At Friday’s Techcrunch Real Time Conference (crunchup), there was a panel on real time search. Several of the commentators argued that real time search was useless without filters and algorithms to refine the search. Philosophically, that approach is similar to Dick Cheney saying it’s okay to wiretap ordinary citizens without any accountability or disclosure because Dick Cheney said so. What these panelists were arguing for was subjective filtering of what people post to social networks based upon what they view as important, relevant, and worth including.
Twitter seems to be leading the charge on this approach. Over the past few weeks, many of the people I follow closely have been unpleasantly surprised when they try a search for posts made from their username (e.g. FROM:karoli). Today was appalling. Many people I consider to be interesting and genuine were removed from search. Actually, according to @delbius, the Twitter spam monitor and public voice, they were ‘filtered’.
One of those ‘filtered’ was my friend Liz Ditz. She was subsequently restored to search, but as you’ll see, the restoration was only for posts after she appealed to Twitter for reinstatement. All of her old messages are gone, evidently for good. Here’s screen captures illustrating the point:
Note that Liz has over 5140 posts to twitter. She follows less than follow her, and in any event the follow ratio is well within acceptable bounds. Here are her search results, captured 10 minutes ago:
That’s it. That’s all of her search results. Over 5,000 tweets and one single page of results. Yes, Twitter “restored” her account, but nothing before 4pm can be accessed via search.
Here’s another example, even more chilling. @zorinsmom tweets information about Iran. It is her mission, as stated on her Twitter profile. She has posted over 3,000 updates. She is focused on the Iran election, and nearly everything she posts is an effort to relay information that promotes a free Iran. A noble mission, right? Have a look at her search result:
That’s right. No results. If you search on the hashtag #IranElection and zorinsmom, you get updates directed TO her but nothing FROM her.
Zorinsmom has been ‘filtered’. Despite her best efforts to get information out onto the network, she has been caught in the net that is Twitter’s subjective algorithm for filtering search results in such a way that certain people are simply….gone.
Imagine the power plays here. Imagine the ability to shape search by purely subjective standards to shape the flow of information. Go ahead and accuse me of being paranoid, and then look at Zorinsmom’s timeline. Or Liz Ditz’s timeline. Both of them are concerned citizens using Twitter to get a message out. In Liz’s case, it’s about autism. In ZorinsMom’s case, it’s about the Iran elections. In BOTH cases, they are legitimate, concerned, engaged users who are hardly spammers. One of my most trusted favorites, RoadKillRefugee was also ‘filtered’ (only 3 pages of search results for an account that has some of the most reliable and interesting links around the political scene there are…thanks, twitter.).
So Twitter has decided that spammy promos for ‘free followers’ (as if anyone actually pays for them?) is fine to leave in search, but legitimate, concerned citizen activists are not? It’s been clear for awhile that Twitter’s business model involves high profile celebrity, including the requisite appearances on Oprah, Conan O’Brien and other shows, but the decision-making process at work here is simply outrageous.
Further, they don’t seem to understand what community is on Twitter. They don’t understand the value of trustworthy links being spread through the network. They obviously have no concern for the value of the one-to-many model when it comes to sharing relevant and important information about really important issues. Twitter’s concern is whether Ashton Kutcher has a million followers or meeting Warren Buffet.
What they have done is take an inherently democratic model and turned it into an autocracy. The standard response is that a link has been posted that Google has identified as ‘malware’. I clicked through as many archival links as I could on these accounts and could find exactly zero links that Google flagged as malware. Not even one. But even if I did find one, shouldn’t users receive some sort of warning that they’ve posted a link rather than simply being ERASED, obliterated, and sent to Twitter prison?
This is a serious enough issue to me that I am seriously considering posting from identi.ca and bridging to twitter. I see absolutely no excuse for wielding power over people like they are in such an arbitrary and capricious fashion, with no appeal and no recourse for past updates. They’re just gone. Links and information that are of value. I’m not talking about the “I had fish for lunch” updates. I’m talking about updates that really ARE valuable. Roadkillrefugee had posted a series of tweets reporting the contents of the newly released report on Cheney’s involvement with illegal wiretapping. Liz Ditz was posting updates with important autism activism information. Zorinsmom was fighting for Iran and fair elections. Yet, in each case, Twitter arbitrarily ‘filtered’ their voice.
We need tools so we can get rid of asshats posting crap about free followers, not so we can silence voices using these tools for good, not evil. More importantly, decisions about what is filtered should not come from the top down; rather, they should come from us, the user, deciding what WE think is relevant in a search.
Shoq (who has been ‘filtered’ more than once) has written a complete step-by-step procedure for checking your own Twitter result and then taking steps to fix it if you’ve been the victim of their ‘filter’. I highly recommend bookmarking it. Chances are, you’ll need it.