Death Machine Learning As A Service And Distribution Of “vertical” Projects: 5 Trends In AI For 2017

Technology investor Bradford Cross has posted on his blog a note about the five major trends among the projects in the field of artificial intelligence in 2017. Edition vc.ru publishes an adapted translation of the material. “The craze for artificial intelligence technologies in 2017 will finally drown — many understand that most such algorithms are yet have nothing to offer.

Paradoxically, 2017 will also be a year of breakthroughs in this area,” writes the author of the note. He believes that in 2017 there will be many startups working in the field of artificial intelligence, which will develop “vertically” — that is, to solve problems in one area, covering it completely. The author has collected five trends in the industry that, in his opinion, revealed in the course of the year.

“Over the past year, representatives of the it industry are actively discussing bots. Generally under bot technicians understand the technology, which has four main properties. It responds to the environment, it is Autonomous, motivated and persistent,” writes the author. But in 2016 it company decided to usurp the term “bot” and began to call all automatiseret some part of their work.

“Automation of business processes, of course, is not going anywhere, but the hype around the bots will subside,” — says Cross. He sees several reasons for this phenomenon: However, Cross says that he does not believe artificial intelligence technology is not good enough to solve some tasks. According to him, the problem is that it is still not clear whether people want to use such systems in everyday life.

“Now everyone is obsessed with deep learning. Deep learning (for those who are not familiar with the terms) is an area of machine learning, and machine learning, in turn, is the area of artificial intelligence. And this it is now growing rapidly and is helping to solve global problems — thats why many companies are adopting the technology of deep learning into their work”.

Technologies for deep learning, for example, help to solve problems in computer vision. “Google, Facebook, Twitter, Uber, Microsoft, Salesforce — the algorithms now use all. Of universities available the mass of experts. 4 years ago the situation was completely different — so I think deep learning is one of the trends 2017″.

Cross says that under the distribution of deep learning implies growth of demand for specialists and developed technology, not the emergence of new companies that create similar algorithms and platforms. Technology Cleantech was always interested in investors due to the fact that they solve global problems and “it may seem that it is possible to earn good money,” writes Cross. “Its not a profitable business. The business can build on something for what it really is to find buyers”.

As a working examples of “cleantech” often lead Tesla and SolarCity, but we must not forget that in these companies “net” technology is only part of the product. You cant build a product completely “clean” technology — this is unlikely to work. Big business puts in first place the customers needs, not technological trend. “Solar energy is not a market and technology.

So artificial intelligence is not market and technology. But venture funds defy the trend and become harder to appreciate the company,” — explains the author of the material. According to Cross, this situation provokes the appearance of immature companies without a product and gexperts on the market. In addition, most of these projects will soon be closed and the investors will lose their money.

Model MLaaS (Machine Learning as a Service), says Cross, has been at least 10 years, but the projects working on it, often fail. “It doesnt work simply because the people who understand why they machine learning and how is it organized, use technology, open source, and those who do not know, will not be able to work with the API,” he explains. At the same time sell their platforms machine learning continue both small and large companies — like Microsoft, Amazon and Google in their cloud-based toolboxes.

But I have not seen companies that actually use these APIs. MLaaS have projects not just the audience. According to Cross, one of the main trends 2017 — the emergence and development of companies that are using artificial intelligence to solve problems from one chosen field, and create the finished product.

“AI will change the whole industry. Such vertical startups will be able to absorb and expertise, and data and production of the product and to give the final result. Which will be in demand among clients. These projects will even be able to find problems that are impossible to solve without the technology of artificial intelligence,” explains Cross. Send your column about how our world will change, [email protected]

Leave a Reply