AI Adoption in 2017

When I was asked to prepare a summary of what happened in the artificial intelligence industry throughout 2016 and what might happen in 2017, the only thing that came to my mind was a quote from Dickens’ “A Tale of Two Cities.”

It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness, it was the epoch of belief, it was the epoch of incredulity, it was the season of light, it was the season of darkness, it was the spring of hope, it was the winter of despair.

We have reached a point where AI is experiencing a period of vibrant flourishing—its age of reason, a time of faith—and at the same time, it is becoming increasingly difficult to distinguish those who are actually engaged in artificial intelligence from those who say they are, treating it as a mere buzzword. I tried to create a summary that would look at the industry from a certain distance, selecting the most interesting facts from the perspective of an airplane pilot—as much as I myself am the CEO of a startup involved in AI.

Crunchbase, the world’s largest list of startups that have received VC funding, currently has as many as 1196 startups that are engaged in artificial intelligence in its ranks. Each one is involved in different areas of the broad field that is AI. Some are creating tools for automatic text analysis; others are teaching neural networks to watch cat images for us; still others are listening to what we have to say. The market giants appreciate the progress. In 2016 alone, players like Google, IBM, Yahoo, Intel, Apple, and Salesforce acquired over 40 AI startups, increasing the number of acquisitions in the industry to over 140 companies sold since 2011. We can predict that the trend will only be upward in 2017.

In Poland too, we can boast several startups that apply machine learning to other industries. Companies such as Ada, Wandlee, and Infermedica can boast interesting products and implementations in the past year—in industries such as real estate rental, sales, and medicine, respectively.

The number of projects is positively influenced by the fact that 2016 was a year of development for platforms related to artificial intelligence. Both Facebook and Microsoft opened their platforms related to chatbots and AI tools. The technical infrastructure, developed by players like Amazon, Microsoft, or NVIDIA, is becoming cheaper and more widely available. Amazon has revived the IoT market, promoting Alexa as an operating system for smart homes—as proven by CES 2016, where perhaps only light bulbs were not voice-controlled.

And speaking of market giants…

Google is certainly on the rise. March was the month when AlphaGo, an artificial intelligence created as part of the Google DeepMind project, defeated the world champion in Go, winning four out of five matches. Google researchers also began to better understand the subtle ways in which data sets used for machine learning can influence the transfer of human biases, such as sexism or racism, to decisions made by theoretically objective algorithms. If we add to this all the results of work related to Google Translate, which showed that their system can learn new languages directly from other languages it already knows, it can indeed be stated that Google had a spectacular year in many respects.

It’s also worth remembering how Amazon is developing its ecosystem of tools supporting developers wanting to operate in the Alexa ecosystem—systems like Amazon Polly, which generates speech from text on demand, come to mind. It’s also worth mentioning projects like OpenAI Universe, an open set of training environments that can serve creators of new AI to test their effectiveness in general-purpose tasks.

What awaits us in 2017?

This question seems the most intriguing. The year 2016 can probably be summarized as the year in which AI invaded the consumer space as a tool that brings new quality to key aspects of products that we know and love. Machine learning gives us more and better “skills”—increasingly accurate dictation, increasingly precise search results, increasingly clever hints, better recommendations, and so on. It’s a value that can’t be denied.

However, the question is whether this is actually where the technology industry should be heading in the future. It seems to me that we are still waiting for the first real startup that will not so much be involved in developing AI skills, but rather decide to base the foundations of its business on artificial intelligence. I’m talking about a company that will be able to base key aspects of its business model in such a way as to continuously optimize profits, minimize losses, and maximize the efficiency of its own products.

Whoever manages to build such a company first will be able to react much faster than the competition to changing market conditions on both macro and micro scales—and only such a company will represent a new quality in the market. Perhaps 2017 will show which startup is on the right track to becoming such a company.