Why Info Push Dominates

Some phenomena to ponder:

  1. Decades ago I gave talks about how the coming world wide web (which we then called “hypertext publishing”) could help people find more info. Academics would actually reply “I don’t need any info tools; my associates will personally tell me about any research worth knowing about.”
  2. Many said the internet would bring a revolution of info pull, where people pay to get the specific info they want, to supplant the info push of ads, where folks pay to get their messages heard. But even Google gets most revenue from info pushers, and our celebrated social media mainly push info too.
  3. Blog conversations put a huge premium on arguments that appear quickly after other arguments. Mostly arguments that appear by themselves a few weeks later might as well not exist, for all they’ll influence future expressed opinions.
  4. When people hear negative rumors about others, they usually believe them, and rarely ask the accused directly for their side of the story. This makes it easy to slander folks who aren’t well connected enough to have friends who will tell them who said what about them.
  5. We usually don’t seem to correct well for “independent” confirming clues that actually come from the same source a few steps back. We also tolerate higher status folks dominating meetings and other communication channels, thereby counting their opinions more. So ad campaigns often have time-correlated channel-redundant bursts with high status associations.

Overall, we tend to wait for others to push info onto us, rather than taking the initiative to pull info in, and we tend to gullibly believe such pushed clues, especially when they come from high status folks, come redundantly, and come correlated in time.

A simple explanation of all this is that our mental habits were designed to get us to accept the opinions of socially well-connected folks. Such opinions may be more likely to be true, but even if not they are more likely to be socially convenient. Pushed info tends to come with the meta clues of who said it when and via what channel. In contrast, pulled info tends to drop many such meta clues, making it harder to covertly adopt the opinions of the well-connected.

GD Star Rating
loading...
Tagged as: , ,
Trackback URL:
  • T

    We know there is utility of prediction but how can one quantify each predictions utility (or in the case of the blog, public opinion) without agent based simulations in our non-linear dynamic world?

  • Doug

    The major exception to this is financial markets. The largest and most sophisticated investors aggressively utilize “info pull” techniques like expert networks, channel checking, political intelligence, mosaic theory, meeting directly with management, and talking directly to analysts covering companies.

    This despite the fact that regulators heavily favor an “info push” model where relevant information is simultaneously announced to all investors at once. Insider trading laws, regulation FD, and central bank announcement protocol are all manifestations of official policy that heavily favor “info push”. Yet probably an order of magnitude more investor effort and resources are devoted to pulling information others don’t have, rather than intelligently digesting pushed information

    • http://overcomingbias.com RobinHanson

      Yes, in finance people seem to care more about simple outcomes like how much money they make.

  • IMASBA

    Every student using the internet to search for stuff to put in their thesis or to solve a homework question is “pulling” info. Sure google favors some content but overall and per unit of effort one will easily be exposed to many more independent sources of information and different viewpoints than if there only was the library or a single recommended book to go on. The same is true for dating websites vs. newspaper ads, or browsing the web for music vs. browsing a record store or browsing the web for a vacation instead of reading a travel agency leaflet.

    People will still favor well-connected or in some other way authoritative sounding sources but they can and often do dig deeper than they could before.

  • Mike Johnson

    Well said. I’d also add that pulling info is much harder and intrinsically more draining than sifting through pushed info.

    With pushed info, we evaluate and act on it. Simple enough. With pulled info, we have the additional steps of trying to figure out what’s worth pulling, of trying to figure out what sources to pull from, etc. There has to be a big quality premium to justify this additional effort.

    I think systems (facebook, google) are starting to solve this problem via implicit curation via trusted friends. But low-effort info pulling is such a hard problem.

    • justin

      Extending on this, it’s hard to imagine getting out of the equilibrium where pushing exceeds pulling. If too many people are pushing info, the return to pulling info decreases because search costs associated with finding the best info to pull become too high. By contrast, if a large number of people are pulling info, the return to pushing increases because the pushed info has a higher probability of reaching someone.

  • http://entitledtoanopinion.wordpress.com TGGP

    People don’t pay much for information because there’s such a huge supply. The incentive for info-sellers (who nowadays charge nothing) is then to maximize viewers and make money from ads.

    • http://overcomingbias.com RobinHanson

      But the question is *why* there’s such a huge supply. Why are people more willing to pay to push info onto passive others than people are willing to pay to actively get the particular info they want?

