Jump to the video Down to contents

Most of us sense something's wrong with social media. A lot of people also say something's wrong with “capitalism”. (I'm using air quotes around capitalism, I’ll be more specific later.) And that something’s wrong with modern politics, liberalism, democracy, education.

People come at these problems from different directions—from climate change, from worries about “coordination failure”, geopolitics...

I came at it through social media.

I started out as an internet optimist. See, in 2008. I worked at Couchsurfing, which was growing fast. So was Meetup.com and Wikipedia. Flashmobs were booming! Anyone remember Improv Everywhere?

It seemed—to me and to others—that the internet was bringing us a better economy. People were working together, on giant projects like Couchsurfing, Wikipedia, and Linux. I was a true believer: Love would be the new motivator. Money would fade away.

So… that was naive. By 2012, I’d sobered up. By then, Facebook’s News Feed and YouTube had not only replaced TV, but had added hours of screen-time per person per day. The “attention economy”. In alarm, with Tristan Harris and others, I cofounded the Center for Humane Tech.

Actually, it wasn’t called “the Center for Humane Tech” at first— it was called “Time Well Spent”. This was part of my proposal to fix the problem. At CouchSurfing, we tried to maximize the amount of meaningful time our users spent with each other, rather than any measure of transactions. Tristan and I thought that, if more people did this—if they maximized “Time Well Spent” instead of “Time Spent”—it might fix social media.

That was still naive..I was young. I thought the problems around me were, technological.

But it goes a lot deeper than that.

Jump to the video

Then, the next three chapters say what needs to change, to bring back togetherness and meaning.

All right, let's get started.

Jump to the video Back to contents

A saying in business is “you make what you measure”. What it means is, as businesses grow, they have to summarize information into metrics or reports to make decisions. The CEO gets summarized information like “how many sales did we make in Europe?” or “how many people returned our new product and asked for their money back?”

The larger the business, the more things need to be summarized, like this.

For that reason, the particular numbers that get measured very important.

We want to make spaces. Good spaces. Meaningful space. If we want to do this in a scalable way, we need to be able to measure meaning.

This, it turns out, is currently impossible. That’s why things tend to get meaningless as they scale. This is one way that business methods are biased towards funnels and tubes.

Say you run a bar. You want it to be a good space for people to relax, and connect with friends.

So, you try to measure whether the bar is a good space for that.

The situation gets even worse, if it becomes a network of bars. If you have to run it like people run startups or cities, using a metrics dashboard.

How can you monitor, remotely which bars are good spaces?

Here’s an idea:

People often smile when they’re relaxed and connecting with friends. You could put up cameras, get a machine learning algorithm doing sentiment analysis, and measure smiles per hour. But the most efficient way to ramp smiles is a funnel that gets people smiling. The local franchise might bring in a comedian to up their score.

Does that make it better for connecting with friends?

The situation gets even worse when bars in the network compete with one another for smiles.

The core problem here is that “being a good space for people to relax, to connect with friends” is vague. Because it’s vague, we end up finding clearer proxy measures like smiles, or beers, or monthly active visitors. Those new measures are clear, but they don’t correspond exactly to what we really care about. And these proxy measures, since they don’t line up exactly with “being a good space for people to relax, to connect with friends”, end up being manipulable.

Actually, it’s easier to explain this as two kinds of manipulation:

So we have behavioral manipulation on one hand; psychological manipulation, on the other.

The fact that we call these things manipulation, I believe, indicates that there is something deeper that we really care about, something deeper, that these manipulations pull us away from.

And that that deeper thing must not be behavioral. And it also cannot be a desire, goal, or feeling, since these are manipulable too.

You can also do this with the people around you—talk to them about their feelings, and what's meaningful. You’ll start to glimpse what I call “the universe of meaning”. The space of all values. Everyone's sources of meaning, together.

A key fact, is the collection of all values is smaller than some other collections—like the collection of all goals, or the collection of all preferences. Let me show you what I mean:

Consider the collection of all goals. It includes many far-off goals. Goals like “get rich like Elon Musk”. Or goals people only have because they hope they’ll lead to something good later. Goals like “impress so-and-so at the bar.”

Values aren’t like that. Values cards refer to things you’ve actually paid attention to in choices, and which you found meaning to pay attention to. You’ve already had the experience.

Far-off goals may relate to our values. Maybe we think accomplishing them will bring us closer to our values. For instance, that if you got rich you could finally spend time playing music.

Goals may also provide a context for values. Someone who finds meaning in visual creativity, might set a goal to publish a weekly comic with a friend.

The goal is something they haven’t done yet, and there are infinitely many of them. But the value is something they already know.

