Hello everyone, My name is Kenshin Fujiwara and I am the CEO and founder of HACARUS Inc.
Through this series of blogs, I will discuss a wide range of topics of AI, from the history of AI to practical tips for the successful application of AI projects. I hope my blog posts will help you gain a better understanding of AI and solve your business issues.
In the last blog, we studied key points for successful AI implementation in the early stages in particular, regarding the organization and human resources. In today’s blog, we will continue to discuss key points for successful AI implementation and potential points of failure for AI introduction.
Knowledge of AI and Business Operations
The key to identifying problems for a company is rooted in knowledge of the company’s operations, not so much the digital expertise of its members. For example, for a company that manufactures automobile parts, it is important to have a deep understanding of the manufacturing process from beginning to end. For an organization in the medical field, it is essential to work with a physician with extensive knowledge and experience in the field of diseases where AI will be introduced.
As a team member, the role of the trained personnel is to draw out their industry knowledge and experience and use it to help identify business issues. Then, they will add feasibility based on the data and recent AI case studies. Finally, they will set the agenda for the AI implementation project.
At this point, there is no need to begin the AI development process. When moving on to the development phase, the AI creation can be delegated to an outside vendor with advanced expertise. This way, the project can proceed quickly.
However, no matter how much knowledge someone has about digital technology, good AI cannot be developed without sufficient knowledge about real operations in the field. As the AI personnel, they are required to have knowledge of the business and effectively communicate with the specialized AI vendors.
Potential Points of Failure for AI Introduction
One pitfall I often see with AI implementation is that the organization tends to throw the entire AI project to an outside vendor that specializes in AI. This is an issue because while the AI vendors may have knowledge of AI and digital technologies, they do not know how the business operates with its customers or about the work environment. Therefore, if the tasks are left entirely to the vendor, there is a risk that the vendor will go in the wrong direction from the beginning.
To share a funny story from my experience, there was a case in which a vendor specializing in AI was hired to build the system, but from the beginning, the vendor didn’t have any of the necessary data. If there were trained AI personnel in-house, the company would have at least been aware of the “data barrier” prior to hiring a vendor.
It is a common idea, especially in Japan, that IT systems are prone to be thrown into the digitalization process. This is also supported by the past. When trying to implement mission-critical systems, especially in large companies, there was a reluctance to have in-house engineers. The main reason for this is that their core systems are only renewed once a year, and an engineer is only needed at this time. This makes the hiring of engineers an inefficient process.
Under Japanese labor laws, it is not an easy task to fire employees just because they no longer have a functional role in the company. As a result, many companies in Japan have minimized the hiring of IT personnel and outsourced this work on a case-by-case basis. As an extension, companies followed the same practices when AI began to become more mainstream.
As I mentioned earlier, the essence of DX is not about digitization but rather about transforming the enterprise. Therefore, if AI is going to be introduced as a way of promoting DX, the organization needs to aim for an aggressive AI construction plan. This means training in-house personnel for the long term.
Beware of the Pitfalls of “Agenda Setting”
For companies introducing AI, sometimes they establish a DX Promotion department. One of the major pitfalls is that due to management, the department tries to conquer a monumental task, and gets lost along the way. In this way, the company fails to produce the desired results.
One example in the manufacturing industry is when a company wants to suddenly become a smart factory. This means modifying the factory so that it is completely IT-enabled and in some cases, operated unmanned.
While it is perfectly acceptable to set this type of goal for the future, if this becomes the goal from the beginning, the project is setting itself up for failure. This type of thinking creates a trap of trial and failure.
In the next blog, I will continue to discuss other key points for successful AI implementation. For your updated knowledge and insight about AI technology, subscribe to our newsletter or visit HACARUS website https://hacarus.com.