the codist - programmerthink

Why I Don't Think Twitter Has a Future

Published: 12/12/2011

It's not that I don't understand the appeal of Twitter and its ability to rapidly spread communications all over the world. I can see how such simple idea makes talking with large groups of people so easy. I even have this blog posting to a Twitter account which I also occasionally add things to as well.

Twitter however has no future in the long run. Sounds pretty stupid to say so but I am sure of it.

The biggest problem it has is exactly what its strength appears to be: the simplicity of the 140 character text message. What can you put in 140 characters? Clearly you can pack quite a bit of information in, with shortcut links and simple #hashtags. The simple text format allows Twitter to rapidly move the information from poster to followers in humongous volumes. Yet the simpleness of the chunk of text will ultimately be its undoing.

Recently it became clear that people in Russia created thousands of Twitter accounts simply for the need to poison the common hashtags people were employing to criticize the government. Tweet enough messages with those hashtags people might be watching for and the real message becomes lost. What can Twitter do to battle this? Not much.

A tweet is not a formal protocol, it has no syntax to speak off other than convention. The whole Twitter concept was an accidental side effect of a different idea entirely. This was the charm of the idea that anyone could broadcast snippets of information and have people read them. We've seen revolutions tweeted in real time and governments watch helplessly to control the flow of information and even try to shut down the entire internet as a last resort.

These last two points seem to tell me that as people in power or people who wish to control their products or idea might use the simplicity of getting a Twitter account combined with powerful tweet generation software and a list of common hashtags to poison the stream with false or random information, as the people in Russia attempted to do. Furthermore I can see people adapting to pushing misinformation or even spam onto the Twitter stream. Imagine if #ows become flooded with Viagra ads and followers duped into following an apparently well-known topic which then switched to all spam all the time, filling up their twitter feed with garbage and making picking out the meaningful stuff impossible.

I am sure that Twitter does have some automatic system to try to deal with this but in the end it will be beaten because there is no real barrier to entry, just like sending emails in bulk is easy. Sure, people come up with ways to stop spam but the spammers will continue to find ways to get through as long as it is profitable to continue. In the case of tweets, especially in countries where the government is wary of its citizens (which is even becoming an issue in the US), the desirability to control what is tweeted is simply too tempting and doesn't even require a whole lot of programming skill.

Imagine paying thousands of people in a poor country to create an email address, obtain a twitter account, and then pass the login information to some third party. The "owner" of these accounts can then build some kind of software which generates tweets which either contain the targeted tags or at least have useful information to attract potential followers. Once people are hooked and some event takes place you want to confuse your massive mis-tweets begin, perhaps all sufficiently different to blunt any filter Twitter might be using, leaving the followers to miss potentially useful information or to render a common hashtag unreliable. Once this works it will be used again and again. The usefulness of posting and following information becomes less valuable the more noisy and unreliable it becomes.

I think of this as a DPTS attack (Distributed Poisoned Tweet Stream). What can Twitter do to combat this? I am sure they have plans in place but like that with email and spam, it's a never ending battle and they are really the only one in position to defend the Twitterverse. At least with email the defenders are themselves distributed and email is not where their value lies. Twitter really has nothing else to offer to support its 8 billion dollar valuation but tiny bits of text being pushed around. If those tiny bits of text are no longer meaningful or reliable or even useful Twitter will vanish as well.

You would think that having control over every tweet that exists would give them an incredible ability to thwart the machinations of the bad guys but the question is where do they draw the line? They need to have clear value to lots of people and not wind up drying up the sources and sinks of tweeting in the name of keeping the stream relevant. So either you filter too much or not enough; somehow I think the likelihood of doing it right is pretty low. Especially if governments really start to invest in poisoning the stream.

If you look at Siri and IBM's Watson's ability to create meaningful sounding dialog and combine with a large number of Twitter accounts it doesn't take much imagination to see how you could create a wall of tweets that look exactly like the people you are trying to drown out or discourage, especially for singular events like revolutions or even bad press.

When that happens, Twitter is toast.

Tag:

submit to reddit submit to hackernews