Preferences also vastly outnumber values, but for a different reason.

Actually, for two reasons.

First, there are many preferences that are about pleasure but not about about life meaning. Perhaps you loooove chocolate cake, but a life filled with chocolate cake wouldn’t be a meaningful one for you.

Second, you’ll have many preferences that are related to your values, but only connected via a long train of inferences. Maybe you love walking in forests, the way the forest light opens your heart, and you think that, if Trump is elected, the forests will be cut down. So, you prefer Biden to Trump. Your Biden preference is connected to your walking-in-forests heart-open source of meaning, but so are a million other preferences.

So, we can save a ton of time by collecting a person’s sources of meaning instead of their goals or preferences. We cut right to what’s precious and immediate, for them, about being alive. The basic ways of living, choosing, and attending, which they find meaningful. The things they need spaces for.

Jump to the video Back to contents

So, we are surrounded by funnel and tube entrepreneur and designers.

I believe it’s fundamentally different to be a spacemaker.

Let's say you want to make a retreat for scientists. You want your retreat to be meaningful for this value: deep work.

Now, this is one of my values. Working for many years to make something—something which may not even be successful in the end.

The middle of the values card covers what people attend to, when they live by this value. Let’s take a look. They attend to

Now, once you’ve picked a value like this, and a context, like making a retreat, there's three types of thinking you can do, to make your retreat meaningful on purpose, for the value that you picked.

First, you can analyze your retreat out into funnels, tubes and spaces.

Let’s follow an individual participant through the retreat.

Once you have a diagram like this one, you can be much clearer about what you’re making.

So, just dividing things clearly like this can help a lot. It’s especially good to match spaces and sources of meaning. Once you do this, your participants can help you shape each space. You can ask them: is the space good for attending to, and choosing by, what’s on the values card? If not (you can ask) how could it be better?

That’s the heart of the design method I teach.

But there’s more you can do.

People can’t do what’s meaningful to them all the time. They have to balance things that are deeply meaningful with other things that are important.

In general, you want to ask: what kinds of goals and norms might crowd out the sources of meaning?

So, that’s two kinds of thinking you can do, to make something meaningful on purpose. Break things into funnels, tubes and spaces. Look for goals and norms that could crowd out values in the spaces.

Using these two kinds of thinking, people can get pretty far. But the real win is from a third kind of thinking. To show it, I have to go deeper into what a value is.

Look closely at this values card, at the middle part. You can see that living by a value doesn’t happen at one moment. Instead, it’s spread out over time. You probably found the curiosities before you found the research methods or colleagues, for instance.

Meaning is not something we experience in an instant—meaning is a realization, an assessment that we have, when we look at our lives, that we were able to attend to what was important to us. Meaning is many moments of choice, woven together through time.

And at each of these points in time, you had to attend to and choose certain things. Some of that attention is meaningful — that’s what goes on the values card. But at each point in time, there are other kinds of, less meaningful attention and choices to make.

I like to use a grid to picture this. In the grid, time goes from left to right. And vertically, we have the different states of attention and choice. In order to attend to something, you need options to attend to, information about each option, and some kind of skill at discerning how to choose. In order to choose something, you often need to finesse an environment or setup to make the choice possible, then take some kind of bold action.

All this, is what it takes to do deep work. The stuff we put on the values card is the meaningful stuff to attend to and choose by, but behind every meaningful moment is a rich backstory.

Now, any of these things might be hard to do. But if you can’t do them, you won’t be able to do deep work.

For that reason, we call them “hard steps”.

Now, if somebody would find meaning in deep work, but they're not doing deep work, it could be that any of these hard steps were blockers for them.

This is one of two reasons that someone might not be able to do deep work.

The mark of a good space, is that—first, goals and expectations don’t get in the way of the value—and second, a good space makes the hard steps easier to take.

Let’s think again about the deep work retreat we want to organize.

These ideas may seem obvious—but my experience is that they’re only obvious once you’ve thought through the hard steps!

And here’s the best part: once you figure out the hard steps for a source of meaning, they can be applied across contexts.

These same hard steps—the ones that improved our retreat—can be used to improve anything for deep work!

To prove this, I’ll use them to improve Twitter.

This may sound like a joke: Twitter’s short posts and feeds are kind of the opposite of deep work. But it’s not entirely a joke: I actually have found deep work colleagues on Twitter.

And I think Twitter could be better for that.

Here I’ve highlighted some of the relevant hard steps.

So, imagine you’re Andy Matuschak, and you’re using Twitter. You see one of your own tweets, but is has this extra box at the bottom asking—“Hey, has you're thinking evolved on this topic?”

