12 tiny worlds

Support for the show comes from the Audible original The Downloaded 2, Ghosts in the Machine. Earth's final days are near, and the clock is ticking for members of the Phoenix colony.

All they need to do is upload their minds back into the quantum computer.

Simple enough, but their plan is disrupted by digital duplicates of their own minds that could destroy their stored consciousness forever.

Listen to Oscar winner Brendan Fraser reprise his role as Roscoe Caduleon in this follow-up to the Audible hit sci-fi thriller, The Downloaded, by Canada's most decorated sci-fi author, Robert J.

Sawyer. What are you willing to lose to save the ones you love? Find out in the thought-provoking sequel, The Downloaded 2, Ghosts in the Machine, available now only from Audible.

With a Spark Cash Plus card from Capital One, you earn unlimited 2% cash back on every purchase. And you get big purchasing power.
So your business can spend more and earn more.

Capital One, what's in your wallet? Find out more at capital1.com slash spark cash plus terms apply.

Imagine for a moment that we have the ability to rewind time.

That we're able to press a button and slide back through the last centuries worth of history

through millennia of music and culture,

back through tens and then hundreds of thousands of years of human existence. Imagine that we keep going.
We go back through the age of the dinosaurs

and the age of the giant insects.

All the way back to around 500 million years ago, when our ancestors, ancestors, ancestors were crawling around in the muck along with other animals.

Creatures like Hallucigenia, which does in fact look like something out of a hallucination, full of strange spikes and weird legs.

Or Opabinia, which has five eyes and a trunk-like appendage with a claw at the end. These are animals that only exist as fossils now because they don't seem to have direct descendants today.

Other animals, like the ones that became our ancestors, they won the evolutionary contest and spread far and wide across the globe instead.

But if we did travel back to that time, back to when all these animals coexisted in the oceans, and if we started over from that point,

would we end up with the same world that we have today?

Or could things evolve completely differently?

Evolution is a process of random mutations interacting with chance happenings, right? So what if if some random roll of the mutational dice went differently?

What if a mudslide wiped out the ancestors of some animals that were key to life as we now know it?

Could some other option, some hallucogenia or opabinia or something else, could they have won the evolutionary lottery and spread far and wide instead of our ancestors?

This is, basically, the question that a paleontologist named Stephen Jay Gould asked in his book, Wonderful Life, back in 1989.

In that book, he argued that the world as we know it is the result of a series of amazing accidents. And if you re-ran things, you might get some very different results.

Other scientists disagreed. One pointed out that over and over and over again, life has gravitated towards certain solutions to being on this planet.

Stuff like eyes, for example, have developed many times in otherwise very different animals.

And so it's not unreasonable to think that things maybe wouldn't look exactly the same, but that there are some strategies that are just like a good idea if you want to thrive on Earth.

And given enough time, natural selection will eventually kind of steer life towards those options. All evolutionary roads lead to a kind of evolutionary roam.

Now, as Gold acknowledged in his book, There's no way to definitively settle this question.

We do not have a time machine or the ability to watch multiple planet Earths unfold across hundreds of millions of years to see if they all end up in the same place.

But what we do have is an experiment that started even before Gold's book came out. This experiment that kind of has the potential to maybe not answer this question,

but at least explore it.

It is an attempt to make something like 12 tiny parallel universes and do a sort of speed run of evolution.

So, this is Unexplainable. I'm Bird Pinkerton.
And today on the show, the Long-Term Evolution Experiment.

The first seeds of the long-term evolution experiment started forming decades ago in the mind of one man. I'm Richard Lenski, and I'm a professor of microbiology here at Michigan State University.

Now, long before Richard became a professor, he was a young scientist, less focused on tiny microbes.

Microbiology seemed kind of complicated, and I was more interested in big picture sorts of questions.

Questions about how evolution worked, how populations changed over time, how quickly or how slowly they changed.

And he was especially interested in a question that was kind of similar to the one that Stephen Jay Gold ended up asking. Basically, how predictable is evolution?

I think it's a really interesting question, which is

to examine the tension between randomness and predictability in evolution. How does that play out?

As he progressed through his career, he started to wonder if he might be able to explore these big questions.

I got really interested in sort of the idea that you might be able to watch evolution as an ongoing process. The problem was that watching evolution at work is very hard to do.

