, in our globalized, hyper-connected world, we also tune out the profound suffering in Xinjiang, Bangladesh and Yemen in the same way as we walk right by that hungry vagrant on the street. We even tune out factual information that might be relevant to our lives because there’s so much of it. My reading list only grows; it never seems to shrink!

Many things in the world cease to exist to to us because they’re outside our respective filter bubbles. Mind you, I’m not suggesting you should feel too bad about this. Of course we must avail ourselves of opportunities to cross-pollinate, and broaden our experiences, but in some sense it’s a losing battle – Every day you have access to exponentially more information than you can process, with the same number of hours in the day, the same number of dollars in your pocket and the same capacity for rational thought. One could be forgiven for thinking, “OK, so what? I’m just surfing my own little wave here.”

Even if we accept our increasingly humble station in the grand scheme, there exist some serious game-theoretic problems that come about due to our decreased relative capacity for sensemaking. Coping with information overload and empathizing with our fellow human beings presents a huge adaptive challenge that stands to massively outpace our evolutionary processes. There are undoubtedly problematic social memes and other phenomena independent from our technology which we as a society must critically assess. Nonetheless, we would be remiss if we didn’t also take a hard look at the significant role technology plays here. We have been tinkering with the fabric of civilization on an unprecedented and expanding scale since the dawn of the industrial revolution. Can anyone honestly argue this effect has plateaued just now?

A Rubric For Sensemaking Systems

There are a lot of commercial and benevolent projects working to address these issues, with widely varying approaches and levels of sophistication. How shall we make sense of these sensemaking systems?

As I see it, we need a yardstick against which to measure them. Here’s mine: 

  1. Personal data – How do you remember things?

    Class 1 sensemaking systems generally entail basic data entry and organization of ideas, tasks, citations, etc., and their relationships. People are clamoring for better ways to manage their personal data. Among to-do lists, messages, reading lists or other notes, there’s a huge demand for personal-data-management systems. Most of the existing ones are okay, but not great. One example of this could be be a notes or to-do app on your phone. Lest we assume this classification system applies only to new technologies, arguably your trusty old moleskine is a Class 1 sensemaking system. You can record, organize and recall your personal data there.

  2. Group data Getting the word out

    Class 2 sensemaking systems enable propagation of data within a group. In larger groups beyond approximately 10-100 people, only a minority of participants are meaningfully able to express themselves on a given topic. That is to say, a low ratio of producers vs. consumers. Such systems might be new and sophisticated, with real-time collaborative features like Google Docs, or more basic collaborative features like a Wiki page. They can even come in more traditional forms, like CNN and The New York Times – hundreds of producers; but a viewership and readership in the millions. Whatever the flavor, a key factor in Class 2 systems is that all but a small subset of participants are prevented from contributing, at least in a practical sense. This might seem like a thoroughly solved problem. But many product offerings are springing up to better serve this market.

  3. Group insights – Getting smarter with friends.

    Group-insight systems expand on Class 2 or group-data systems to include third-party data sourcing and somewhat more sophisticated semantic representations of their data. A major goal of which is to apply basic machine-learning techniques to help users better synthesize new insights. Such systems can be useful to organizations looking to make better decisions with data, thus increasing their efficiency. These insights will prove useful. Still, Class 3 systems suffer many of the same limitations as Class 2 systems in terms of practical limitations on the number of participants able to contribute at any given time. There are a few market entrants beginning to work here. For most of them, it’s fairly early in their development lifecycle.

  4. Discourse substrates – Making sense with strangers.

    Contrasted with the previous classes of sensemaking systems in which a majority of participants are necessarily passive due to practical considerations, now we are beginning to enable a broader base of active contribution. The key challenge of such systems: Organizing and consolidating multiple inputs into digestible output. Such systems must be resistant not only to cognitive biases and populistic effects but also to Sybil and other kinds of attacks. They might have rudimentary filtration based on tone/quality to keep things factual and focused, or sophisticated moderation strategies. Such systems might be useful to better organize and civilize existing topics of discourse. But, they will always top out with a relatively small number of overall participants before the cost of the moderation becomes too great to keep up. If done well, these systems might help improve society to a limited degree. There are some projects already working on this, but not many. One example is Kialo.

  5. Crowd-wisdom substrates – A Ouija board for your thoughts.

    Class 5 sensemaking systems are where things start to get really interesting. These systems will exhibit limited emergent phenomena resulting in group synthesis of new insights that none of the individual participants would be capable of seeing of their own accord. It is, in effect, the realization of the memex and the beginning of a whole new level of sensemaking. It will require a much more advanced system of semantic representations based on analogical reasoning rather than the coarse grained knowledge representations that much of the industry uses today. I will talk about some of these design elements in a subsequent post.

  6. Strong AI and you – These are the droids we’re looking for.

    AI as we know it today is only intelligent in the narrowest sense. We train these predictive models with huge volumes of data — far more than we would use to train a baby how to walk or a teen how to drive. Today’s AI systems are impressive in many ways, but that power is superficial. We neither understand them nor commune with them in any meaningful way. Class 6 sensemaking systems will begin to offload meaningful expressions of individual opinions, thoughts and feelings. They will gradually begin to serve as extensions of our personhood and our intellect. The main advantage of such systems is not their potential to help all of us achieve super-genius level intelligence, but rather to better understand and interface with the rest of our collective than we could ever hope to achieve alone. We might not become geniuses individually, but we can be smarter together.

  7. Instruments of human ascendence – We’re not in Kansas anymore, Toto.

    Class 7 sensemaking systems are where things get really weird. Such systems would constitute a much stronger and accelerated warping of the lines between humanity and machine, so much so that humanity will become unrecognizable — and perhaps something else entirely. It’s hard to appreciate the levels of sophistication we might achieve, rather like an ant trying to comprehend the entire contents of the internet. It’s a scary and weird thought, indeed, and it might take hundreds or thousands of years — provided we don’t destroy ourselves first.

Suffice to say, Classes 5, 6 and 7 each constitute huge categorical leaps, and present significant room for debate in their premises. I hope you’ll agree however that they are useful to think about, if for no other reason than to begin to systematically compare existing and future systems.

Over Soul -

Over Soul -

Some might understandably view these as an entirely negative scenarios: merging with AI and all the “interesting times” implications thereof. And they might not be wrong either. That said, elements thereof might provide a path toward a more universal exercise of empathy, which could be a huge boon when it comes to making better decisions at a civilization scale. Whether it’s climate change, war, or wealth inequality, we as a civilization are pretty bad at sensemaking. In turn, we’re pretty bad at making good decisions about the wellbeing of the whole.

At a planetary scale, we suffer greatly from the externalization of cost and the tragedy of the commons. Rather than continuing with our ineffectual and baroque interventions against these tendencies, we should consider what it would take to do better. Might it be possible to create a post-incentive economy in which we make decisions based on care rather than self-interest? Obviously, it would be desirable for members of our civilization to make better choices, but we can’t make them do it. Authoritarianism definitely isn’t the answer.

The really interesting alternative is to strive to create instruments of universal empathy. When we can manage to feel our neighbor’s joy and pain in some fashion, we might be capable of making meaningful inroads toward a more just civilization — one that might be capable of being self-sustaining rather than self-defeating.