You think, “yeah, it has” so you click the “capture” button and choose some earlier tweets on the same topic. These tweets are precursors to the more recent one. Together, they show a timeline of how your thinking has evolved.

You also pick some twitter users who helped you evolve your thinking, maybe based on who replied to the earlier tweets. Maybe you also type something, about how your thinking changed, and post it!

If Twitter had this, you’d see these “evolution threads” as you scrolled.

I think it would help with all these things:

If twitter had that feature, probably more deep work would happen. Because it addresses some of the hard steps of connecting with colleagues and communities for deep work.

So in this chapter, I've walked through three ways you can make things meaningful, on purpose.

I've made them into a kind of “bento box”, which can guide you, as you design.

If you use the bento box as you design, you’ll make a space where people can live by their source of meaning. A meaningful space, for the value you picked.

Many people are using this bento box to make more deeply meaningful things.

And there are new stories like this all the time.

These people are making spacemaking into a craft. They are spacemakers.

I believe spacemakers will soon be quite a prestigious group.

Just as neurosurgeons need to take great care while operating on their patient’s brains, spacemakers take a similar care with what gives us meaning in life.

Those who can be trusted to do this—to keep things meaningful—will be honored. They will be put in positions of great responsibility. They will be trusted to repair the institutions that have been destroyed by funnel thinking—trusted to fix the problems with research, with democracy, with education, and so on.

I call forth the spacemakers. It’s your time to shine.

Jump to the video Back to contents

So, onto the last chapter!

With values cards, in chapter 2, I showed how we can be more intentional about meaning in our own lives. With hard steps in chapter 3, I showed how we can make meaningful spaces for others.

Imagine if many individuals articulated their values, and many spacemakers made beautiful spaces for them. Imagine a kind of a renaissance of spaces.

Such a renaissance would only go so far, because large-scale systems—like social media, markets, and recommender systems—act by piling up strangers. And because they suffer from what I called “recommender disdain”—the problem where social media recommenders don’t recommend spaces as much we’d benefit from them.

So, at large scales, funnel and tube entrepreneurs will remain more successful than space makers.

That's not my goal. I don't want space makers to be niche. I want them to be the most successful entrepreneurs in the world. I want the latent demand for meaning and togetherness to be completely unlocked.

So, in this last chapter, I want to sketch out an alternative vision for those large scale systems—for social media, for markets, and for recommenders. I want to show that they could be deeply changed.

I should say up front that I don’t have a plan to actually get them changed. To do that, would, I think, require a vast cultural shift. Like, if it became common sense that sources of meaning are what we should organize our lives around, and if the public was demanding a reorganization of economic, political, and technological systems to suit this. Demanding to return spaces to their rightful place.

I know, that might be wishful thinking. But if it were true, we’d have a question, which is, what should those new systems look like. In this chapter, I want to sketch an answer. At least the beginning of one.

My hope is just to convince you that such systems could be made.

That’s all pretty exciting. But I think we can do even better. To get into it, let’s go back to chapter 1, when I talked about piling up strangers.

What we need is something in the middle.

I cannot stress enough how important this is.

So many structures in our society pile up strangers:

If we can come up with an alternative structure—that means all of these things get reinvented. A profound societal shift.

So let’s see what we can do.

When I have a problem like this, I often try thinking of something that’s not scalable, but meets the other criteria.

So, we need a social structure where there's lasting relationships, deep possibilities are surfaced, and the roles of creator and consumer are blurred. But it also should be open to strangers, and surface excellence.

Can you think of something like that?

There's more than one answer, but here’s my favorite one.

A breakdance circle.

Also called a “cypher”, this just a circle of breakdancers where one at a time someone goes into the circle and shows off, and everyone cheers them on.

Thinking about what makes a cypher work, I think it's three things.

We can compare the collaborative production in a cypher, with what’d happen if these people were just dancing with their closest friends.

So cyphers have some of the advantages we’re looking for.

But at cyphers, the “consumption” isn’t separate from creation. Everyone’s there together, with the same group mission, and deep possibilities are discovered and pursued: new aesthetic values, new styles, new ways of relating, new ways of jamming.

I’ll show another example. Something more scalable.

Science! Science has a lot in common with breakdance. Look at writing a scientific paper. It travels through a sequence of spaces, each of them is a bit like a breakdance cypher. There are shared values, non-transactional relationships, and a group mission.

This as a series of groups. which cover a continuum between production and consumption. Everyone kind of does both, but there’s more production on the left side and more consumption on the right.