Because environments in nature are constantly changing. There's not just two species, but hundreds of species potentially interacting.
So seeing it as an ongoing process is very challenging.

Richard, in fact, experienced firsthand how challenging it could be.

For my my PhD, I worked at the University of North Carolina on beetle populations in forests and in clear-cut areas, trying to understand the competition between species and the growth of populations and that sort of thing.

And the project went pretty well, but it was also very frustrating to me. The weather didn't always cooperate.
The beetles didn't always cooperate. And it was a lot of work.

What does an uncooperative beetle look like? Well, among other things,

they can squirt some nasty stuff at you, at least these particular ones.

So you have to handle them a little bit gingerly.

If Richard wanted to get answers to his questions about evolution, he realized that he needed to strip things down to get rid of the literal mess of squirting beetles, but also the figurative messiness of them.

all the other things that they might encounter during their little beetle lives, like bad weather or good weather or lots of food or not enough of it.

Richard needed to create an experiment so simple, basically, that he had a realistic chance of keeping track of everything in it, of watching evolution play out without distractions.

I wanted to strip evolution down to its bare bones. And to do that, he moved away from beetles, away from animals entirely, and he landed instead on bacteria, and specifically E.
coli. E.

coli coli is the lab rat of bacteria. Many people have sort of a visceral reaction to E.
coli saying, oh, that's a nasty bug, don't get near me.

But in fact, every one of us, every human on the planet has roughly a billion E. coli in their lower digestive tract.
Most strains of E. coli are not harmful.

In general, it's just a little bacteria with a big sweet tooth, right? It gobbles up sugar and then just kind of minds its own business.

And it's very easy to work with in a lab, not just because it's not squirting stuff at you, but also because it's easy to look after and it reproduces really fast.

It can go through six generations in a day, which would be like if you could produce great, great, great, great grandchildren in 24 hours.

That's pretty perfect for someone who's trying to figure out how things evolve over time. So it was pretty perfect for Richard.

He couldn't go back and see how the Earth might have evolved differently if you re-ran it in different ways, but he could create 12 almost identical situations and then watch each of them play out, like making 12 parallel universes if those universes were in laboratory flasks.

Little flasks, little tiny flasks that hold up to 50 milliliters. They're only a couple of inches tall, little glass flasks that have little beakers sitting on top of them.

And so 37 years ago, Richard took 12 little piles of E. coli, each of which had grown from a single cell.
And on February 24th, 1988, he put one pile into each of 12 flasks.

And inside each flask, there was a mixture of water and sugar and some other ingredients that you can think of kind of like a broth or a soup, I guess.

It was a tweaked version of a historical recipe that lots of other microbiologists had used to raise E. coli.

And this broth mixture, again, was pretty much identical in each flask, or as close to identical as possible. And then after the E.

coli were added, each flask was put into an incubator, this big box kept at 37 degrees Celsius that shakes very slightly to keep everything kind of moving around a bit.

And when the flasks are sitting in the incubator with their little glass beakers on top, they make a very delightful little jingling sound, very quiet to the ear. Too bad I don't have any here.

I could shake one for the recording,

but it's a very pleasant sound. Inside each flask on that first day in 1988, the E.
coli basically did their thing.

They ate glucose, they reproduced, they reproduced some more, and by the next day, they'd eaten the glucose in their flasks.

So Richard took 12 new flasks, put fresh broth in each of them, and then took some of the E. coli from the first flask and transferred it carefully into a new one.

Then he did the same with the second flask. And you do that for all the populations.
It takes a few minutes, once you're good at it at least, and you pop those back into the incubator.

And there, the E. coli continued eating and reproducing and eating and reproducing some more.
And the next day, he transferred the E. coli again and again.

Eventually, other people in his lab were doing the transfers too, but it was always the same routine. Each day they would transfer the E.
coli to new flasks, and then the next day they do it again,

and the next day, again, and again, and again, and the next day, and again, again, and again, again, and again, again, and the next day.

Weekends and holidays, and as close to 365 days a year as possible for first weeks, and then months, and then years.

And with each day and each round of reproduction, the populations in each flask started to peel away from each other.

Because in each flask population, even though everything else was almost the same, different random mutations were happening, tiny changes to the E. coli genes.

Some of those mutations did not help the E. coli, maybe even harm them.
But some mutations were helpful, and those E. coli were more likely to reproduce and pass their mutations on.

