05 Nov MIT Researcher Predicts Twitter Trends, Destroys Space and Time
In “holy ****” news (to me at least), an MIT researcher has built an algorithm that can accurately predict Twitter trends well before Twitter’s own processes pick them out. While not quite seeing the future, if this technology pans out, it’s going to pretty freaking close to paranormal – as Arthur C. Clarke once said, any sufficiently advanced technology is indistinguishable from magic.
This breakthrough obviously has huge ramifications for how Twitter approaches advertising associated with trending topics, so it’s highly likely that Twitter will snatch up the algorithm before some like, oh, say Google, buys it for a cool billion dollars. No coincidence, however, that the GigaOM article points out that a Twitter employee was a grad assistant for the primary researcher.
Ignoring Twitter for just a moment, if this algorithm works on a grander scale (and Twitter is pretty damn grand), imagine what this could mean for stocks, commodities and any dynamically changing price. To nerd out for a minute, imagine if this could be used across all volumes of dynamically generated data – seismic, climate, health-related, etc. – to accurately predict natural disasters and other catastrophes before they happen. We’re talking a level of forewarning well beyond the dreams of Minority Report.
I’m not saying that this algorithm will get that far – it’s likely that when exposed to the full volume of Twitter it fails – but imagine if this is the first turn of the key that unlocks data integration on a global scale. The world as we know it would change overnight…and Twitter would be responsible. How’s that taste, Google?
What do you think? Is this just another Twitter gewgaw or a herald to a deeper understanding of data?