As something moves from group to group it's produced, refined, discussed, assessed, annotated, etc. My favorite word for this is—it’s being legitimated. By the time it gets all the way to the right it’s a full-fledged, legitimate, published, scientific paper.

It gets this way by accumulating little labels. Labels like grounded claims! and field redefining!. They are assessments of quality, and they have to do with the sources of meaning of scientists.

This whole system scales in an interesting way. It scales in three directions.

Also like breakdance,

I want to call this kind of structure — where people in a series of spaces collaborate to get something co-produced, refined, annotated, and discussed — where production and consumption are mingled and social — and where work, proceeding forwards, accumulates these kinds of values-aligned social assessments of excellence. I want to call that structure a space train. Because when I zoom out, to me it looks a bit like a train. A train made of spaces.

And also because. Well. SPACE TRAIN.

Space trains are a third path, between friends-only structures and piling up strangers. In space trains, you’ve got lasting relationships, deep possibilities get surfaced, and roles stay flexible. Plus, they’re open to strangers, excellence-surfacing, and scalable.

If Sandra, William and Margaret, our spacemakers from chapter 1, had scaled their operations inside space trains, they could have all these advantages.

Sandra likes to throw events. She’d be part of a collective of event designers, and would refine her event structure, together with others in that group. When her event is ready for prime time, it wouldn't be Sandra alone who signs off on it. The other members of her group would make some assertions about her event. They’d attest that it creates a vulnerable environment, or it opens up creativity, or that it’s a good space for localism and food. These tags would be based on what Sandra's event refining collective specializes in. The tags would propel the event structure or description forward into another space of values-aligned event curators. Maybe one of them would come to experience Sandra’s event in person. They’d discuss it, provide some more tags, more related to their specialty. That would propel the event into friend groups who like to experience these kinds of events together. They’d become aware of the event structure, and could attend Sandra's event together.

Note:

We can imagine something similar with William’s poetry—using a train of tags and of spaces of poetry appreciation, his poems will find their way to whomever’s hearts they will set a fire.

Space trains can be used in many places.

Social media like TikTok has the piles of strangers, creator and fans type of structure.

Let’s redesign it.

First, I’m going to make it clear that I’m in a group. I do the standard up and down swipe to swipe videos. And I can swipe left and right to change groups.

Here, I’m in this “Solace and Wisdom” group, together with this guy, John Green.

In this group, we stitch-respond to people having a hard time.

Before posting, we pass around our drafts. John’s just passed us a draft of his. He talks to a woman who's crying, about the Emily Dickinson poem “Hope”.

I love how John’s done it. So, I select some tags we use in the group to mark out the kinds of excellence we look for in solace and wisdom.

Alternatively, I could have made some suggestions for John, to help him achieve the excellences we’re going for.

But he knocked it out of the park on his first try.

Those tags will make the video appear in other groups. It might even be seen by the woman he stitched with.

OK, now I'm caught up on solace and wisdom.

I want to post something. Let me move the post button over here so we can tell it posts into the group.

Now I can swipe right, to see another group. This is the programming languages and interfaces group, where we demo new programming languages and interface ideas. Oh, Adam Wiggins has something new!

I’ll watch that later. I've got something to share.

I hope the other members will refine it with me! Or tag it with the excellences that we care about, here in programming languages and interfaces.

That’s my TikTok redesign. In it, I focused on two of my interests. But the same tagging and sharing interface could work for dance videos, comedy, etc. So long as each group can sync up on values.

I doubt TikTok will make a big change like this. But someone will.

All I want to show here, right now, is that it’s possible to make a social media structure that doesn’t pile up strangers. That creates space trains instead.

Stories pass through groups and accumulate labels. Consumption also takes place in groups. In fact, the distinction between creation and consumption is blurred, just like in a breakdance circle.

So you might be thinking:

If space trains are so great, why aren’t they already the default way that things get done? Why is the default structure of social media, markets, and recommenders one that piles up strangers?

I think there are three main reasons.

First, space trains require that the group inside each space, that they sync up on values. On which types of excellence they want to co-produce, and curate.

Historically, this has been quite hard. It takes communication and introspection. And I said in chapter 1, we’ve lacked a common vocabulary with which to do it.

My hope, is that new technology—like values cards, from chapter 2, and like large language models—can make it much easier to sync up on values.

The second reason, is that space trains are made of spaces. Recently, spaces have been hard to make, so space trains have also been hard to make.

Maybe the bento box, from chapter 3, can help. It made it easier to make that “deep work” science retreat. It can also make it easier to make the spaces in a space train.

Take, collaborators, working on a paper. Ideally, this would be a space for deep work. We can use the middle part of the bento box and look for goals that crowd out deep work.

