Artificial intelligence technologies are gradually reaching a productivity plateau. Among the first swallows are speech recognition services, on the basis of which chat bots work. Users are also interested in automated machine learning technologies and business applications with built-in AI mechanisms. There is a growing demand for artificial intelligence platforms provided as a service and related cloud services. But some applications, such as in autonomous vehicles, will not be realized until 10 years from now.
Artificial intelligence goes to the masses
Gartner’s latest research on the development of artificial intelligence reveals the wide variety of applications of AI in enterprises. And this is logical, given that, according to surveys conducted by this company, in 2019 the share of organizations that implemented AI increased from 4% to 14% compared to last year. And, despite the relative youth of the AI market as a whole, Gartner analysts placed two technologies at once in the “productivity plateau” section – “ speech recognition ” and “GPU-based AI accelerators” (the latter are much better suited for creating artificial intelligence systems than processors “general use”).
Among other applications of AI that are predicted to succeed soon are conversational AI tools, fueled by the success of virtual assistants like Amazon Alexa, Google Assistant , etc. There is interest in new technologies such as augmented intelligence, edge AI , whose popularity is growing along with the popularity of peripheral computing itself, automated data labeling and “explainable” AI ( artificial intelligence system , the solutions of which people can explain). But autonomous vehicles , which, as many believe, are about to appear on the roads, according to Gartner, will “leave” the productivity plateaumore than 10 years later.
I – AI blue chips
In general, many new technologies have appeared on the artificial intelligence hype curve, and a significant proportion of them are marked with blue circles, indicating that Gartner hopes they will soon reach a productivity plateau. Moreover, many of them received a forecast of “two to five years before implementation”, while still climbing the peak of hopes.
However, as analysts note at the same time, not all of the new technologies have a clear application and are able to benefit the business. And we must try to realistically approach the forecasts and analysis of the prospects for implementation.
One way or another, analysts advise companies that strive to keep up with the times to at least prepare a financial and economic justification for the introduction of AI. And those who have already carried out initial implementations should think about scaling projects.
The Gartner Curve for Artificial Intelligence
Artificial intelligence technologies worth paying special attention to
Among all the AI technologies , Gartner analysts highlighted five that could most seriously change business processes in the foreseeable future, and advise CIOs to closely monitor their development.
II – Augmented Intelligence
Analysts refer to augmented intelligence systems as automation tools that help increase the productivity of human mental work. They help organize a “partnership” between humans and AI , in which the former play a dominant role.
The use of artificial intelligence in this capacity helps to reduce the amount of routine work and, accordingly, the number of errors in the course of its implementation. And human participation, in turn, will reduce the risk associated with automated decision making – due to the fact that a person will be able to solve questions that AI has not yet been trained to answer.
IT personnel for the industry will be trained in a special educational centerIT INDUSTRY SUPPORT
Chatbots, the face of artificial intelligence that we encounter almost daily, are also changing the way we interact with customers. For example, at Kia , they help 115,000 car owners solve problems every week, and at the German discount chain Lidl , a bot named Margot gives shoppers tips on choosing wines and snacks.
Chatbots can be text and voice, they answer standard questions according to a script compiled in advance, taking into account the experience gained by live operators. They can be used to solve the problems of the human resources department or help desk , help employees adapt to a new place, etc. But to the greatest extent, these AI solutions have changed the customer service process. If earlier the user usually had to study the interface of interaction with the system, now the chatbot “ studies” the user, “guessing” his intentions and suggesting further actions.
III – Machine learning
Among the tasks that machine learning can solve are customer service personalization, dynamic pricing, disease diagnosis, anti-money laundering, and much more. The principle of operation of machine learning tools is the detection of patterns present in the data using mathematical models. Machine learning is being used more and more, fueled by the rapid growth of data in organizations and the rapid development of computing infrastructures.Sergey Pimkov, Selectel: What IaaS solutions do businesses need in 2022CLOUD SERVICES
Machine learning helps to optimize processes and find new solutions to business problems in a wide variety of industries. At American Express , for example, machine learning algorithms and analytics detect fraud attempts in near real time, saving the company millions by preventing losses. And at Volvo , analytical systems predict probable failures and the need for repair and maintenance of various vehicle components, helping to improve their safety.
IV – AI control system
According to experts, it is impossible to neglect the creation of an AI governance system in enterprises. This is necessary, among other things, to understand and control the potential risks associated with regulation and the possibility of damage to reputation. As explained in Gartner , the AI management system is based on specially developed policies to prevent systemic errors (“bias”) of AI, discrimination of users or groups of users on certain grounds, and other possible negative consequences of using artificial intelligence.
When developing an AI governance system, experts recommend that analytics leaders and CIOs pay attention to three areas: trust, transparency, and the principles of ethno-cultural diversity (diversity). The need to ensure the ability to trust data sources and the results of AI systems is one of the cornerstones of their successful implementation, and the development of transparency requirements for data sources and algorithms will reduce risks. Concern for diversity in data and algorithms contributes to the ethics and accuracy of the results of AI solutions.
V – Intelligent Applications
For a few more years, the only way to introduce artificial intelligence tools was to independently develop AI systems. Today, however, most organizations prefer not to develop such solutions or even purchase “stand-alone” AI systems, but to receive AI tools as part of enterprise applications.
Initially, the most “intelligent” were analytics tools with built-in AI technologies. Recently, however, vendors of a wide variety of enterprise applications – ERP systems , CRM systems , human resources management and office suites – are embedding AI tools in them and are starting to create AI platforms. So Gartner analysts advise CIOs to require software vendors to include AI tools in their product development plans, including advanced analytics tools and tools that optimize user experiences.