And so Even though the E. coli in each flask started with an almost identical ancestor and close to identical and close to identical conditions.

As they mutated, they started to differ from each other just a little tiny bit.

So this is how Richard and his colleagues set up essentially a miniature model of evolution, these 12 tiny dolls' house worlds full of bacteria.

But in order to figure out how those worlds were changing over time, they also needed a record, right? Snapshots of the dolls' houses across the generations.

So they also made a doll's house version of a fossil record.

Every 75 days, after they moved the small fraction of E. coli from last day's flasks into a new one, they took some of the leftover E.
coli from the old flasks and they froze it.

Because E. coli have one more superpower that makes them especially good research subjects.
We can freeze the bacteria and they are still alive.

You can think of them like they're in suspended animation almost. So if you thaw them out again down the line, they come right back to life.

And that allows us to do what I like to call time travel. We can directly compare strains of our bacteria that lived at different times in the past.

A paleontologist who's studying the normal fossil record, they have to look at stuff like old skeletons, right? Those skeletons are often incomplete, sometimes just single bones.

They're creatures we don't have any bones from at all. And fossil bones are, obviously, fossilized, so they've turned to rock.
All to say, our fossil record is a puzzle with a lot of pieces missing.

But in Richard's lab, the record of E. coli past is beautifully thorough.
Yeah, we call it the frozen fossil record.

So not only do we have images of the cells, not only do we have their DNA preserved, but we can actually bring them back to life.

And so, as I said, we can compare things that lived at different times, directly compete them head to head. Well, I guess bacteria don't have heads, but cell to cell.

If they froze bacteria one day, then 75 days later, or 150 days later, or 3,000 days later, they could pull out that first frozen sample.

thaw out those bacteria, and then compare them to current bacteria.

It'd be a little bit like if we could thaw out cryogenically frozen but somehow living humans from 10,000 years ago or something, right?

And then see what would happen if we compared ourselves to them directly.

What if we played football with them? What if we played chess? Or what if we had to hunt for food?

Who would win in a hot dog eating contest? Me or my 500th great grandmother? In Richard's lab, they can meaningfully ask a version of that question.

It's just, who would win in a a glucose seeding contest, this bacterium or its 500th great ancestor?

And doing work like this gave them some really fascinating insight into some of Richard's questions about how evolution actually works.

Over the years, they've been able to track how beneficial mutations spread through a population, and they even saw that some E.

coli seem to increase their mutation rate, as though to increase their odds of hitting on a lucky mutation that could help them out.

It is all really fascinating work. But a few years in, after the E.

coli evolved and evolved and evolved some more, Richard and his colleagues were also able to start answering the big question that they'd had about what the effects of random chance mutations might be.

They could compare the 12 parallel populations, the doll's house universes, if you will, and see

Did all the bacteria take sort of different evolutionary routes to wind up in essentially the same place or not? When he set up the experiment, Richard assumed that the answer was going to be not.

I was imagining that if I set up these replicate populations in the laboratory under identical conditions and with the exact same ancestor, I was thinking they'd kind of all go off in very different directions.

And what we found, a little bit to my consternation, was that all roads seemed to lead to pretty similar improvements in fitness.

Basically, the common solution seemed to be to eat as much glucose as you can and as fast as you can and get really big, fast-growing cells.

With DNA analysis tools, Richard and various collaborators could peek under the hood of the E. coli cells too.

And there, they also saw that the bacteria in the different populations were taking different approaches, and there were definitely some significant differences between them, but they were overall landing in pretty similar places.

And moreover, not only were they doing things remarkably similar, not identical, but it was really hard to sort of see the differences.

They were also slowing down.

It seemed like all roads led to maybe different suburbs of Rome, but still Rome-ish.

Now, that doesn't mean that Stephen Jay Gold was wrong to say that the past matters and things that happened millions of years ago shape the future, right?

Our Earth is a lot more complex than a little box of flasks.

It contains a lot of different species all interacting with each other, not just one species of E. coli.

And it also changes over time. In Gold's thought experiment, part of what created different paths for different universes was random chance stuff beyond just mutations, right?

Like a lake drying up or an asteroid hitting the Earth. That stuff is not happening in this experiment.

Still, by about 15 years in, it did seem like at this small, super simple scale, the question was somewhat settled.

