While SXSW isn’t where tech launches, it’s where it typically gets key momentum. Twitter. FourSquare. Vine. Meerkat. While none of these launch at SXSW, they all gained their initial momentum here. This year is a little bit different, the key topic of discussion in panels this year isn’t a single application, but a much broader piece of technology – artificial intelligence (AI), and rather than gaining momentum people are talking about how they’ve been using it for some time. But first things first – how did we get here?
According to Inmar Givoni, there were three key factors:
1. The increasingly widespread availability of hardware: while we’ve historically been limited by the amount of computing power that we can afford to buy, access to cloud-based servers (also known as “someone else’s computer”) means that it is now possible for companies, including startups, to scale up to the required level of computations that AI requires both quickly and cost-efficiently.
2. Massive amounts of training data: while it is essential to AI to have access to machines with sufficient computational power, all the hardware in the world will be for naught without training data. Lots and lots and lots of training data. With people increasingly storing their files in the cloud, and importantly with people tagging those items with meta data (everything from descriptions of their cat photos to the genre of the songs in their music collection) there has been sufficient data for companies to develop their AI programs (so that’s how Facebook recognizes my friends when I upload photos of them!).
3. The availability of software and algorithms: while one size does’t fit all, the sheer volume of software products and services that are now available off the (virtual) shelf have again lowered barriers to entry in the implementation of AI.