So, one thing we’d want in the space train is alternate funding structures, institutional roles, or career trajectories, to reduce these pressures

Ok, let’s look at the hard steps.

One can imagine addressing both the crowding out and these hard steps together. What if prestigious awards or positions could come just from asking questions or proposing methods. That could be an alternative to racking up citations, or to pursuing safe, incremental publications.

These changes would be hard to make across all of science, but they’d be pretty easy to attach to one collaborator group, as software infrastructure. And if it was common knowledge that deep work was important for science, and these hard steps were blocking it, there’d be a reason to scale that software, attach funding to it, and so on.

So, that’s two reasons why space trains haven’t already become the default. And with both of them, I see some reason to hope they’ll change.

The third reason I’m a little more stuck on, but I don’t think it’s insurmountable.

So, that’s space trains. Space trains are my deeper proposal to repair markets, recommenders, and social media. Currently, they see you mostly alone, mostly interacting with funnels and tubes, and they’re unlikely to learn your sources of meaning. I think space trains would change that.

Space trains are, I hope, also how we can fix democracy, education, science, and art.

In each of those categories, certain values or sources of meaning make the whole system work. One of those values, in science, is deep work. Big advances in science usually involve deep work.

Science has gotten worse for deep work. And for other important values of scientists. A similar decay happened in education, in democracy, and art. In each case, we can draw out the field as a space train, then make each space meaningful for those key values.

That makes things more meaningful for scientists, for politicians, for artists, for educators.

But it also makes those fields more productive.

What we're talking about, is a massive transformation of nearly every aspect of society. We're talking about billions of people changing the vocabulary they use, their ways of relating. We're talking about the redesign of institutions all the way from local businesses to nation states, around meaning and values.

So, this is not a thing that a small group can do.

But every change starts with a small group.

What could a small group do?

Okay, so now we’re entering the realm of strategy. Strategy is hard! Without an absolutely inspired strategy, this project is doomed.

To defend an idealistic movement against forces like these is really, really hard.

I can’t tell you, yet, how we’ll do it. But there are some plans in the works, and we’re gathering the best strategic thinkers in the world to make those plans better.

Here, in the last few minutes of this talk, I just want to give a little taste of those plans.

So, phase 1 involves building prototypes, and a community with a common vision. This talk begins to sketch that vision. But much more is needed:

To sketch these things out, we’ll need conferences, meetups, online forums, full of advocates and researchers, working to rebuild society on meaning.

And there’s prototyping to be done:

The next phase is one of rapid expansion.

We want to help a million people identify their sources of meaning. To discover which kinds of meaning they’re yearning for. I believe this can be an incredible experience, and that it will unlock tremendous demand. Once people understand what makes their life meaningful, they’ll want to drop the bullshit and focus on the meaning.

By unlocking that demand, we can link those million people into an ecosystem to help them do just that.

What would such an ecosystem include?

So, meaning profiles, and an ecosystem. You can view this as a technical shift. But you can also view it as a cultural transformation.

This is a new mythical structure for society. A new story to tell ourselves, about who we are, where we belong, and how we can get along.

My colleague, Ellie Hain, is all about articulating that new myth. She has a talk about that you can watch on our website, rebuildingmeaning.org.

At some point, as this expands, we're going to run into conflicts with business as usual:

Here, that new social vision will be important. The more people take to it, the harder it is to fake. Also important will be robust metrics and auditing structures, so people can see where the real meaning economy is happening, and where it's being faked. Report cards—like the one I showed in chapter four.

By then, we’ll have in-roads in academia, continuing work I started here to redesign systems like voting and markets around meaning. There will be renowned professors, think tanks, and policy people.

To support this, we will have amassed some economic power. There’s already a school for training spacemakers—the school for social design, and an open source textbook. The school can partner with bootcamps, accelerators and funds. There can be incubators for space-based startups, and a kind of meaning mafia.

This will eventually translate to a broad social and political movement. But… such a movement will need to be more robust against co-option than, say, effective altruism or social justice has been. How can we do that? We don’t have a concrete plan, but there are some unusual techniques for managing status and power in the community, that we’re excited to try.

I have just sketched a hard road ahead. But I believe it’s the hot path to restoring social trust, repairing democracy, science, and education, reducing hustle and ideological warfare, to creating corporations and recommenders and other structures that people can trust with their lives and data. And to creating a society that has less trauma, less manipulation, more meaning and togetherness.

But… can we do it?

You might be the difference between this happening or not.

We need to rapidly build a support network, advisory network, and team.

Here’s two ways you can help: