Nova Spivack - My Public Twine Nova Spivack - My Public Twine / Items

Can Twitter Survive What is About to Happen to It?

Get Feed

I am worried about Twitter. I love it the way it is today. But it's about to change big time, and I wonder whether it can survive the transition.

Twitter is still relatively small in terms of users, and most of the content is still being added by people. But not for long. Two things are beginning to happen that will change Twitter massively:

  1. Mainstream Adoption. Tens of millions of new users are going to flood into the service. It is going to fill up with mainstream consumers. Many of them won't have a clue how to use Twitter.
  2. Notifications Galore. Every service on the Web is going to rush to pump notifications and invites into Twitter.

Twitter reminds me of CB radio -- and that is a double-edged blessing. In Twitter the "radio frequencies" are people and hashtags. If you post to your Twitter account, or do an @reply to someone else, you are broadcasting to all the followers of that account. Similarly, if you tweet something and add hashtags to it, you are broadcasting that to everyone who follows those hashtags.

This reminds me of something I found out about in New York City a few years back. If you have ever been in a taxi in NYC you may have noticed that your driver was chatting on the radio with other drivers -- not the taxi dispatch radio, but a second radio that many of them have in their cabs. It turns out the taxi drivers were tuned into a short range radio frequency for chatting with each other -- essentially a pirate CB radio channel.

This channel was full of taxi driver banter in various languages and seemed to be quite active. But there was a problem. Every five minutes or so, the normal taxi chatter would be punctuated by someone shouting insults at all the taxi drivers.

When I asked my driver about this he said, "Yes, that is very annoying. Some guy has a high powered radio somewhere in Manhattan and he sits there all day on this channel and just shouts insults at us." This is the problem that Twitter may soon face. Open channels are great because they are open. They also can become aweful, because they are open.

The fact that Twitter has open channels for communication is great. But these channels are fragile and are at risk from several kinds of overload:

  • Hypertweeting. Some Twitter users tweet legitimately, but far too much. Or the content they tweet is just innane. In doing so they market themselves and dominate everyone's attention with their presence.
  • Hashtag Spam. For example, an advertiser could easily pump out tweets that market their products, and simply attach popular hashtags to them, thus spamming those "channels" with ads. Similarly, clueless users could do the same thing.
  • @reply Spam. Another way that spammers could create annoyances in Twitter is by doing @replies to people, with ads for products, or simply to make trouble.
  • Twitter Chains. It is easy to package a highly contagious meme as a tweet that spreads linearly or exponentially. Some variations of this are highly self-replicating and can quickly spread to millions of people. There are various ways to design such memes to spread exponentially and across multi-level relationships with extreme virality. Multi-level marketers and others could take advantage of this to create havoc, and even potentially flood Twitter with multi-level messages to the point of crashing it.
  • Notification Overload. Another issue is the rise of Twitter bots from various services, whether benign in nature or deliberately spammy:
    • It won't be long before every social network starts pumping updates into Twitter.
    • News and content sites are starting to pump updates into Twitter for every article they publish.
    • Games and MMPORG's are starting to pump notifications for things that take place in their worlds, into Twitter (e.g. player x just defeated player y in a battle) 
    • A variety of other desktop, online and mobile apps will be pumping notifications into Twitter

There is soon going to be vastly more content in Twitter, and too much of it will be noise.

The Solution: New Ways to Filter Twitter

The solution to this is filtering. But filtering capabilities are weak at best in existing Twitter apps. And even if app developers start adding them, there are limitations built into Twitter's messaging system that make it difficult to do sophisticated filtering.

Number of Followers as a Filter. One way to filter would be to use social filtering to infer the value of content. For example, content by people with more followers might have a higher reputation score. But let's face it, there are people on Twitter who are acquiring followers using all sorts of tricky techniques -- like using auto-follow or simply following everyone they can find in the hopes that they will be followed back. Or offering money or prizes to followers -- a recent trend. The number of followers someone has does not necessarily reflect reputation.

Re-Tweeting Activity as a Filter. A better measure of reputation might be how many times someone is re-tweeted. RT's definitely indicate whether someone is adding value to the network. That is worth considering.

Social Network Analysis as a Filter. One might also analyze the social graph to build filters. For example, by looking at who is followed by who. Something similar to Google PageRank might even be possible in Twitter. You could figure out that for certain topics, certain people are more central than others,  by analyzing how many other people who tweet about those topics are following them. Ok good. Nobody can patent this now.

Metadata for Filtering. But we are going to need more than inferred filtering I believe. We are going to need ways to filter Twitter messages by sender, type of content, size, publisher, trust, popularity, content rating, MIME type, etc. This is going to require metadata in Twitter, ultimately.

Broadly speaking there are two main ways that metadata could be added to Twitter:

  1. Metadata Added Outside Twitter. Twitter messages could simply be URLs that point to further resources that in turn carry the actual body and metadata of each message. Thus a message might just be a single URL. Clicking that URL would yield a web page with the content and then XML or RDF metadata about the message. If this were to happen, Twitter messages would be simply URLs created and sent by outside client software -- and they would require outside software (special Twitter clients) to unpack and read them.
  2. Metadata Added Inside Twitter. Another solution would be for Twitter to extend their message schema so every Twitter message has two parts, a 140 char body and a metadata section with a certain amount of space as well. This would be great. It would be a good move for the people at Twitter to jump the gun by enabling this sooner rather than later. It will help them protect their control over their own franchise.

One thing is certain. In the next 2 years Twitter is going to fill up with so much information, spam and noise that it will become unusable. Just like much of USENET. The solution will be to enable better filtering of Twitter, and this will require metadata about each tweet.

Someone IS going to do this -- perhaps it will come from third-party developers who make Twitter clients, or perhaps from the folks who make Twitter itself. It has to happen.

(To followup on this find me at http://twitter.com/novaspivack)

Now read Part II: Best Practices - Proposed Do's and Don't's for Using Twitter

See Also:

This article on CNET

A new article on CNET mentioning this article

Comments

  • Public Comments

    • jebworks jebworks
      10 months ago


      Totally agree with the opinion expressed. Twitter is already getting quite unwieldy in many respects and filtering will become a necessity. Another one is the ability to tweet to closed groups, a kind of SMS based on Twitter.
      Nova Spivack - My Public Twine

    • Barry O'Gorman Barry O'Gorman
      10 months ago


      Could not agree more. Really only began to use Twitter a few months ago. Great way of connecting with people interested in same (and related) areas of knowledge/research. Unfortunately about to be bombed out of existence if steps not taken.
      Nova Spivack - My Public Twine

    • François Dongier
      10 months ago


      Yeah, that's the problem with push technologies. Couldn't you say exactly the same about RSS?
      Nova Spivack - My Public Twine

      • Nova Spivack Nova Spivack
        10 months ago


        No because RSS isn't a bunch of community channels -- in Twitter anyone can broadcast onto a hashtag or even by simply doing an @reply to you. Everyone has to see that. RSS feeds are one-way -- there is no way to reply on them, and there is no equivalent of hashtags.
        Nova Spivack - My Public Twine

        • François Dongier
          10 months ago


          I use RSS to follow communities and I suspect you do too: I subscribe to the comment streams of several blogs, to the change page of several wikis, and last but not least, I also subscribe to several twine feeds. If someone writes junk in a wiki, I get notified. If I subscribe to a twine, I get in my feed-reader all new items and comments posted to this twine. Including spam. This spam by the way will also appear in my daily digest and in my interest feed. Isn't subscribing to a twine very similar to subscribing to a hashtag? I also subscribe to the "new twines" feed and regularly get spammed there. I also have a feed that brings me all the comments in Twine, and "unwanted material" sometimes also pops up there as well.
          I'm afraid all the open systems that somehow "push" their new (non-moderated) content will every so often push garbage. That's the risk you take by subscribing to such non-moderated open sources..
          Nova Spivack - My Public Twine

          • Phil Duby Phil Duby
            10 months ago


            As described, that is still a bit different. The twine version of subscribing to a hashtag would be to subscribe to the RSS feed for a twine search, where the search is tied to a twine tag. That would then get all content with that tag, regardless of which twine(s) it was created in and shared to.
            Nova Spivack - My Public Twine

    • nerxs
      10 months ago


      Interesting read. Never thought about this. I always thought about twitter as almost resistant against spammers because of its subscription based character unlike email.
      Nova Spivack - My Public Twine

    • François Dongier
      10 months ago


      Why wouldn't some sort of collaborative spam filtering work here (without the metadata overload)? If it works for email, why not for the poor man's email?
      Nova Spivack - My Public Twine

    • 10 months ago


      Metadata and filtering and all of this kind of stuff would destroy the twitter spirit. Never saw a bird filtering, tagging, adding metadata to their chirps... ;-)
      Notification overload is indeed a problem but its up to you to tell/ask them to reduce or simply dont follow or block them. Or am I missing something?
      Nova Spivack - My Public Twine

    • sandybrode sandybrode
      10 months ago


      As a recent facebooker and new twitterer this week. I've already learned to unfollower hogs of the system. I like to read many posts and not waste time sifting through abusers of attention. In fact, I just joined twine this morning so I could comment on this like minded article. Not a big fan of joining tons of sites or blogs- to be seen or heard I'm however, already impressed by this one!
      Nova Spivack - My Public Twine