I kind of imagined that the experiment had run its course, and I even talked to some colleagues.

You know,

maybe we should end this experiment.

And then one of the 12 flash populations stopped looking like all the other populations. It stopped going to Rome.
One of them went on a trans-Pacific flight and did something completely different.

That's after the break.

Support for the show comes from one password. If you don't work in IT security, you might have a Hollywood vision of what it looks like.

This lone IT professional who probably looks like mid-90s Sandra Bullock, feverishly clickety-clacking away, trying to stop the hacker from creating a backdoor to the mainframe.

Or maybe you've got a Bollywood vision, which is basically the same thing, but with better songs. The reality, though, is way more complicated.

Between keeping devices and applications safe and dealing with shadow IT and securely onboarding and offboarding employees, your IT team is doing tons of work every day.

And you can help them with their real security needs with Trileka by OnePassword. Trileka by OnePassword combines a complete solution for SaaS access governance.

And it's just one of the ways that extended access management helps teams strengthen compliance and security.

OnePassword's award-winning password manager is trusted by millions of users and more than 150,000 businesses from IBM to Slack.

So take the first step to better security for your team by securing credentials and protecting every application, even unmanaged shadow IT. Learn more at onepassword.com/slash unexplainable.

That's the number one password.com slash

unexplainable.

With the Spark Cash Plus card from Capital One, you can earn unlimited 2% cash back on every purchase and you get big purchasing power. So your business can spend more and earn more.

Stephen Brandon and Bruno, the business owners of SandCloud, reinvested their 2% cash back to help build their retail presence. Now, that's serious business.

What could the Spark Cash Plus card from Capital One do for your business? Capital One, what's in your wallet? Find out more at capital1.com/slash SparkCash Plus. Terms apply.

Corporate credit cards, cash flow management, centralizing spending, help with catching fraud.

A powerful corporate card program needs to do a lot, so it should be designed to help work with your business.

That's why with the American Express corporate program, you can customize key elements to match your company's needs. Explore and apply from a variety of card types based on employee roles.

Select liability and billing options that fit your needs. Add spending controls to support compliance.

And decide whether earning rewards or cash back in the form of statement credit delivers more value for your business.

Learn more and start customizing your corporate program at americanexpress.com/slash corporate. Terms apply.

Richard actually almost missed one of the most exciting discoveries of the long-term evolution experiment. Yeah, it's kind of a complicated story and one that I almost blew.

And history gives us do-overs in some cases. It all started one day in January of 2003.

A researcher in Richard's lab was going through the usual process of pulling out the 12 flasks to transfer the E. coli in them into new flasks when he noticed something odd.

So

the person who

first noticed this increased turbidity was or increased cloudiness of the culture

was a postdoc.

And this happened to be on a weekend. So I was just looking back at the lab notebook.
I have it right here beside me, in fact. So it was a weekend.
and on Saturday, he said, hmm, something seems off.

One of these cultures is more turbid than it should be. Cloudiness like this, it means that so much stuff is growing in the flask that it started to block out the light.
E.

coli multiply as they eat glucose and reproduce, but there wasn't enough glucose in the flask for E.

coli to grow to light-blocking numbers, which suggested that there might be a contaminant in the flask, right? Or that something had gone wrong somehow. And on Sunday, same thing.

It was still too turbid. So on Monday, the transfers were being done by Nirja Hajila, a former technician in my lab, who actually has done more transfers than anybody else.

She worked with me for over 20 years. And Nirja showed it to me and I said, something's wrong.
We got a contaminant in there. Let's go back into the freezer and restart that population.

Basically, they did the equivalent of going to your last save place in a video game and starting again.

They went back into the freezer, grabbed their last frozen population, and that let them start from a freezer save point, essentially, and continue on. And do I remember that?

Only vaguely. What I remember much more

is,

so the first discovery of the turbidity, the cloudiness, was in January. And then in April, the same population that had been restarted by going back into the freezer suddenly got very cloudy again.

And I said, whoa.

And I started doing things like looking more closely at the colonies, doing literally a little bit of a smell test, how they smelled and how they looked on plates, and checking some genetic markers.

We knew certain things that would tell us these were really E. coli doing this.
His analysis of the bacteria that were in this cloudy flask ultimately showed that they were, in fact, E.

coli and descendants of that first ancestor. There were just a lot more of them than there should be.
And the explanation had to do with the broth that Richard had been putting in these flasks.

This broth, again, was basically just a recipe that Richard had pulled from other historical experiments on E. coli, right? A mixture of water with some glucose.

But there was another thing that I left in the recipe, which is something called citrate, which is a compound that gives lemons and oranges and citrus fruits their sort of tart, sour taste.

And that was in the recipe for historical reasons. And I just left it in.
I had sort of played around over the years with getting rid of it, but I just left it in.

But I didn't reduce the amount of it. So there was a lot of citrate, but it didn't really matter because E.
coli can't grow on citrate. At least, not in these conditions.
In these conditions?

It eats the glucose and then goes to bed every night until we transfer it, not knowing that there's this delicious lemony substance, this citrate, this lemony dessert that was sitting there to be had.

And for 15 years, none of them figured it out. And then suddenly one of the populations figured it out.
This one flask full of E. coli had figured out a way to eat this lemony citrate stuff.

That's not totally unheard of. There are other examples in the world of E.
coli that have developed the ability to eat citrate.

But Richard and his colleagues were really interested to see that even though there was a lot of citrate around in this broth to eat, and even though the E.

coli had had over a decade to evolve and had gone through many thousands of generations of evolution at this point, only one of the flask populations of E. coli had made this leap.

It seemed strange that, if it was possible, it would have taken so long and only happened one time.

Something weird was going on,

and we wanted to get to the bottom of that weirdness. Which is kind of an exciting task to have as a bunch of scientists.
I'm still really excited about it. So, yeah, yay.

Yeah, actually, I still get

almost silly with excitement whenever I give talks about this, which is unusual because I'm actually a very shy person, but I don't seem that way when I'm talking about this.

Zach Blount was not actually there for the discovery of the lemony citrate eating behavior in the C. coli because he joined Richard Lensky's lab as a PhD student a little later on.

But through a series of lucky breaks, he wound up directly involved in studying the citrate eaters. And the entire project fell into my lab.

Zach wound up spending the next few years of his life trying to figure out what exactly had happened with these E. coli, how they had developed the ability to eat this lemony citrate stuff.

And to do that, he went back into the freezer to unfreeze some of the past generations of E. coli to then see if he could get them to do this again, essentially.

It was a lot of waiting and worrying, is this going to work? There's actually a picture of him sitting next to more than 13,000 little trays that he used to grow just a whole lot of E.

coli, which he then kept careful track of in a bunch of notebooks and spreadsheets. waiting to see if the E.
coli would re-evolve the ability to eat this lemony dessert.

I'd have to look back my calculation, but I think I went through something like 10 trillion cells and found 13 independent re-evolutions of the citrate-eucine bacteria.

At what point did you finally feel like, okay,

this is real? I did, I like, this is real. I'm going to have a pretty good PhD on my hands.

Well, I'm a worry wart, so I kept thinking, maybe I should go a little bit further. Maybe I should go a little bit further.
And finally, Rich said, Zach, it's time to move on.

There are other questions to answer. I think you got this one.

Zach was, ultimately, able to work out that this newfound citrate eating ability didn't happen all at once in a single mutation, but instead came from a series of mutations that built on top of each other.

Becoming able to use the citrate required multiple steps, multiple mutations.

And over the course of this one population's history, it just happened to pick up a number of the mutations that were needed. So it only needed one or two more mutations to finally be able to do it.

Basically, in order to wind up not in Rome, these E. coli took a bunch of turns along the way.

They had to take one fork in the path of potential mutations and then another fork on that that evolutionary path, and then a few more fork choices until they eventually landed on the path that led them away from Rome and any of its suburbs and towards somewhere else,

their evolutionary Tokyo.

All this left me wondering what this might mean for the big question that Stephen Jay Gould asked, right?

The question of, would our world look different if we re-ran it again, or would it look the same?

Or to put it another way,

how much does history matter?

If you take one fork in the road versus another at some point early on, does that affect anything, or will things eventually loop back to such a similar place that your historical choices are kind of unimportant?

In this extremely simplified, experimental version of evolution,

it seems like in most of the tiny populations of bacteria,

history kind of didn't matter, right?

They didn't wind up in exactly the same place, but they are still eating glucose 37 years in.

And yet, in one population, history did matter.

