Skip to main content
Search form
  • 00:05

    [Moran Cerf Discusses Developmentsin Computational Neuroscience]

  • 00:10

    MORAN CERF: My name is Moran Cerf.[Moran Cerf, Associate Professor, Kellogg Schoolof Management, Northwestern University]I'm an associate professor here at Kellogg.And I'm a neuroscientist by training.And I teach how to use neurosciencein the context of business.[What is the most exciting innovation in computationalsocial science?]I'm biased a little bit, but I thinkthat what's happening gradually is

  • 00:31

    MORAN CERF [continued]: that more and more computational social scientists are realizingthat they need to look at the brain to understand behavior,and that the brain is coding many of the thingsthat we explained before using psychology.So we're looking at things like machine learning, whichis a buzzword that is very popular rightnow among computational social scientists.

  • 00:51

    MORAN CERF [continued]: If you try to kind of track the origin of this word,it's actually coming from neuroscience.It's how the brain learns.This is true for a lot of terms, like deep learning,and generally neural networks.And many of the things that are out there in the field,they come from neuroscientists whodidn't know that their tools would be used for anythingbut explaining the brain's behavior.

  • 01:14

    MORAN CERF [continued]: That's kind of the methods part.And on the behavior part, generally, Ithink that we're starting to look into the brainand realizing that a lot of the anomalies of behaviorcan be explained there.So I'll give you example that's concrete.For a while, there is this field that'sgrowing called behavioral economics, whichtries to predict how people will do things when they do them

  • 01:34

    MORAN CERF [continued]: in ways that are irrational.So they're supposed to go right, but they go left,because somehow something made them go left that theycould not explain perfectly.And when we look at the design of the structure,we figure out that we could predict that.Behavior economics has given us a lot of explanationsto people's behavior.But it also fails in two domains.

  • 01:55

    MORAN CERF [continued]: It fails to explain why are we doing things thatare irrational, which when we tell us what we did, we say,oh, I want to change.And also, it doesn't help us change.So we make mistakes again and again.And we can't fix them.And what we're learning is that if you take one step backand you look at the brain, you can actuallypredict the behavior, and even understand what

  • 02:16

    MORAN CERF [continued]: is the root of the behavior.And you can actually change.You can actually take a person, the extreme version of that,bring her over to the lab, have her take a nap of two hoursand wake up different.And that is where we're aiming right now,taking the field that propagate in computational socialscience, but adding the strawberry or the icingon the cake that will actually explain things fully.

  • 02:38

    MORAN CERF [continued]: [How will innovation in computational social sciencechange life as we know it?]I think that we're beginning to understandthe nature of something that is profound,which is what humans are for, and whatis the future of humanity.That's a big opening.I think that up to now we kind of looked at our livesin the context of hisstory and tried to draw conclusions

  • 02:59

    MORAN CERF [continued]: from where we're going.But for the first time, we're startingto take ownership of the future.We're saying, OK, evolution took that many years to get usto become who we are right now.And if we want to have wings and we rely on evolution to do it,it's going to take millions of yearsand probably not going to be in our lifetimes.But wait.We know how to do it ourselves.So suddenly we can take ownership of evolution,

  • 03:20

    MORAN CERF [continued]: and build things for us by the way.When we think about learning, for a while,learning was a thing that we tried to get better,but we tried to improve the school system or educationor small scale things.And now we say, what is learning?It's actually getting content into the brain.Maybe you can just do it efficientlyby plugging things directly.

  • 03:40

    MORAN CERF [continued]: Communication.If you take a person from the 100,000 yearsago, Neanderthals, they still used their mouth to speak.This is one of the few things thatdidn't change at all in the last 100,000 years.We're still using the same methods to communicate.But we're realizing right now that communication is justtaking an idea in my brain and putting it in your brain,and language is just the most efficient thing

  • 04:01

    MORAN CERF [continued]: we had up to now.Now it's time to say, wait, we can do better.We can actually connect the brains using alternative waysand communicate ideas directly.And in doing so, we actually aren't justcommunicating ideas perfectly, moving thoughts from one placeto another perfectly.We are also speeding up this thing.When I speak, even though I speak very fast,I probably speak up to 150 words per minute.