    • 10 months ago


      Just an idea. Tim o´Reilly overloads my twitter with a lot of updates, but I love it, because they contain a lot of information and intersting thoughts for me.
      Information relevant for me. So the quality, context, content, relevance and authority is a way to weight or rate twitter updates. Full automatic or customizable.
      Funny how the rules and patterns of communication (signal-to-noise-ratio, from neuroscience to twitter) always emerge even in the smallest niche.
      Nova Spivack - My Public Twine

    • arnt eriksen arnt eriksen
      10 months ago


      I disagree to some extent, because - unlike a lot of other social media platforms, you choose the people you recive the noice from. If someone is overloading your twitterfeed with 'spam' or information you don't want, you just unfollow them and problem is solved. Twitter is a open social media network, with the choice to find people with information and interesst similar to yourself. The reason twitter works as well as it does, is it's simplicity - and it is open to everyone that want to be apart of the conversation.
      Nova Spivack - My Public Twine

    • Jon Garfunkel Jon Garfunkel
      10 months ago


      Nova --

      Agreed with your problem definition.
      See "Semantic Social Media Construction" for a solution I offer:
      http://bit.ly/ssmc

      Jon

    • 10 months ago


      Agreed, as written in my former comment. I was just thinking aloud about an idea of how to get in control of the overloading (of people I am deliberatly following, especially when you have lots of them) in my twitter accout.
      e.g. I could rate my mothers 100+ chirps a day less important than Tims 10 chirps a day. May be something like a "chrips cloud" with the high ratet chirps "louder" in foreground and the less important damped or muted in background but still there.
      It is a bit like in the real birds world, where numbers of Zebra Finches twitter in the bushes, but the female finch listens to the male finches and the one that twitters most colourful will get her full attention. The rest is just noticed as a muted twitter cloud ;-)
      Nova Spivack - My Public Twine

      • Kissa Watson
        10 months ago


        I like this idea. I wish there was a way to categorize who you follow into sections. There are a couple people that tweet a lot, and I like their tweets, but it drowns out the others that don't tweet as much.
        Nova Spivack - My Public Twine

        • annette-r annette-r
          10 months ago


          There is, Kissa! It's called TweetDeck. You might want to have a look at this list of various useful applications to make managing Twitter easier: http://tinyurl.com/d9d3nr
          Hope you find that helpful. :-)
          Annette.
          Nova Spivack - My Public Twine

    • Bean Fairbanks
      10 months ago


      I like the idea of the "chirp cloud". I follow less than 200 people but I find it overwhelming at times. I use groups so I never miss a student tweet and another group so I can browse through the sheer volume of Chris Brogan's tweets for the gems hidden within when I have the time. It is easy to unfollow someone that is pure noise. The trick is filtering someone whose some interests are the same but not having to spend the time to scan thru their basketball tweets or the tweets on their dinner prep.
      Nova Spivack - My Public Twine

    • dekaysion dekaysion
      10 months ago


      Nova, great analysis. Here's some thoughts.

      Twitters 140 char limit is great as it means spammers can never really overpower anything. You all know the spam comments on blogs with mighty many URLs etc. Not possible here, the volume of spam will need to be divided into many tweets.

      Twitter clients can do most of the filtering. Not just black/white like many clients do these days, but once they learn about tags, about poster profiles (lookup on mr. tweedeck, etc.) they really can shine: what is a tweet worth from someone who has massive stuff auto-piped into the system? Maybe allot, when it is clear the tweet was NOT auto-piped, i.e. contains individual content.

      What I would like to see is a client that analyzes tweets based on tags and word (families). Imagine you are at SXSW and have a tag cloud of the real stuff going on. Not just the #tags, but also other words that get used often, used in combination with #tags. There's a lot of information to be used well. Open the iPhone and see what people really are tweeting. No more individual tweets, but tweet heat maps of concepts, coincidences and co-concurrences.
      Nova, would that not be something to create?

      …alas enough for today.
      Nova Spivack - My Public Twine

    • Nova Spivack Nova Spivack
      10 months ago


      I've updated this article with some points about social filtering as an option as well.
      Nova Spivack - My Public Twine

    • Nick Vidal
      10 months ago


      Instant Syndicating Standards provides a mechanism to solve this problem.
      Nova Spivack - My Public Twine

    • Kenley Neufeld Kenley Neufeld
      10 months ago


      It happened to Facebook...massive explosion of users...and it still seems somewhat usable. Granted, the user numbers for Twitter are astronomically smaller than Facebook. However, the traffic on Twitter is significantly heavier per user than on Facebook (at least in my observation). I post to Twitter 10x a day and to Facebook 1x a day.

