Chuck-In Soup
Image shows someone cooking a large pot of steaming soup
Let me take you on a little journey to explain one of my main problems with AI in music and the arts, and it starts in my childhood.
When I was young, I remember chatting with someone at school. As happens to us all when we are little, you become aware that your family traditions and habits are different from other families. Slowly a whole new world opens up and you learn of all the exciting and totally different rituals and routines other families call normal life.
Another pupil explained that their mum made what they call ‘Chuck-in soup’. Not only did this seem exotic, but I also realised what a great idea! it is the amalgamation of what is left over, what needs using up. Every batch is subtly different, but helps to clean the fridge out and provides a meal for the family. Perfect.
In modern times, this soup is the perfect thing to ensure no food goes to waste and we can all save a few pennies, but back in the 90s it was just what a family did every week. I’ve regularly made my own versions, and enjoyed the satisfaction of feeding our family while ensuring we don’t waste food or money.
But. Despite sounding free and without rules, it is in fact a meal that follows a structure and learned pattern. A professional chef can make a Michellin starred soup of leftovers taste like heaven. As a home cook who loves food, I know if I add too much of any ingredient the soup is ruined. Too much seasoning/not enough will ruin it. If I add a certain herb I can recreate a favourite taste, add another and my household will act like I’ve served poison. So I use my experience, learning, knowledge of foreign food and tv cookery shows, recipes from other soups and meals to guide me, along with experience of previous versions.
This is the problem that we have with AI. AI works as a big pot of ‘Chuck-in soup’. The ingredients may be top-notch quality - the best there is - or they could be of very poor quality. The quantities may be the perfect balance, or they could be overly influenced by one source. Who is to judge or correct?
If used and run without learning, experience, quality control, understanding of the subject, knowledge of historic sources, it is not culturally relevant or reflective of society. It will not aid true creativity when it is driven and presented by those who do not know and understand the subject outwith their AI job.
If you use AI for your work, you are no doubt aware of these things, and have found times where you have chosen how/why to make adjustments to what you have made. The problem will arise when it is used without knowledge outwith the system of AI.
As the arts are being eroded in education and we are being led to believe that lower quality is desirable, who will be able to judge? Why pay for human musicality and live performances if an untrained non-musician can make a cheaper ‘performance experience’ that can be repeated forever from a recording?