So what does that mean?

I think the answer I have come up with is history doesn't matter except for when it really does. That natural selection largely is going to drive evolution to a certain set of outcomes.

But sometimes history is going to lead it down a different path, and there's going to be something radically different.

In some ways, this reminded me of the experiment itself. Because it's also full of a lot of historical accidents, right?

Some feel like they didn't necessarily change the outcome, like Richard dismissing that first cloudy flask as contaminated.

But others, like Richard using that old recipe that happened to have citrate in it, those seem to have had a big effect.

So it feels like kind of the same thing. History didn't matter, except when it really did.

But before we draw some grand, sweeping conclusions about the nature of history or of evolution, I think it's important to remember that we don't know the final outcome here, because the experiment isn't over yet.

It's still running.

And maybe, given enough time, some of the other 11 flask populations will develop the ability to eat citrate too.

Maybe all of them will. Maybe all roads will actually lead to citrate eating if we wait long enough.

How can we know?

We don't know. That's one reason I want this experiment to continue into perpetuity.

You could imagine that, as difficult and background-dependent, and rare mutation-dependent as it is, that none of them will get there over any whatever time scale you want to imagine. Or perhaps,

well, one every 35 years is our best estimate. So maybe it will happen in some of the others.

And I'd love to know the answer to that.

A few years ago, Richard decided to pass the experiment on. So it's entering its second scientific generation with a new director.

He is committed to continue the experiment, to keep watching and waiting and seeing what happens next with these bacteria.

When I spoke to him, he told me that he liked the fact that every day with this experiment was kind of like pulling a slot machine, waiting to see if by some random chance, today might be the day that we get some answers to our questions or learn something interesting.

And I guess, if not today, then maybe tomorrow or the day after that.

Or the day after that, or after that, or the next day, or the day after that, the day after that, or after that,

or the next day,

the day after that,

the day after that, or the next day, or that, the day after that, after that, the day after that. There is so much more to read about the long-term evolution experiment.

And I really encourage encourage you to look up their research website and dig in we will link to that in the transcript and we will also link to a great review paper in nature written by james stroud and william ratcliffe it gives an overview of what long-term studies have taught us about evolution so far

this episode was produced by me bird pinkerton it was edited by julia langoria with help from jorge just and paige vega meredith hodnot runs our show noam hasenfeld makes our music.

Christian Ayala did the mixing and the sound design. Melissa Hirsch checked the facts.
And I am always, always grateful to Brian Resnick for co-creating the show.

I'm also grateful this week to the many people who gave me their time to help me better understand some of the science here.

Thanks so much to Jeff Barrick, the new director of the Long-Term Evolution Experiment, to William Radcliffe, James Stroud, Joe Wolf, and Jean-Bernard Caron.

An extra huge shout out to Zach Blount as well for all his time.

If this episode made you think, or if you have something you'd like to share with us, write in or send us a recording. We are at unexplainable at vox.com.
I love hearing from you all so much.

Please write in. If you want to support the show and help us keep making it, please join our membership program.

It is at vox.com/slash members, and you get ad-free podcasts, unlimited access to Vox journalism. What more could you want?

If you sign up because you love Unexplainable, please tell our bosses. It makes a big difference.

You can also support us by leaving a nice rating or a review, perhaps, or just by telling the people in your life to listen. Thank you for listening.

And Unexplainable is part of the Vox Media Podcast Network. We will be back in your ears next week.

Support for this show comes from S.C. Johnson.
We've all been there.

Choosing not to wear your new white shoes because there's a 10% chance of rain, bending awkwardly over the tiny coffee table to enjoy a sip of your latte, not ordering the red sauce.

Those feelings of dread are what we call stainsiety.

But now you can break free from your stainsiety with Shout's Triple Acting Spray that has stain-fighting ingredients to remove a huge variety of stains so you can live in the moment and clean up later.

Just breathe and shout with Shout Triple Acting Spray. Learn more at shoutitout.com.

If you're tired of database limitations and architectures that that break when you scale, then it's time to think outside rows and columns. MongoDB is the database built for developers, by developers.

It's asset-compliant, enterprise-ready, and fluent in AI. That's why so many of the Fortune 500 trust MongoDB with their most critical workloads.
Ready to think outside rows and columns?

Start building faster at mongodb.com/slash build.