      Just adding to the discussion with a few more thoughts. I appreciate Nova's insights here and am intrigued by the social filtering elements. I use Tweetdeck to manage my Twitter feed but the groups are not portable so I need to recreate them on each of the machines I use. That is not workable.
      Nova Spivack - My Public Twine

    • Winston Neutel
      10 months ago


      I don't understand the problem. Twitter only shows me that which I've chosen to subscribe to. If what someone posts annoys me, I unfollow them.

      It's already the case that 99.99% of what's on Twitter is of no interest to me. I don't see how it affects me if that number becomes 99.9999%. It won't change what the people I currently follow post, and it won't change what type of thing I choose to follow in the future.
      Nova Spivack - My Public Twine

    • Kevin Makice Kevin Makice
      10 months ago


      I disagree with the implication that Twitter can't survive growth without adding filtering. Filter is implicit in the most powerful feature in the system: the unfollow.

      For other controls, such as eliminating exposure to @replies, #hashtags, retweets or other structures, the third-party development continues to respond to these individual needs and iterate better ways to improve the user experience. As users, we are not limited to a single interface (the main Twitter website) or even a device. In fact, every individual stream is unique; there are not duplication of content, only overlaps.

      My suspicion is that the people who follow thousands of people will one day realize that there is little value in that. Search of tweets is improving, giving everyone the potential to customize and filter the public stream. Even if that doesn't happen, so what? Twitter allows everyone to make their experience her own. For all of the lists of how-to's and attempts at top-down norms, the bottom-line reality is that no one can craft my follow network but me, and no one can post to my own timeline but me. For those who don't like or grow tired of my contribution, they, too, have the right to exercise unfollow.
      Nova Spivack - My Public Twine

    • 10 months ago


      I wrote a small bayes classifier (a spam filter, esentially) web app for tweets that you can try out at http://tweetvote.infofaktur.de. The classification does good scoring (given that I only have a very small training base) so I actually get some useful recommendations out of it. This is especially helpful to catch up with messages that accumulated as one was offline.

      The main problem here seems to be usability: how can I capture opinions on specific tweets in a very quick and simple manner? How do we show the results of the scoring process?

      I'd like to invite people to play around with this approach: the code is on github, there's a REST API that allows both automated voting and plugging in your own scoring/rating/... approaches. I'd love to receive your feedback. (Also, a nicer name/URL is needed ;-)
      Nova Spivack - My Public Twine

    • david sanger david sanger
      10 months ago


      Twitter really is several things. 1) a public stream (really a firehose) of tweets, each from an identified author 2) a set of identities with followers. 1) and 2) and their history form the dataset they maintain.

      Twitter also provides an API to access the dataset. Their own website is one user of the API but there are many other better.

      So can the system handle a realtime conversation between 7 billion people and who knows how many machines?

      For the underlying dataset it is a technology question. Right now the system produces 5 million tweets per day ( Gigatweet)

      Can the dataset scale to 50, 500, 5 billion, 5 trillion a day ( plus history)? This is basically a database issue.

      Can the API allow realtime access to that dataset? My guess is that each tweet is retrieved many many times.

      It all boils down to infrastructure. Can the dataset and access massively scale? and if not then what happens? does it stop growing, restrict input or access or history, granularize?

      The second main issue then is filtering. Can a reader get just enough info to be useful and not get overwhelmed. I think this is the easier issue. Google handles it reasonably well now for its type of data. There's 600Gpages on California and I can query just fine. Setting up multiple tweetreaders with different filters and filters on filters. Developing reputation and authority screens like some have mentioned above. My guess is that search/filtering solutions will evolve quickly.

      The last issue is the business model. It has to be free, or at least as much of a public utility to be easliy affordable. Advertising though would just add more noise. So the issue is how is it funded.
      Nova Spivack - My Public Twine

    • Bob Calder Bob Calder
      10 months ago


      Nova, what do you say about the dB limitation Sanger mentions? Is there anyone who works in high energy physics data infrastructure that wants to comment? They are using some exotic hardware at ANL.
      Nova Spivack - My Public Twine

    • Dave Sag Dave Sag
      10 months ago


      Fascinating discussion. I am wondering if using digital sigs and making tweets authenticated would help here. A reason Spam is such a problem with email is that senders are never authenticated - anyone can post an email, from any address. If tweets were all digitally signed then senders would be accountable for their posts. I would support a move to authenticated tweets.

      the idea of adding a meta-data layer is good too. too much of my posts are taken up by URLs and hashtags. Like Flickr saw what users were doing with Geotagging and decided to hide the 'machine-tags' to improve the user experience, so too we could hide meta-data from the body of a tweet unless the reader wished to see it.
      Nova Spivack - My Public Twine

    • John Bosley John Bosley
      10 months ago


      Can Twitter survive? Yes. Why, how? Because just like in nature the strongest survive. I beleive that as the system is overloaded, the people who are true fans and community of Twitter will adapt and make it better because they are the ones who trully made it great over the short life span of Twitter so far. Twitter is trully a community based site that follows the needs of the people who are loyal to it. The Tweeps won't let the spammers win!
      Nova Spivack - My Public Twine

      • david sanger david sanger
        10 months ago


        Actually I don't think Twitter is a community. It is a raw infrastructure that facilitates conversation, just like Google facilitates search (and Google isn't a community either).

        The conversations take place among thousands of different communities; in fact each user has his own community of follows and followers, which more or less overlap among like-minded people.

        If this company doesn't do it another one will.
        Nova Spivack - My Public Twine

      • Toby Champion Toby Champion
        10 months ago


        Nah, surely. Not the strongest. Those able to adapt. See for example,
        http://onlyagame.typepad.com/only_a_game/2009/01/myths-of-evolution-3-only-the-strong-survive.html
        Nova Spivack - My Public Twine

        • david sanger david sanger
          10 months ago


          There's Twitter the company and there's the fundamental technology/phenomenon of conversation, recorded, filterable and searchable.

          Twitter the company may survive, fail or adapt. But the phenomenon of realtime digital conversation is here to stay.
          Nova Spivack - My Public Twine

    • Ade A.K.A. Twittergator Ade A.K.A. Twittergator
      10 months ago


      An interesting read, especially for those using Twitter for marketing. How will this effect you and your tweetin'?
      Social Media Marketing

    • jtnix
      10 months ago


      Twitter is a phenomenon to me, and I see all the concerns you have raised to be imminent as well.

      I hope they start with user filtering and grouping abilities ASAP. I would love to be able to group those I am following and regulate the flow of twitters based on time slots throughout the day/week/month and have digests emailed to me for the high volume tweeters.

      I do agree with John Bosley that I think (to some extent) the avid twitter fans will take advantage of the filters quickly to help reduce the noise, and reduce the amount of query and outbound traffic overall.

      I know I will!
      Nova Spivack - My Public Twine

    • Alex Williams Alex Williams
      10 months ago


      Thanks all. As a new user of Twitter I am seeing these potential problems even after only a couple of weeks. It reminds me of the time AOL came online in the early Nineties. Much ignorance and rudeness followed. Hopefully education and information provided to those getting started now will mitigate some of the problems.
      Nova Spivack - My Public Twine

    • Daniel Tunkelang Daniel Tunkelang
      10 months ago


      It seems to me that pure search-based filters are a transient phenomenon, precisely because they are so spammable. This problem isn't specific to Twitter--it applies more generally to alerting (see http://thenoisychannel.com/2008/10/13/alerting-push-or-pull/ for an example of spamming via Google alerts).

      Where Twitter does best is in explicitly choosing whom to follow. But of course that is limiting. More sophisticated filtering mechanisms would be nice. I'm not fan of follower counts, though I have proposed what I believe to be a more robust influence measure (see http://thenoisychannel.com/2009/01/13/a-twitter-analog-to-pagerank/ for details and http://tunkrank.com/ for an implementation). In any case, as Clay Shirky tells us, the problem is not information overload but filter failure.
      Nova Spivack - My Public Twine

    • Christopher Rines
      10 months ago


      Better filtering for multiple communication streams is coming, it's one of the areas we will see a lot of movement (finally) this year.
      Nova Spivack - My Public Twine

    • Jon Garfunkel Jon Garfunkel
      10 months ago


      Nova-- Just to expand on what I wrote above... I have dubbed these sorts of messages "flam" -- see http://bit.ly/flam

      "Number of Followers as a Filter." Loic Le Meur proposed this in December, and got beat down for it, since it would perpetuate A-listism.

      "This is going to require metadata in Twitter, ultimately." Yes, that's what I propose at http://bit.ly/ssmc , with a working example at http://bit.ly/Star-Priority -- which simply uses codes like *Q0 (question, answer anytime) to indicate their priority.
      Anybody interested in semantic social web should be help contribute to this. And then it's up to the Twitter clients to implement this.

    • istara istara
      10 months ago


      This is partly why I keep my Friends list limited to people I know, and those who post interesting Tweets but not too many. I started out adding a lot of news feeds, but ended up unsubscribing to many because the volume of updates was too high. If I used Tweetdeck or similar, then a high volume might be easier to manage. But I mainly use a couple of iPhone clients so I don't want too much noise.

      Re-Tweeting Activity as a Filter - sadly I see this being very easily abused, very easily botted. It will be just like Google/Blogger link spam. So easy to set up a few dozen accounts and automate them re-tweeting one another.

      One feature I would like is some way to tweet "premium" or "public" tweets versus "private" or "personal" tweets. That way if you wanted to twitter your every thought, you could do it, but keep it out of your main stream of public tweets. Some blog services have Public/Private/Friends Only options. That would be very useful for Twitter. Friends could also perhaps choose whether to monitor your non-premium tweets, depending how close they really were to you, and how much they really cared about what you ate for breakfast/what toothpaste you used/how long your morning commute was/etc.

      I see a split as something like this:

      Premium: tweets on interesting links, interesting photos, unique thoughts, original observations. The kind of stuff a normal person would take the time to write an interesting blog post about if they had more time.

      Personal: tweets on the minutiae of your daily life, your frustrations with minor things, a hi-score you just got in a game, a film or TV programme you just saw. The kind of stuff that is of little value or interest to anyone except the person it happened to, and possibly a couple of their friends or relatives.

      This would also help professional twitterers who have work colleagues and clients and so forth reading their tweets. You can let people know about your products, your research, etc, and keep a more clean and professional Twitter profile/image by having your personal life tweets go to select personal friends. Currently one needs two accounts for this, which Twitter doesn't make very easy.
      Nova Spivack - My Public Twine

    • Jacob Friis Saxberg Jacob Friis Saxberg
      10 months ago


      I agree, but the solution could be universal. Companies like Google should be able to fix this with their approach to spam.
      Nova Spivack - My Public Twine

    • roweben
      10 months ago


      Very interesting post Nova,

      Filtering could help, but ultimately I'm worried that Twitter could collapse simply because it becomes too mainstream.

      I propose another solution: I actually think it might be better if the folks at Twitter stopped allowing new accounts. Limit new users from joining, and preserving Twitter as it is today. Here's my recent post about it (business model included)

      http://benrowesblog.wordpress.com/2009/03/05/help-me-save-twitter/
      Nova Spivack - My Public Twine

      • Naomi Most
        10 months ago


        That seems like a sure way to kill Twitter. How do you think the average Twitter-user would feel knowing that no one else from their social groups will ever be allowed to join them? At this early stage in the game, I'm willing to bet that a huge proportion of people on Twitter are patiently hoping that more friends and family will join them on Twitter. Not everyone's in it for the celebrity worship.

        There's a reason why our culture more or less equates stagnation with death.
        Nova Spivack - My Public Twine

    • Daniel Tunkelang Daniel Tunkelang
      10 months ago


      Ben, if your only concern is filtering, then allowing users to be able to filter based on account creation date would obviate a benefit from disallowing new accounts. As for the business model of creating a marketplace in scarce accounts, I think Twitter could do better by monetizing premium access to the unprotected stream.
      Nova Spivack - My Public Twine

    • Jon Garfunkel Jon Garfunkel
      10 months ago


      This is weird. You posted a version of this article yesterday, and here was my response:

      Nova-- Just to expand on what I wrote above... I have dubbed these sorts of messages "flam" -- see http://bit.ly/flam

      "Number of Followers as a Filter." Loic Le Meur proposed this in December, and got beat down for it, since it would perpetuate A-listism.

      "This is going to require metadata in Twitter, ultimately." Yes, that's what I propose at http://bit.ly/ssmc , with a working example at http://bit.ly/Star-Priority -- which simply uses codes like *Q0 (question, answer anytime) to indicate their priority.
      Anybody interested in semantic social web should be help contribute to this. And then it's up to the Twitter clients to implement this.
      Nova Spivack - My Public Twine

    • Craig Jolicoeur Craig Jolicoeur
      10 months ago


      If you are looking for Twitter filtering, check out the Twalala client (http://twalala.com). It's "Twitter with a mute button."
      Nova Spivack - My Public Twine

    • Imran Anwar Imran Anwar
      10 months ago


      Though you may overstate a few "problems" (complexity of Twitter {far easier than MySpace/FaceBook for novice to use}), you make some excellent and valid points in the solutions.

      That is why this is a great analysis, more focus on real solutions that are in consumers and Twitter's interest.

      Good work. All the best.

      Imran
      Will Your Life's Work "Live, Forever"? http://neternity.org

    • Pamela Jacob Pamela Jacob
      10 months ago


      I feel that way about Facebook. But the one good thing is we can follow and unfollow so we really can cut down on the people that we feel are only in it for them...
      Nova Spivack - My Public Twine

    • Concetta Phillipps
      10 months ago


      I found your article very interesting, but I think what you're asking for is highly subjective. For example, problem 1:

      "Hypertweeting. Some Twitter users tweet legitimately, but far too much. Or the content they tweet is just innane. In doing so they market themselves and dominate everyone's attention with their presence."

      How much is too much tweeting? Is it one post, or ten, or a hundred? Someone like me who posts 10-15 times a day - would that be too much? Its a mix of personal "what are you doing" with retweets, conversations and added value things such as article links.

