Cloudy With A Chance Of AI
Artificial Intelligence’s as-a-service makeover. . . & it’s all in the cloud
Cloud computing first began as a way to access cheap computing power and data storage, allowing users to rent space in data centres via the internet. Now, the market is undergoing a transition. In China, Alibaba, Tencent and Amazon are just a few of the companies vying to dominate the scene. In the West, Google and Microsoft are attempting to move in on Amazon Web Services. To make their cloud services more competitive, tech leaders are developing new cloud computing tools that can be used by developers. But what has this got to do with Artificial Intelligence, and will it lead to the as a service adoption of AI?
The consumerism of cloud computing
Things are heating up in the cloud. In a recent report, Gartner predicted that revenue from worldwide public cloud services would reach $260bn by the end of this year. By 2021, this figure is expected to rise to $411bn. But why are so many organisations transitioning to the cloud? One of the main motivations is monetary. There’s no need for companies to build their own data centres, and therefore no need to invest in hardware, software, utilities and expertise. The cloud also enables a flexible, collaborative style of working. Teams can communicate more efficiently using real time updates and data sharing applications, speeding up the development of products and services. As well as this, the cloud can provide data backup if power is shut off or equipment is damaged. So where does Artificial Intelligence come in? A surge in popularity has meant that the cloud needs to get smarter, and use more powerful methods. Providers have turned to deep learning, a subfield of machine learning. Machine learning itself is a statistical approach to creating AI. Google’s Cloud Vision API, for example, uses image recognition for visual content. Cloud storage company Box has partnered with Google to use Vision, highlighting the dominance of major technology corporations. Amazon has also introduced a new service called SageMaker, which provides tools for developers to write and deploy algorithms for tasks like speech recognition. In other words, it looks like the as a service model is knocking at AI’s door.
AI for the masses?
Cloud computing is evolving. As the cloud deals with more and more data, cloud computing providers have applied powerful machine learning techniques. The progressive introduction of these tools is making the cloud more efficient, encouraging businesses to transition from traditional IT. Gradually, the wider use of cloud computing will lead to the distribution of artificially intelligent technology. More and more people will be able to use AI related techniques, leading to the development of an as a service model in which corporations make certain tools available to customers. In some respects, this could be viewed as part of the democratisation of AI, reflecting the ongoing open source movement in which developers collaborate together to get better results. Opening up these techniques for public use would also allow major providers to see how developers choose to apply them. AI as a service could have wider implications for productivity simply by equipping businesses and developers with smarter tools. In turn, this could positively impact the economy. The AI as a service market, driven by cloud computing, presents a huge opportunity for big providers to entice more customers, but equally for these customers to experiment with machine learning tools they otherwise wouldn’t have had access to. The challenge now is to ensure that this precious data remains safe in the face of cybersecurity threats.
There can be no denying that Artificial Intelligence is still largely in the hands of big tech companies. Despite this, the expansion of cloud computing has necessitated the distribution of machine learning techniques to enable the cloud to process mass data. Developers who may not have the resources to experiment with the technology will now do so via the cloud, making AI more accessible than ever. Even so, by providing the cloud computing services, the likes of Amazon, Google, Microsoft and their challengers in the East will retain their power. However, AI is the hotly debated technology that some believe will take over the world. If it is to become a widely available service, it might be worth creating a code of practice for those who use it.
Should Artificial Intelligence be made widely available? What other disruption could be caused by AI as a service? Will Amazon be able to stand its ground in both the US and Chinese cloud markets? Share your thoughts and opinions.