There’s nary a digital service or platform that doesn’t touch the cloud in some form or fashion. From file sharing services like Google Drive or Dropbox to web hosting and remote server setups from the likes of Microsoft Azure or industry world-beater Amazon AWS, the cloud pervades everything we do in our digital lives (which is getting harder to delineate from our physical lives, but that’s a post for another time). It provides computing power and services to small businesses that might not be able to afford or possess the technical acumen to assemble the physical infrastructure required to make a digital business thrive (or even exist for that matter). But, the massive business opportunity the cloud represents has the relevant tech giants plowing full-steam ahead to gain a competitive edge over their cloud rivals.
The next frontier they’re turning to? Why, artificial intelligence of course!
In much the same way small- and medium-sized businesses often don’t have the resources or acumen to build a digital infrastructure in house, so too do those same businesses lack the access or ability to build tools capable of tapping into real A.I. horsepower. It’s simply too nascent a field with too high a barrier to entry. But why does it matter in the first place?
Because the forward-thinking companies are going to start demanding A.I. functionality be integrated into their custom software solutions, and developers will need access to the tools and resources to deliver on those requests.
Now, that may not impact all that many companies right now, considering the market for A.I. development tools isn’t exactly huge… yet. But as companies start to see their competitors — specifically the innovators within their respective sector — partnering with A.I. experts to bake A.I. functionalities into their digital offerings? The demand for open-source, affordable A.I. development environments will skyrocket — hence all the major players throwing their hats into the proverbial ring.
When we say the barrier to entry is high, what we primarily mean is A.I. in general, and deep learning/neural networks in particular, require a huge amount of a specific type of computing power to undertake. And most companies, even cutting edge software developers, often don’t have the resources to build the physical systems necessary to test A.I. algorithms and teach a neural network how to do a specific task a client might want/require. But if the tech giants who already own the cloud can integrate A.I. functionality into their respective cloud offering, that will certainly entice developers and the companies they represent to move/keep their respective business with the best comprehensive cloud provider.
According to MIT Technology Review, Amazon and its AWS unit is leading the way in this regard. In the last year, Amazon unveiled “Amazon Cloud 9, an integrated development environment (IDE) that plugs directly into its cloud platform. It also announced a host of new AI tools that can turn speech in audio files into time-stamped text, for example, as well as translate between seven languages and track people, activities, and objects in video.”
Google might lag behind AWS and Microsoft’s Azure, but it’s making an A.I. market share play itself with TensorFlow, which is essentially open-source AI software capable of building other machine-learning software. According to MIT, since TensorFlow went public, it’s become “the AI platform of choice for many developers, and it underpins many new artificial-intelligence projects.” To that end, Google created its own chips, called Tensor Processing Units, or TPUs; they’re designed to efficiently process TensorFlow and cut down on energy needs.
Of course, Microsoft and Amazon aren’t sitting on their respective loins while their biggest cloud competitors pick up steam. In fact, they’re teaming up: “The two launched an open-source deep-learning library called Gluon that works a lot like TensorFlow and is meant to make it as easy to build and train neural networks as it is to make an app.”
The market is clearly moving in this direction — with the speed of adoption in custom software, it’s only a matter of time before A.I. becomes commonplace. The cloud provider that helps usher in that era may very well be the biggest winner of them all.