      For you, I might be a hypertweeter. For some people, I'm not even close. For twitter to be able to deal with that, it would have to have some kind of rating cloud where you could rate the value of a post so you could see more or less of them (similar to the old Facebook system that's now gone).

      Regarding #hashtags - I think the question is on how many people actually use them and search for them. I don't bother with them in most cases because I have followed the people I know will post useful content and I don't need to worry about finding something in the ocean of hashtag responses.

      Hashtags seem to be holdovers from other forms of communication like blogs, where you would click through a multi-subject blog and want to find only one particular topic. With so many specialized Tweeters/Twitterers, I don't see the need for them in many cases.

      Number of followers definitely doesn't work as a filter because of the amount of people gaming the system. I've noticed that if I take time to investigate a particular Twitterer/Tweeter, say more than 1 day, I'll already be unfollowed by many people because they are looking for only people to auto-follow them so they can inflate their follower count. Other topics going around this include getting on the "suggested" list of people to follow so that you can get millions of ignorant followers who are brand new to the service.

      Notification is probably the only filter I use. If someone is only posting links to their blog, I'll unfollow because I more than likely already read their blog and know that there was a new posting. But there are services that I follow because the Twitter account is a better notification than other systems.
      Nova Spivack - My Public Twine

    • shane mcloughlin shane mcloughlin
      10 months ago


      There are other consequences of openness on twitter too. Here's the blog http://relativemusings.blogspot.com to read it. Basically, the openness of data on twitter leaves the door wide open for anyone to analyse, interpret and capitalise on your data in any number of ways. Twitter needs to look at the bounds of privacy in addition to the issue of filtering.
      Nova Spivack - My Public Twine

    • Keren Dagan Keren Dagan
      10 months ago


      I think that since Twitter is so open it will also be easy to find whom not to follow (and by that filtering your personal tweetline). Twitter communities can find ways to keep there channels clean.
      Adding metadata to each tweets does not sound like a scalable solution to me. It does sound like an interesting solution for #hashtags and Trending topics. In many cases it takes time to find the context for these tags.
      Keren
      Nova Spivack - My Public Twine

    • Maddie Grant Maddie Grant
      10 months ago


      Another possibility is that we'll all start to protect our updates so only people we really vet can see them - which would be sad, but maybe necessary. The good thing about social media is its' flexibility in adapting to problems like these. I think we just have to keep fingers crossed that the developers will stay one step ahead.

    • maggielmcg maggielmcg
      10 months ago


      This is a great, really thought-provoking post.

      The one saving grace of Twitter is that you can choose who to follow and therefore control the number of tweets in your stream. Therefore, to some extent, at least, people can control how much "noise" goes on at least as far as they are personally concerned. Granted, you can't control "@" messages from people you don't follow, but you can block a person who sends you direct or @ messages if you don't want to hear from them.

      The one positive in the realities you point out is that maybe it will eliminate--or at least lessen--the numbers=popularity dynamic of Twitter, where the more followers you have, the higher up in the "Twitter elite" you are. Twitter is in no small part about vanity and wanting to feel like people want to listen to you. So if people stop the practice of following just so they can be followed, they may be more inclined to be very selective about the people they follow.

    • Bill Bittner Bill Bittner
      10 months ago


      I don't understand. In your main view, you only see tweets of those you follow. So don't follow spammers or the inane.

      However, the one problem I do see now is the amazing number of new follower notifications I get in my email. And most of them are "consultants," "coaches," etc. Was their a life coaches convention recently where they said to follow everybody on Twitter?

      But, I must be fair, most of these followers were not spammers. They may try to promote their service, but at least they waited every 20 tweets or so.
      Nova Spivack - My Public Twine

      • Ss Sss
        10 months ago


        why not turn off your notifications? that would seem to resolve that problem

        and some self proclaimed gurus have written articles that recommend following everyone who follows you - so its a big game of follow, and for some follow/unfollow/follow/unfollow to game the rankings

        the bigger problem i would argue is that Google will eventually buy Twitter, they have all but suggested that recently with their classic sorta denial, but they will; and whence they do an enormous amount of data from email to health records to purchases to internet searches to phone to maps will give someone a nearly complete view of some individuals - this is 100x more scary than any Patriot Act

        sure, Google will add all the features that Twitter refuses to add and stabilize the infrastructure but it will give them unprecedented access to an array of data unseen by one entity in history, even the govt
        Nova Spivack - My Public Twine

    • pallian pallian
      10 months ago


      I think Twitter will survive and adapt pretty quickly. With celebrities endorsing it non-stop these days, and with our capability to embed twitter groups on any website and blogs (http://www.tweetizen.com/) - it'll be so mainstream, ppl will wonder how they lived without it in the first place.
      Nova Spivack - My Public Twine

