In the interview with Ben Shneiderman from 2018, he had several prognostications about AI. However, one that stuck with me was his worry about AI's integration into manufacturing systems, financial systems, loans, mortgages, and medicine. He believed that AI should be studied and improved, but these sectors should have levels of protection against AI. This was an intriguing topic to me because I believe he is right in this statement. AI has drastically increased productivity recently by becoming a mainstream tool and integrating into a person’s day-to-day life. This has led to extreme success in many fields, like manufacturing.
I do agree with his statement that we should recognize the risks of unchecked AI in sensitive areas like medicine or lending. This is especially true when these areas involve significant decisions in people's lives. I can understand the risk where people are not comfortable with its help in making those decisions for them. AI still makes mistakes at its current stage, and in these fields it is critical to be cautious and slow in its integration even with its improvements every year.
I also agree with his views of safe integration. He suggested that insurance should vet the implementation of AI in these sensitive industries to prevent reckless integration. With safe integration, AI has significantly helped the sensitive industries he mentioned. AI in manufacturing has been shown to significantly increase productivity and optimize profits and sustainability. General Electric has helped prevent injury or mistakes with its AI that monitors the health of the equipment. Such technology can actually help save lives in the dangerous business of manufacturing. Finance has successfully integrated AI. An example of this is Mastercard using AI to help detect fraud, which is a significant benefit to its customers.
I understand Ben Shneiderman’s worry, but well-measured and regulated AI integration can unlock a future with extreme benefits in our society. We hope it can be safely integrated into these critical systems to help more people than we can now.
Link to the video of the interview: