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 better understand AI and solve your business issues.
In the last blog, we discussed how important accountability factor is in AI systems. Today, we will discuss the current progress and future responses regarding AI ethics.
In Japan, the Cabinet Office has recently established the Council on Social Principles of Human-Centric AI, which has proposed seven principles. These principles include “human-centered principles”, “principles for ensuring privacy”, “principles for fairness, accountability, and transparency (FAT)”, and others that society should pay attention to accept and appropriately use AI.
The Differences between the Current State and Future Responses for AI
Looking at one example, within the principle document, there is a section titled “The Relationship Between the Use of AI-Based Programs to Support Diagnosis and Treatment vs. The Provisions of the Medical Practitioners Act”. Next, I want to provide a summary of the current issues and future responses regarding Article 3 of this act.
- The Current State of AI
AI is merely a support tool used to improve the efficiency and presentation of information to medical professionals to aid in making decisions in the medical treatment process. The physician still has the final say in medical diagnosis for the time being.
- Future Responses for AI
The physician is responsible for both the diagnosis and treatment. The physician is responsible for the final diagnosis. The medical treatment in question is performed as a medical practice under Article 17 of the Medical Practitioners Act.
Looking at these two sections, we can see that the phrase “support tool” has been removed from the future response section. While adhering to the main principle that medical professionals should control the final decision in medical matters, this also indicates the direction that AI will take in the future.
I think that AI will be actively utilized in the areas of diagnosis and treatment in the future. This discussion of AI ethics is being discussed not only in Japan but also in the U.S. and other European governments and companies.
Collaboration between Humans and AI
The concept of “Human-in-the-Loop” is considered a competitive advantage for Japanese companies As the debate on AI ethics grows, there is a growing movement known as “human-in-the-loop”. This movement aims to promote collaboration between humans and AI.
In the world of visual inspection, the current mainstream AI based on deep learning autonomously learns the characteristics of an object and creates categorization rules using large amounts of data. In other words, this AI operates without human intervention.
This is partly due to the fact that human understanding cannot keep up with AI or understand how it makes decisions. However, the absence of human input can result in a variety of problems, such as the ones I have previously mentioned.
Sparse Modeling as a Optimal Solution for Human-in-the-loop Methodology
Human-in-the-loop, on the other hand, is a way of thinking to solve problems that arise in the operation of AI by daring to involve humans in the decision-making process. Naturally, those issues include areas such as social correctness and ethics.
Personally, I believe that this human-in-loop methodology is a space where Japanese companies can compete when it comes to the AI race. By incorporating human knowledge in areas where the quantity and quality of data are lacking, we dare to establish areas of human involvement in AI decisions. By doing so, we believe we can cover the aforementioned shortcomings of AI.
In the U.S., many have adopted a way of thinking that pushes humans to allow AI to operate on its own. However, I think that Japan should move in the direction of building systems that are easy for humans to use in order to promote active human involvement alongside AI systems.
From Japan’s position, it is impossible to compete in terms of technological innovation, data availability, and financial resources. Therefore, it is necessary for us to bring our strategies into the battle of the analog domain.
This is also where HACARUS shines since sparse modeling is one of the optimal solutions for the human-in-loop methodology. As mentioned above, this is because it can be built and operated using small amounts of data and it can be viewed from a different perspective as a mechanism that facilitates human involvement.
In the next blog, I will continue to discuss the advantages of sparse modeling in the context of Human-in-the-loop methodology. For your updated knowledge and insight about AI technology, subscribe to our newsletter or visit HACARUS website https://hacarus.com.