    • Casey McKinney Casey McKinney
      10 months ago


      Good post, trying to figure out twitter and the difference between status updates on facebook now and this, and if I should do separate posts for each, but yeah some of the iphone apps, they are almost useless because of the lack of filtering, so it'll be interesting how it evolves.
      Nova Spivack - My Public Twine

    • .LAG .LAG
      10 months ago


      there are some great ideas in your initial post, nova, and in the sheer volume of comments, clearly this is a topic of interest to many. ultimately, the goal for everyone is the maximize signal and minimize noise. sounds like an opportunity for someone -- or many someones -- to solve this problem.
      Nova Spivack - My Public Twine

    • Rebecca Johnson
      10 months ago


      I completely agree, and with Twitter looking for new ways to monetize their service it's not long before we're exposed to paid advertising on our Tweets.
      Social Media Marketing

    • Tim Stiffler-Dean Tim Stiffler-Dean
      10 months ago


      What if it doesn't happen? Then we have to turn to alternative services or suffer with the insanity that twitter is quickly becoming. I doubt that the developers of something so simple as Twitter will want to add in such advanced ideas as metadata and filtering. TweetDeck (and being able to separate friends into groups) is quickly becoming the way to remove much of the noise, and I feel like that's about the only thing that can be done.

      The awesome thing about twitter is that it is so simple to use that anyone can get it. It doesn't take more work than to type out a thought. I feel like adding more of the features that you're talking about would cause twitter to become... twine. :)
      Web Industry Trends

    • Alex Schleber Alex Schleber
      10 months ago


      I've been experimenting with piping the "with friends" RSS feed into Thunderbird (Mozilla's Email client), and doing some Filtering Rules stuff on keywords, etc. in it. Thereby serving to alerted to AND to get archiving of all tweets by keyword, and from there on down to folders of “tweets with links” and “RTs".

      Basically going from the very specific to the non-specific, and ending up with a residual pool of keyword-less, link-less, RT-less tweets that can still be sliced & diced for research purposes, but that has to be deleted every so often to not slow T-bird down too much (I’ve had up to 150k tweets in that residual folder, and it was getting too slow to search).

      Occasionally you will come across a new keyword to filter for, or the occasional nugget that I would classify under "quote-worthy". It will comfortably hold 100k tweets at a time, but then starts to get slowish for searches (on my average laptop). They'll still work though, and at about 250MB memory usage holding nearly 200k tweets, the app itself does not crash like Tweetdeck tends to. Extra bonus: no issues with the API limit running out before the hour is up, and tweets being "missed".

      So far I haven't seen any better solution, though Thunderbird could stand someone creating a special plugin to enable REGEX search on tweets, and integrate status page lookup (HTML rendering) plus a few extra tools like RT links a la Tweetdeck. Because in truth even the tweets of just 100 or so active people one might follow are too much to fully attend to. Let alone 1,000 or 10,000.

      What is however relevant is the distinction of tweets by keywords coming from people you are following, or Twitter overall. E.g. for a brand I would recommend following back all of your followers, and use the above or similar method to see what they are saying about your brand/company, rather than just listen to the complete Twitter timeline via search.twitter.com.

      Also note that following people by keyword is NOT the same as ignoring them (you still want to reply to @replies and thank for RTs etc. of course).

      Another thing to watch in this regard is the question of whether Twitter will ultimately move to shut off backwards access to the full timeline via Search.twitter.com, e.g. provide results only say 7 days back. This is already happening during daytime heavy loads, presumably for stability reasons. But they could make this permanent at any moment, in order to charge for access to serious corporate/etc. data miners.

      In this light, archiving both your own and your friends ("folllowing") tweets is taking on added significance.
      Nova Spivack - My Public Twine

    • Michael Lewkowitz
      10 months ago


      Nova - late to this post. Love the depth you've hit with this and am particularly interested in the meta-data, inline tagging, and filtering. We've just put http://thread.io up as a public alpha. With it we see tags as threading twitter while the links weave the web. There are clearly challenges of the public micro-message container but also think some things in it represent the potential to become a new foundation for distributed collaboration and collective intelligence. I've posted some thoughts on that here http://igniter.com/post406 and http://snurl.com/svc-preso-post. If this is true I'm excited about what it means for our future. If not... well it will be a fun experiment. Cheers! (@igniter on twitter)
      Nova Spivack - My Public Twine
    Add a Comment
Report This

Twine is about discovering, collecting and sharing the content that interests you. Learn More

Join Twine

Stats

First Posted By

Nova Spivack

First Comment By

jebworks

Tags

Community Tags

Who's Interested In This?

Forgot your password?