How social networks make us smarter | Alex 'Sandy' Pentland | TEDxBeaconStreet



so I'm motivated by what is the biggest problem in the world and it's not global warming and it's not war instead I think the biggest problem in the world is how do we get all of humanity to work together to make good decisions and follow through on them you know we could solve global warming tomorrow if we could just get our act together and carry through on it now the impossibility of getting everybody to work together sounds so utopian right means we don't even think about it but there is hope in the last decade for the first time ever almost all of humanity is connected by cell phones computers and other networks that means for the first time ever humanity has a voice and we have social networks to connect each other and share data unfortunately those social networks seem to be more the cause of panics and crashes than the source of wisdom to really get them to make us smarter to solve this big problem we have to figure out how we work how we share ideas as humans so that we can work when they are Humanity and not against it so for the last two centuries since Adam Smith we've thought of ourselves as individuals competing we write our government that way we run companies that way oh you and I also know that we're much more than that we're social beings we interact with each other and in fact one of the main things we do is we learn from each other that's the original social network and that's the social network that we have to leverage if we're gonna make ourselves smarter now that's been difficult because the science is really tough I do you make measurements but there's been a revolution and the revolution is big data a big data is creepy in many ways but for the social scientists it finally gives us the day from social networks from cellphones from credit cards that we need to be able to understand human behavior and as a consequence we can actually write down equations that talk about ideas flow from person to person through social networks we can predict things now don't worry I'm not going to talk about math here the big story is that this science is emerging and I call it social physics because that was the name given to it two centuries ago by the person that founded sociology and he believed that society was shaped by a progression of ideas we believe that those ideas would emerge in lockstep uh and that's not true but the new science the mathematical predictive science tells us that something nearly that way is true and that is is that the propagation of ideas from person to person is the source of a collective intelligence that we often call culture so first us with colleagues at MIT we gathered up hundreds of teams of regular people and put them through their paces solving all sorts of problems and what we discovered is is that teams of people indeed have a collective intelligence that is independent of the individual intelligent and greater than the individual intelligence and it's not a small fact it's a really big effect and the thing to think about it is is that it's the pattern of idea flow that makes these teams smarter so when there are lots of ideas coming in and everybody is equally engaged then the collective intelligence of the group goes up something to remember the same rule works in companies so what we can do is we can make visualizations of idea flow so this is perhaps the world's first visualization it's a German bank it's a creative department that does advertisement and things like that and the little boxes are things like the managers and customer service and development and between them you see in blue all of the email they're sending an email back and forth is they design this new ad campaign and something you've never seen before is the red stuff those are all the face-to-face interactions not the words just who talked to who and we measure that using these little electronic name badges that we built so we can see how ideas propagate around in the organization and what we find when we analyze this versus the output of the organization is that the electronic stuff doesn't really make all that much difference but the face-to-face stuff does and when we do a graph of the face-to-face stuff we find the work group is this stuff in red and just as in the team's the more cohesive that is the more engaged people are the better the productivity of the team often forty percent of the variation is just that one pattern of idea flow what they're acting like is a little ideal machine that takes ideas processes them against each other and selects the wisdom that will move them forward and I call this engagement nice word people understand it and what you find just as we're the team's is if you can improve engagement you improve the productivity of the group but a group like that that just has engagement is very static and not very creative creativity comes from ideas outside of the work group so something I call exploration the exploration feeds in new ideas into this idea machine which chews them up and figures out new ways to do things and in drug discovery and creative teams it's this exploration together with the engagement that produces creative output so you can vary this in various ways you can improve an organization by shaping the idea flow first us in this German Bank it turned out that the customer service group was not in the loop the idea flow didn't really reach them so we changed the seating around so that they would be much more in the center of the idea flow and lo and behold things work lots better in a call center we changed the coffee break structure so the people got more informal conversation lo and behold the productivity went up and in creative groups like drug discovery the same thing happened so it really is a causal thing this idea flow is what generates both productivity and innovation in small groups in dozen of companies and even in entire cities so we gathered data for 300 cities half in the EU half in the US and we use things like Foursquare to measure the idea flow face to face idea flow in the city and we found that just from that you could predict the wealth of the city you didn't have to look at education level you didn't have to look at class structure you didn't have to look at specialization just idea flow so it's not looking so good for the digital social networks because they weren't involved in that they weren't very important in the companies what's going on how can we use the lessons we have from the face-to-face networks to make the digital social networks work better they answer that let me give you a specific example that has a very clear cut edge to it so this is millions of people who were on a social network called eToro that are doing things like buying stocks trading them for dollars or euros buying gold silver so their financial day traders and what I'll do is for this network there's millions of people you can see the idea flow because they copy each other they follow each other all the normal sorts of things and so for each day you can calculate the idea flow for the millions of people and then at the end of the day you can ask how good were their decisions because at the end of the day they either made money or lost money and what you see off at the left is people with very low idea flow so a day's very low idea flow people are mostly independent they're isolated and they don't make any money their decisions aren't very good as idea flow goes up remember these are millions of people whole days millions of decisions this idea flow goes up they make huge amounts more money and that's compounded so it's actually an exponential process but then as you get even more idea flow something weird happens they entered the echo chamber so when idea flow is too fast you get these avalanches of things that just rush through the whole network and as a consequence you get the same ideas over in Oregon that's why it's called echo chamber and the accuracy of the decisions goes down and in fact what this doesn't show it but this is also where bubbles and panics come from where people lose everything so social networks can generate wisdom of the crowd but they can also generate the madness of the crowd through these echo chambers so the thing to remember is no echos good decisions idea flow works up to a point and then stops also know what happens in face-to-face networks well in face-to-face networks we don't see this echo chamber as much not that it doesn't happen but you're pretty good at detecting it when the cab driver starts giving you stock tips you should be suspicious when everybody in the crowd is chanting the same slogan over and over again you should stop back and say well maybe it's a good slogan but you know we're not thinking independently here are we right so we have good ways to detect echo chambers in the real world but we don't in social networks and as a consequence if we ask this question how can social networks make us smarter how can we come together as a species and make better decisions we find that our current social networks have the problem of echo chambers and these echo chambers destroy the wisdom of the crowd so we're left with face-to-face connections the thing we evolved to make us smarter and those really do work they change minds but what can we do in the future well one of the things we can do is take our face-to-face skills for detecting echo chambers and apply them to digital social networks so you each can do that you're really good at detecting echo chambers in the real world you should use some of the same logic when you look at social networks and in the longer term where we do as we should build social networks they'd have echo cancellation to protect us from the madness of the crowd [Applause]

5 thoughts on “How social networks make us smarter | Alex 'Sandy' Pentland | TEDxBeaconStreet

  1. Just from reading the title I had to write this hateful comment: "Social Networks make NSA and CIA smarter – not us".

  2. This guy is on the board of trustees for instagram. No wonder he would say something like this. At first I thought he must be joking and the conclusion of the talk would have a sarcastic overtone of the title, which I thought was obviously a joke and I though he'll accordingly conclude with the sentence 'social networks don't make you smarter but dumber'. But hey what do you know he is actually trying to pull this one through and make it look like socials networks make us smarter. Then I did research and found out he he is on the board of a social app, go figure. In short he is promoting social networks because he makes money off of them. Meaning, this is a commercial you're just now watching. Mind blown. I know right. You can only be this smart if you're not on a social network.

  3. Isn't he involved with the Network Science Collaborative Technology Alliance (NS CTA), which openly says that "its purpose is to perform foundational cross-cutting research on network science, resulting in greatly enhanced human performance for network-enabled warfare and in greatly enhanced speed and precision for complex military operations", and that it seeks to "Understand, Predict, and Influence" social interactions? Is he collaborating on the weaponization of the knowledge he tries to share here?

  4. Reminds me of the Asimov's Foundation Series – Hari Seldon's Psychohistory & mathematical sociology being born now…
    http://en.wikipedia.org/wiki/Foundation_series

Leave a Reply

Your email address will not be published. Required fields are marked *