  • 04:23

    MORAN CERF [continued]: That is really slow compared to how fast my brain can think.My brain can think about anythingbetween 400 and 500 thoughts per minute,or some say if it's just visual thoughts, it's 60,000.So we can really do things much faster.And we're slowed by our old way of communicating.So here is a way that we can improve that.And that's not just a mechanical way.

  • 04:43

    MORAN CERF [continued]: It's actually entirely changing the wayhumans are going to interact, because--and this is where I'm going to end, this kind of inspiringthing about the future--many people think that what makes humankind unique, whatmakes us the crown of creation, what makes us stand outin nature is that we have a component in our brainthat no other animal has that allows us to not just live

  • 05:06

    MORAN CERF [continued]: in the moment, but actually reflect on things,think about the future, and imagine futures that aren'tthere, and really know without even trying that ice cream thatwill be made of olive oil might not be as tasty as one madeof vanilla.We don't need to taste it.We can just imagine it, and it's going to work.But it also allows us to invent complex ideasand show them without actually having a physicality to them,

  • 05:29

    MORAN CERF [continued]: ideas like democracy.It doesn't exist, really.We can't wear it, or eat it.But I can invent in my mind, tell you,and you and I live through a world with that.We can invent complex ideas, like the blockchain,or whatever.Things that you hear a lot of people thinking about,they don't really exist in the sensethat they're anything but an idea.You can't communicate them to a puppy,

  • 05:49

    MORAN CERF [continued]: because they just don't have any sense of things that aren'tphysical or aren't tangible.And in that sense, humans have spiked in evolution,because we started to believe in things that were more complex.And living through them, this is a networkof social ideas, social people living by ideasthat actually communicate them.This is like what a lot of the world that we live in rightnow is enjoying, memes, and complex concepts that travel

  • 06:13

    MORAN CERF [continued]: faster than anything else.When we realize that this is one of the powers of humanity,to have those ideas, and that they actuallycreate a network of humans that is bigger than an individual,I think we're seeing the idea of where it could go.It could go into a world where humansbecome one big entity, rather than many, many unitsof individuals.[How has computational social science affected your work?]

  • 06:38

    MORAN CERF [continued]: I'm trying to think about neuroscience in connectionwith business.And this is what I'm doing here at Kellogg.And one of the things that I'm doing that is relevantis I'm trying to see if there's a power to many brainsthat is bigger than the power of an individual brainin solving problems that are difficult for us to introspecton alone.So I'll give you an example that's concrete.When you watch a movie, and I am the filmmaker

  • 07:02

    MORAN CERF [continued]: and I want to know if the movie is good,there are many ways I can know that.I can see if you recommended your friends to go watch it.I can see if you paid for the second movie thatcomes out in the series.I can also ask you.And I can also have people with dials do it.But what we learn is that all of those are not really reliable.It turns out that sometimes people really wantto watch a movie, but there's another one that's

  • 07:22

    MORAN CERF [continued]: actually good, and they go to watch the other one.And your movie lost not because it's not good,just because there's another one that also is good coming out.Maybe people don't remember things down the line.Maybe people are biased by the ending.It's a really boring movie, but she fell in lovewith the vampire in the end, and we immediatelysay, oh, it's amazing.I loved everything.So a lot of things go into that.And a movie is just an example I gave.It's true for music, a conversation, any experience

  • 07:45

    MORAN CERF [continued]: that is kind of dynamic is reallysomething that's hard for us to introspect on,because to do it properly, we need to stop the experience,and ask about it while you're going through it.So if I want to know how interesting I am to right now,the best way to do it is to stop right now and to ask you.But in doing that, I already ruin something,because the experience is no longer lived,but it's kind of reflected on.

  • 08:07

    MORAN CERF [continued]: So the question is, is there a way for usto look at an experience and know whether you're enjoyingit, whether it's interesting for you without asking youand without stopping it and without waiting to the endand wait to see how you reflect on the entire historyand maybe be mistaken by the end?It turns out that the answer to how interesting I am to youand how interesting this experience is is in our brain.

  • 08:27

    MORAN CERF [continued]: And if I can read it from your braindirectly while you go through it, I don't have to stop you.And that is what we are trying to do.But what we learned is that it turns outit's really, really hard to ask one brain.If I just look at your brain, it's really hard to know,like there's going to be a lot of things therethat are going to go up and down.And there are going to be momentswhen we talk that your brain is going to be metaphorically verysilent, which I would say, oh, my god, she's really bored.

  • 08:47

    MORAN CERF [continued]: But actually, this silence is what lives--what makes you kind of take a pause, if you want,so you can enjoy the moment after much more.So all of those kind of moments of silence or increasedactivity are actually part of the entire experience.And if we just try to look at them as like moment to moment,we would be mistaken.What we found is that the best way to actually know

  • 09:08

    MORAN CERF [continued]: if an experience is good is to lookat many, many brains who went through the experiencesimultaneously.So if I want to know if the movie is really, really good,I shouldn't just look at one brain,but I should look at the entire room, all the brains of allthe audiences, and see how similar they are to each other.The more similar they are each other,the better the movie is, because the moviewas able to tap into every person's brain in the same way

  • 09:29

    MORAN CERF [continued]: despite who they are.So an old person, a young girl--I can't-- people who love the movie,people who hate the movie.Somehow when they see a Hitchcock film,they all look the same.Their brains look the same, because hewas a master in tapping into everyone's brain,figuring out who we really are, and making us all look alikeso our brains kind of move in the same waves,

  • 09:49

    MORAN CERF [continued]: including the silences all perfectly crafted by him.Mozart's genius was that he couldsit in his room and kind of compose something,knowing somehow how the brains of many, many other peoplewould look years after he was dead.Somehow the genius of people is to beable to realize how to go into people's brains.So what we see is that if you look at many brains

  • 10:12

    MORAN CERF [continued]: and you see how similar they are,you can actually know how good the content they experiencedis.So this is one of the first timesthat we realized that I can't look at one brain,but I can look at a group of brainsand measure not their response to a movie,but how they look together.And as a group, there's a lot more information therethat tells me that the movie was good or bad.More similar, better movie.Less similar, less good, less engaging movie.

  • 10:34

    MORAN CERF [continued]: It's true for conversations, for music, for political opinions,anything you can imagine.We can figure out by looking at many, many brains who they'regoing to vote for, how they're going to changetheir mind, what's going to affect them,what they're going to find emotional, what not,what they're going to find annoying.All of that by just looking at how similar their brains are.I mentioned looking at the brain.The brain controls their heart rate

  • 10:55

    MORAN CERF [continued]: and the respiration and skin conductance and your eyemovement.So one of the things we look at isall kinds of residuals, proxies of the brain.And they all become very much alike to the extreme versionthat some of the scientists who are looking into it right nowfound that if you go back to moviesas a really good kind of test bed for content

  • 11:15

    MORAN CERF [continued]: and how interesting it is, the air conditioner in the movietheater, if you put a sensor on it, you can actuallyknow how good the movie is by the effectof the movie on everyone else, because they allgasp at the same time.And they suck the air from the room.And you can actually figure out which moviethey're watching just by looking at the air conditioner.

  • 11:35

    MORAN CERF [continued]: They all kind of sigh in the same moment.They all emit some smells and some odors from their bodywhen they get sweaty or when theyget-- like they move too much.So you know a lot just by how our body responds to content.[What advice would you give to business leaders about usingcomputational social science?]I feel like I'm biased, but I actuallythink that the next kind of big step in how we would think

  • 11:58

    MORAN CERF [continued]: about all the problems in the business worldwould be adding neuroscience to them,would be actually not asking you questions, but evaluating youby looking at your brain.This is true for HR, how we're going to interview people.This is true for how we assess people when they work.This is true for how we learn about our customers.Instead of asking them, we're going to look at them.So I think that if I were an entrepreneur right now,

  • 12:20

    MORAN CERF [continued]: I would make a point to befriend a neuroscientist.It's not there yet in the sense that we'redeveloping the tools.But they're still too expensive.They're too slow.They're too niche.They're in academia, but not in the business world.So I don't think that a lot of companiescan afford to incorporate that in their day-to-day experience.The big ones, Google, Facebook, Microsoft can,

  • 12:42

    MORAN CERF [continued]: and that's kind of where the world is heading right now.Silicon Valley is really using that a lot.And they're going to drive this innovation.But I think that if I were an entrepreneur,I would kind of keep my finger on the pulse.I would say every two weeks I'm going to kind of Google what'sup with neuroscience and business,and see if there's something I can use.I would become friends with a neuroscientist

  • 13:02

    MORAN CERF [continued]: and ask him or her every kind of six months, hey,what's new in your field?Because I think that at some point,it's going to be a big drive.And it's already shown that machine learning and deeplearning that came from neurosciencebecame like the world du jour, if you want,of like almost any company does analytics.I think it's going to happen more and more frequently.And if I were starting a company right now,

  • 13:24

    MORAN CERF [continued]: I would if not use it already, at leastkeep my finger on the pulse so I wouldknow what's happening in this fieldso I can be an early adopter.[What was a project you've worked on where you facedmethodological challenges?]The part that I-- one of the parts that Ireally, really liked that I still believeis going to change the world is trying to understand dreams.

  • 13:45

    MORAN CERF [continued]: So dreams are something that most of us experience.They're very meaningful to us, in a sensethat we put a lot of meaning into them when we wake up.If your boyfriend woke up and tells youthat he has a dream about his ex-girlfriend,you're going to be mad at the awake self.You wouldn't say, oh, it's your brain whocame up with this dream.I would blame the awake you.

  • 14:06

    MORAN CERF [continued]: The Bible is full of stories of people whowent to war because of dreams.History is full of hieroglyphs in caves of peopledrawing their dreams.It's something that since we know ourselveswe've felt it is important.And yet, we're very, very poor in actually reaching outto them.All we know is what you tell us when you woke up.That's basically where we are since the days of Freud

  • 14:26

    MORAN CERF [continued]: and up to now, the last 100 years.We don't have really a good way to extract him, let alonechange them.If you say, "I have bad ones.I want to change them," we're really notthere yet when it comes to doing things to them.We're getting close.The last 10 years have seen stepstowards actually looking directly at brains of people

  • 14:48

    MORAN CERF [continued]: and extracting content and visualizing it and gettingaccess to the narrative of your dream.And that's when the technical problem comes.It's really, really hard to do this research,because you're asleep.And I need to know when you're dreaming.I need to know it in real time.I need to know that you're dreaming right now.I can't say, oh, she has dreamt in the last 20 minutes.

  • 15:08

    MORAN CERF [continued]: And the dream is already lost.I have to know right away.I have to not wake you up when I try to extract things.And I'm outside your brain, not inside mostly.I can't really drill holes in your headif I'm not doing it in a clinical setting.So there's a lot of challenges there.And this is vague.Here's a specific one.One of the difficulties we have right now isto actually have a person sleep next to us with EEG--

  • 15:29

    MORAN CERF [continued]: this is the device that usually reads the brain activity--and have someone else look at their brain activitylive while they're sleeping, and say,right now the dream started.Because there is no real need for thatin the clinical world-- in the clinical worldwhen you do sleep studies, you areOK with just waiting for the end of the night,and then looking at the entire nightbackwards-- there are very few peoplewho actually developed tools to do it in real time,to get a reading of the moment the dream is happening

  • 15:51

    MORAN CERF [continued]: right now.And the tools that we have right now are not great.And this is a struggle that me, my students, my colleaguesare suffering with.It's very specific.It needs a push by people, like the people in your group,and the people who might watch this, who say, oh, that'san interesting technical problem that I think I can solve.

  • 16:11

    MORAN CERF [continued]: It requires more data.We need more sleeping people.We need people that sleep with no problems.We need-- there's a few challenges thatare all technical.Once we overcome this thing-- and it should be overcome.It's embarrassing that we didn't solve it yet.I think we're going to make the biggest leap in understandinghow dreams are, which I can assure the viewers every time Igive a talk and I speak about the things thatare interesting, dreams make everyone's eyes pop.

  • 16:32

    MORAN CERF [continued]: Everyone finds it interesting.So it's something that we should just get over.


Moran Cerf, PhD, Neuroscientist and Associate Professor at the Kellogg School of Management, Northwestern University, discusses developments in computational neuroscience, including exciting innovations in the field, how these will change life as we know it, how it has affected his work, advice for business leaders, and methodological challenges faced in this type of research.

Looks like you do not have access to this content.

Moran Cerf Discusses Developments in Computational Neuroscience

Moran Cerf, PhD, Neuroscientist and Associate Professor at the Kellogg School of Management, Northwestern University, discusses developments in computational neuroscience, including exciting innovations in the field, how these will change life as we know it, how it has affected his work, advice for business leaders, and methodological challenges faced in this type of research.

Copy and paste the following HTML into your website