      • Ronfar

        Distributing information has near-zero marginal cost. Of course there will be a huge supply of it.

      • BenGolden1

        I’m not sure how exactly to quantify the comparison, but my sense is that more resources are spent pulling info than pushing it.

        People spend an awful lot of money buying magazines, newspapers, books, college educations, etc., which I’d tend to classify as actively pulling.

      • http://overcomingbias.com RobinHanson

        Magazines and newspapers are pretty unspecific as info sources. Buying access to them is more buying access to a channel that others will push stuff through.

      • IMASBA

        With newspapers at least you do often know their overall stances and they do name their sources, of course the same is true for the internet when you have a lot of experience with it.

      • solipsist

        People do pay to pull information, and handsomely. If I require legal advice, I seek the council of a lawyer. If I require advice on a merger, I hire an investment banker. If I require about widgets, I hire a widget consultant. These services require expert humans who can understand their clients needs and guide their clients into asking the right questions*.

        *and, far less importantly, answer those questions

      • http://entitledtoanopinion.wordpress.com TGGP

        I was thinking of individual blogs, magazines, etc as being sources of information that are deliberately “pulled” by audiences, with the ads being “pushed”. Also, what Ronfar said about the cost of copying/distributing information.

      • http://overcomingbias.com RobinHanson

        If you bought access to a newspaper database and then searched that when you have a specific topic of interest, that is info pull. When you just read whatever articles they put on the front page every day, that is info push.

      • http://www.gwern.net/ gwern

        If searches are ‘pull’, doesn’t that imply there’s an absolutely enormous amount of pull going on every day? Google doesn’t have a market cap of $378b for nothing.

      • STINKY

        Probably because I’m half-reading these comments on an iPad while finishing my breakfast in my building’s coffee shop before I go upstairs to work for the day. Passivity and info push seem to pair well.

  • http://www.frivjuego.co/ Friv juego

    Each person will have different views. Each student used the internet to search engines for inclusion in their thesis or to solve a homework question is “pull” information. Surely google support but some content overall and per unit effort one would easily be exposed to more independent sources of information and different viewpoints than if there is only a library or a book the only suggestion to go on. Very supportive and I thank you for your sharing this.

  • Matthew

    It’s unfortunate but that is how our generation has become. I would love for more people to seek out information and question every answer they receive but it seems that technology has made people lazy. What I find disturbing is that we live in a time where we have so many resources at our disposal that we chose not to use. Or what’s even more disturbing is that many people chose to ignore the issues. I wrote about this a year ago and it wasn’t until recently that I posted my writing in a blog, http://unconventionalviews.blogspot.com/2014/01/problems-with-our-current-generation.html

    • IMASBA

      That’s what the ancient Greeks used to say about their youth. The truth is it wasn’t that different in the past, nostalgia bias doesn’t belong on this website.

  • http://www.kizijogos.co/ Kizi jogos

    Can it be considered a pretty perfect way to boost the efficiency of the work. I think it really brings many benefits, not just a useless method

  • Vicificient

    Several thoughts:

    1) People really like attention, so that’s an incentive to share interesting information you already have for free. And if most other people are sharing information for free, it’s hard to compete unless you have far superior information.

    2) If information is new, then there’s no way for you to know you want to pull for it until somebody has already pushed it at you, is there? Finding worthwhile stories we haven’t heard of is what we pay journalists for.

    3) Trying to think of a pull-type service that’s successful, Amazon comes to mind. You can choose what you want more information about and pay for a book about it.

  • Christian Kleineidam

    A lot of nonfiction authors and bloggers don’t support themselves by advertising but make money by public speaking or consulting. Pushing means you can give everyone the same product. You can do that for cheap. On the other hand you can charge good money for specific consulting.

    I don’t know exactly the business model of this blog but I see no ads even if I shut down my adblocker.

    Google might make most of it’s money via advertising but most of the time the user of Google still clicks on a authentic link instead of clicking on the advertised result.

    Another huge trend that you missed is the switch from TV to Netflix. We see people subscribe to Spotify to replace radio consumption.

    It might also be a problem to see everything in terms of money. Wikipedia provides content for people who pull articles that are relevant to what the people want. Wikipedia just isn’t a commercial enterprise. Programmers who pull information about a problem that they have by asking a question on stackoverflow are also interacting with a complex system where money is only a side issue.