I love music and I feel lucky that I grew up through the Punk/New Wave/New Romantic. It formed the basis for my love of a certain type of music. For me that meant an enduring love of Japan (the group) and subsequently David Sylvian as well as many other bands of the era, such as Duran Duran, Spandau Ballet, Marillion, etc.

So recently I thought it might be an interesting exercise to compile a list of the major albums that were released in each year of my birth against my age. I could see what albums I was into when I was 16 or 17 for example.

This, I thought, would be a perfect opportunity to use AI to compile this information. As I am most familiar with ChatGPT I asked it to compile a list. The top ten albums that year for a specified range of years.

It couldn’t do it. It said that the information wasn’t publicly available and therefore it couldn’t compile it.

So I turned to Claude by Anthropic and I asked it for the same thing. Not only did it compile it after asking me pertinent questions about the content, it then produced a fully formatted Word document.

I was impressed.

It was an object lesson in choosing the right tool because a hammer is not a screwdriver and not all AI systems are the same. Now this may well be obvious but it only became obvious through usage.