AI Strategy for Successful Business

Original article was published on Artificial Intelligence on Medium

AI Strategy for Successful Business

AI technology is no longer science fiction, but a reality facing businesses. Today, companies are applying a variety of AI technologies to their businesses in new and interesting ways, including machine learning algorithms and smart connected products. According to a recent survey conducted by Deloitte on companies that actively adopt these cognitive technologies, 76 percent of respondents said cognitive technologies, including AI, would substantially change their businesses within three years. However, there is still a big gap between this aspiration of the company and reality. Gartner analyst Nick Heudecker said that in reality, about 85 percent of big data projects fail. So how can a company achieve successful business results through an artificial intelligence strategy?

Establish clear goals and a shared vision

First, the purpose of introducing artificial intelligence should be clear. Just as it is necessary to determine whether a person is suitable for the company when hiring new employees, it is necessary to carefully judge how the application of AI technology will affect actual business results.

According to Roman Stanek, CEO of big data analytics firm GoodData, the first question to think about is “what business performance are companies pursuing?” Because AI projects generally begin with implementing specific technologies, front-line managers and staff often do not understand its usefulness unless the vision for the business goals they want to achieve is shared. Therefore, it is important to clearly state the goals within the organization.

What will you automate?

Second, it is necessary to select carefully what to automate. Many people are worried that cognitive systems such as artificial intelligence will cause people to lose their jobs. But MIT economist David Autor sees it differently. It doesn’t mean jobs are lost, it just means that there is a distinction between ‘routine work’ and ‘non-routine work’. Just as machines automated human physical labor during the Industrial Revolution, artificial intelligence only quickly automates the cognitive processes we identify as “busy work.” If you go to an Apple store, you can easily understand these priorities. It is not artificial intelligence that greets us when we visit Apple stores, but the warm welcome and reception of employees. This shows that automation within Apple is in the backend, while interaction with customers is still human territory.

Understanding Data and Acting Responsibly

Third, data must be fully understood and selected. For quite a long time, there has been a perception that more data is better. That’s why companies have been trying to collect as much data as possible to inject into sophisticated algorithms to create a more accurate predictive model. But it is now clear that indiscriminate data collection is by no means the best practice. If companies enter systems without fully understanding the data, the data bias problem can be serious. Consequently, the entity will need to create a model that will be fully aware of the data sources it intends to use, one that people can understand and explain.

The value of labor

Finally, we must recognize and transform the view of a person not as a unit of labor, but as a more valuable asset to society. Automation of some work means that the labor force used for work in one area can then be utilized in other areas. Using artificial intelligence simply to replace human labor and reduce costs can not lead to the best results.

Tesla CEO Elon Musk, who seeked to implement a futuristic factory with a fully automated system using AI-based robots, ended up failing to meet orders and forced production to stop. Musk tweeted: “That’s right. Tesla’s over-automation was a mistake. To be exact, it was my mistake to underestimate humans.

The solution to this was the expansion of jobs. Tesla hired hundreds of workers to handle production process modifications, robot training and replacement work.

A global survey of more than 1,000 companies at the forefront of implementing AI systems found that the performance when humans and machines worked together was far higher than when machines replaced humans. Through this collaborative relationship, one can further develop the performance of the machine, and the machine can help the person achieve a higher level of performance.

Amid this trend in the artificial intelligence market, Mind AI is also developing technologies that can contribute to industry and business, and plans to release the best artificial intelligence in the second half of this year that complements job creation and ethical issues that humans are worried about.