The Commoditization of AI
It first started with the cloud: cloud computing became standardized and cheaper – a common technology, offered by increasingly more suppliers. As a consequence, consumers benefit from lower prices. Today, there is also a trend towards the commoditization of Artificial Intelligence, better known as AI.
The Commoditization of AI
AI is already touching our lives and it will continue to become more and more relevant in our day-to-day lives. The commoditization of AI is caused by major breakthroughs in machine learning. The adoption of cloud computing and a great improvement in algorithms have made access to massive computing power and data stacks a lot easier.
Just consider how the average consumer is already benefitting from AI: one of the two ways that are probably pre-eminent is natural language. Natural language understanding means that one can talk in complete standard sentences and have a machine make sense of them. That would not have been possible seven years ago, except in the very highest echelon or maybe in educational institutions, like universities, where they have access to the necessary data. Now, consumers can also enjoy the benefits of this: talk to your bank account to find out what your bank balance is. Or, ask Apple’s Siri to assist you with a reminder. Microsoft’s Cortana may help you with advice. Or, ask for information from Google Voice Search, or Amazon’s Alexa. There’s no coincidence these things have all appeared together – it’s because of the commoditization of AI.
The second most obvious way AI has been commoditized is through computer vision. AI enables the understanding of what is in a picture or photograph, and the commoditization of AI allowed for Deep Learning, which is used on a daily basis. It may be an algorithm that can look at paintings, interpret their style, and then redraw it in the style of the artist (so, you could have a Van Gogh that was never painted by Van Gogh!) Or, it may be technology that identifies specific people or objects in a photograph to enable the consumer to extract a subset of pictures (by means of machine-learning) to suit the consumer’s requirements. This kind of recognition of people in pictures was once a novelty, but it is almost trivial to do now – Facebook has been doing it for a few years and so does Google Photos. Upload your photos to Google Photos, search, and it will correctly find what you are looking for!
Cloud Computing: Algorithms & Data
As AI algorithms improve and data expands, access to AI will become accessible everywhere. Big companies who make use of AI (because they have the means to develop and deploy AI) will not, for much longer, enjoy an advantage over those companies who do not have the means to develop AI. That is because big players will give competitors the same capabilities!
The rise of cloud computing plays a significant role in the commoditization of AI. It is distributing the power required to make AI accessible to all. Ordinary companies no longer need to have a huge stack or massive processing power, because they can just pay the likes of Amazon, Google or Microsoft to do the processing work for them. And, thanks to cloud computing, the data is readily available. Moreover, improvements in algorithms also help to make consumer-facing apps smaller.
In order for AI to be an effective and useful tool, it needs data. The more data available, the more commoditized AI will be. Big players like Facebook, Google, Microsoft, IBM, etc., can deliver really complicated solutions in easy-to-use consumer-facing APIs. Smaller companies can also leverage this infrastructure, as a service, in order to run their own algorithms and services. These developments all make it possible for start-ups to leverage the same power and then sell them out to big players that take enterprises further. And so, we are seeing a big wave of independent small businesses delivering what was, until a few years ago, totally out of their reach.
Seven years ago, if a small company was looking into the development of a smart chat bot system, an entire custom stack would have to be built. It was possible, but complicated and expensive. It just was not easy for small companies, let alone individuals, to enjoy what AI had to offer. But then, major players (like Facebook, IBM, Amazon, Microsoft and Google) having the means to develop AI, made it more accessible. In addition to big improvements in algorithms, the AI systems of big players have the massive amounts of data that is required to make powerful data processing possible.
AI is incredibly smart and systems are just getting smarter as they are fed more data! Up until recently, to ordinary consumers “Artificial Intelligence” was just a mere academic term. However, the commoditization of AI means that easy access to consumer-facing AI services is already in our homes! There are a variety of different AI services that we are beginning to take for granted…AI is becoming commoditized and consumers are already getting very familiar with AI features!