The Cloud appears to be spreading to all of our latest and greatest technological creations, but what if it did more than just store data in a convenient place?
Remember HAL from 2001: A Space Odyssey? Or Gort from The Day the Earth Stood Still?
What about David from Spielberg’s A.I. or the rain-soaked dystopian jacket of Roy Batty in Blade Runner? Aliens’ Bishop, Star Wars’ C3PO, Terminator’s T-1000 or Short Circuit’s Johnny-5, perhaps?
From 1927’s Metropolis through to recent blockbusters like Ex-Machina and Chappie, humanity has painted a singular vision for the future of artificial intelligence (AI) in its popular culture: A lone, tangible, real world construct.
But all those visionaries got it wrong – that is not where the future of AI appears to be heading.
As network infrastructure improves and devices become more connected with each other, a machine’s AI need not be tethered to and limited by one physical construct.
AI can exist in the cloud, and reach across the whole world.
It’s become one of the most used terms in modern technology, but many consumers don’t fully understand “The Cloud.”
Not that long ago, when you were operating a computer or mobile device and performed an action – such as opening a program, playing a game or reading an email – the computer looked for that information on its hard drive.
When the internet came along, you could go online to perform actions – rudimentary activities like searching for information, or pictures, or perhaps watching a video.
While these devices would connect directly with remote computers – called servers – to pull down this information, the heavy-lifting was still being done on the local computer.
The local hardware was where the data was being understood by the AI and displayed.
However, as network infrastructure has improved, more complicated activities have been able to occur remotely, yet still be displayed on your home or mobile device without the lag that would make the experience uninviting or unwieldly.
In addition, these servers can be accessed and manipulated by multiple remote users simultaneously.
This is The Cloud.
It’s not literally a collection of particles floating somewhere above us, but a collection of mammoth data centres, located all around the world.
And essentially it allows your device to see and engage with files remotely and, potentially, beyond its own internal capacity to run.
For example, you don’t need to have Google Sheets installed on your computer – you can run Google Sheets directly from the cloud, and your data is stored there.
If your computer wasn’t powerful enough to run Google Sheets, and you didn’t have enough hard drive space to save a document, it wouldn’t matter.
Google Sheets is being displayed on your device, but the program is being run on a super computer on the other side of the world with no shortage of storage space.
What if the AI that understands and interprets all the data wasn’t confined and bottlenecked by your single machine?
What if it also had no specific hardware confines and existed in this databank?
The next evolution of the cloud is on-demand AI, sometimes referred to as “AI as a service.”
The four major players in this space – Google, Amazon, Microsoft and IBM – are deep into their research and development on what is possible with cloud AI.
Early examples you may have already used include Apple’s Siri, Google Assistant (previously Google Now) and Microsoft’s Cortana.
Cloud AI goes beyond the idea of simply storing data or hosting programs that run on a remote device.
It can actually analyse the data, make its own decisions and define what gets displayed to connected devices.
Image analysis and speech recognition are two such forms of input these cloud AI services currently evaluate.
Able to store and trawl through phenomenal amounts of data – way more than any commercial computer or mobile device – to answer questions asked of it.
But where things get exciting – or scary, depending on your perspective – is the current push to bring sophisticated machine-learning to cloud AI.
This is where the AI can learn without programming simply from the data it receives.
Trend forecasting and translations are some processes that have begun to explore machine-learning cloud AI, improving their accuracy with every user query.
We’re only just seeing the tip of the cloud AI iceberg at present, and its application is expected to expand dramatically in the coming years.
As broadband infrastructure improves, Australians will become more connected with the cloud and able to not only send, but receive, larger amounts of data for analysis.
This is great news for local businesses, who can invest less in onsite technology, and fantastic for consumers of AI-driven content like video games.
It has the potential to improve day-to-day life by offering fast, accurate and personalised information on demand.
Just don’t call it Skynet…
There are already a variety of personal digital assistants that rely on The Cloud. Check out a quick summary of three of the most popular ones available right now, and see how they compare.