ARTIFICIAL INTELLIGENCE
I've used computer analysis to solve complex water supply and distribution schemes. In my day, this was known as "Computer Modeling." In any case, a computer, whether artificially intelligent or not, has no intelligence. We, as living and thinking beings with consciousness, are the entities ultimately responsible for understanding and making decisions, not any form of computing.
Consequently, I'm delighted to say a Computer Engineer confirmed to me; Computer Modelling is a form of AI.
At this point, in describing this method, as AI may lead many to underestimate the limitations of computers, and that's my initial concern.
My use of Computer Modeling was significant because the computer model resolved problems in a way similar to how human brain cells would naturally address various analyses and calculations. The key difference is that we, as professionals in hydraulics, understood where the computer had its limitations. My concern, once again, is that AI may proliferate without many users possessing either of these essential understandings - a lack of scientific knowledge in the relevant field and the mistaken belief in the infallibility of computers.
In my case, Computer Modeling didn't replace employees; instead, it significantly expedited our work. Without this type of computer analysis, it would have taken considerably longer or even been impossible to perform this kind of analysis manually.
Furthermore, the economic aspect of AI introduces concerns. Companies may maximise profits by employing individuals who merely press computer buttons, as they cost significantly less than highly-qualified experts. Consequently, this could lead to a decrease in the number of professionally qualified employees. Another potential consequence is that AI could increase the proportion of the population ill-prepared to handle such responsibilities.
My concerns are reinforced by the privatisation of the water industry, which led to the outsourcing of work to private consultants. Some of these consultants lacked training in hydraulics and an understanding of computer limitations. An incident where a consultant relied heavily on computer-generated results without discernment highlighted these concerns. By way of an example: after the water industry was privatised, there was a directive for work to be awarded to outside private Consultants. This resulted in staff who were not trained in hydraulics or the limitations of a computer.
One such instance was when I had to assist one of their ‘foot soldiers’ - who was described as a Consultant.
The Consultant showed me a questionnaire, which requested the return to him for various measured water quantities, to be originally sent-out individually to our three area offices. As I and The Consultant’s team, operated from the same HQ offices I advised that how it was worded, led it open to different interpretations. The employee replied: "The computer will sort this out”. I replied: “A computer hasn’t the ability to discern and evaluate from the collective results of possible various individual subjectivity”! He didn’t take my advice, and similarly, later on when he presented me with other questionnaires, open to interpretations to be issued to our area offices again. Moreover, I don’t think he was even capable of understanding my advice to him!
Final Conclusions:
I still endorse my concerns as previously alluded to
2. While I support the use of AI in cases where it can expedite life-saving diagnoses, it's crucial that fully qualified medical consultants provide the initial information for analysis and re-evaluate the final computer-generated conclusions.
3. Even with the required expertise, computers and human beings are fallible. If a dedicated consultant using AI makes an incorrect diagnosis, compassion and understanding are necessary for both the patient and the consultant. AI could have the potential to reduce the risk of mis-diagnosis
M. E. Thorpe