This year it seems pretty certain that Twitter will start completing the necessary paperwork for an IPO in 2014.

Many people do not realise that Twitter as a business is almost 7 years old. I was one of the first waves of people to have an account due to someone I knew who was hearing a lot about Twitter at the time in the Valley.

In his words, this is going to be big. During the 7years, I have fallen in and out of love with Twitter. At times using it a lot and then not at all for months. In fact in November I actually deleted my account, simply because it is such a public forum and, to be honest, while there are people who use Twitter in the sprit it was designed there are a minority who do not. Plus I am not entirely sure how many people actually care that I am at a conference or want to see me tweet a random, most likely dull, photo that means nothing to anyone except me. Don’t even get me started on Facebook…

It has got me thinking though, where does Twitter’s suggested IPO valuation of $15 billion come from? Where is the true value?

Granted it was unknown to me until recently that Twitter revenues were some $80 million in 2011, $250 million in 2012 and are expected to be in excess of $500 million for 2013.

Twitter uses human intelligence to make sense of its search result data. As a supporter of artificial intelligence, although it is still worth remembering that we are still at a point where us humans actually provide the best view.

Twitter has matured a lot in the past 7 years. From a simple post by a user, to changing the way users share news or even rebel against the establishment.

With more iterations of product and changes to conversations between users and what they expect, it’s been key for Twitter to make sure the right information is being delivered into the right streams.

So I now see a key challenge for Twitter if they ultimately want to sell users advertising. And its one that has challenged the majority of developers - how do you get a computer to understand what a user is saying or typing? How do you make a machine understand emotions and context?

To get to artificial intelligence, as suggested in many a movie , the technology space needs to continue to create improvement efforts to solve the issues of taking human thoughts and emotions into a way or format that can be used by machines. Going back to Twitter, how do they know what ads to place against which hashtag.

So I was interested to learn that, right now, Twitter does indeed use humans to make those decisions. Not sure why I was surprised by this, but maybe I never appreciated the scale of twitter and it’s back office.

As you can see from Twitters own blog here, they have an automated process which farms out work to humans.

Twitter’s systems are great at picking up spiking, but what does the context of the spike actually mean? Hence the involvement of a human, in an automated process that is far cheaper right now than using a computer to figure it out. What a human will not be able to do though, is plough through millions of tweets to see for a short period of time, the height of the impact of particular spikes.

Maybe the way for AI to become better and grow is actually not to try and make humans redundant, but to look at what they are good at and how it can help the process of the human making a decision.

I also asked myself a question, would Twitter be good at breaking news stories? Twitter could take a subject that was trending and potentially develop news pages, far quicker than a human editor could. Automation of the breaking news space, could be a strategy for monetisation in the period coming up to any IPO.

One of my more ‘out there’ thoughts for Twitter is around the gift card and discount space. Having previously suggested that Apple may try and turn iTunes into a way of purchasing anything from a burger to a tickets soon, taking into account the patent filed in December 2012, Integrated coupon storage, discover and redemption system", there is a possibility too for Twitter to join up with pushing deals via passbook or allowing users to buy gift cards, via Twitter. Its probably a crazy idea, but its one that came to me while piecing this blog together.

One thing that is clear is that maybe Twitter is the tortoise in a race with the hare, but slowing is going to win the race against its competition.

Enhanced by Zemanta

Find your next job with techworld jobs