ChatGPT & Live Music: Useful or not?

We were impressed with the functionality of ChatGPT and asked ourselves how relevant it could be for the live music industry. This is part 1 of 2 of our answer to this question.

ChatGPT & Live Music: Useful or not?
A screenshot of one of our chats with ChatGPT.


Like everyone in and around the tech world these days, we have been playing around with a new innovation called ChatGPT.

In case you’ve missed what all the fuss is about, here’s what the tool says when you ask it “what is ChatGPT”:

ChatGPT is a variant of the GPT-3 language model developed by OpenAI. It is designed for use in natural language processing tasks such as language translation, question answering, and text generation. ChatGPT is particularly well-suited for chatbot applications, as it is able to generate responses to user prompts in a conversational style.

Like other language models, ChatGPT is trained on a large dataset of text and uses machine learning algorithms to predict the most likely next word or sequence of words given a certain context. This allows it to generate responses that are natural and coherent, even when given a prompt that it has not seen before.

It’s a very impressive piece of technology, accessible through a simple text field, and it’s also very fun to play around with...
And it got us thinking: could ChatGPT be relevant for the live music industry? Could it help tour managers and touring artists be more successful?

Will ChatGPT change the live music industry?

Based on our personal experience and all the conversations we have had with tour managers, artists, tour accountants and managers, our hypothesis is that ChatGPT will not be very helpful for the live music industry.
And there are 3 reasons why:

1. This industry requires highly specific knowledge and experience.

ChatGPT, just like a good old Googling session, will be able to provide superficial information about for example the different players in the industry, and what their role is. If you’ve never heard of the term “promoter” or “venue”, this will help you understand their role and why these players exist. However, it won’t provide a shortcut to understanding what motivates these people, how a tour manager ideally works together with a promoter, and how to build mutual trust. And it definitely won’t help you figure all of that out for the specific artist team you work with. The level of specificity (and the human connection and experience required) is too high for ChatGPT.

2. There is a lack of written information.

ChatGPT told us itself: it works based on a large scale analysis of text data. In other words, in order to work well, it relies on a written account of that specific knowledge and experience that we referenced in point 1. And as far as we are aware, this simply does not exist in written form. Artists and their teams require privacy about what happens on the road and how the financial side of their business works. Also, the tour manager life is a busy life. We haven’t encountered too many of them who have time to sit down and write stories of how they settled shows or how they created trust with the management team they worked with. So the information ChatGPT needs to function well is simply not available outside of the minds of the talent in this industry.

3. The conditions are ever changing.

You don’t have to work in the music industry to appreciate that it’s a very dynamic environment. Things like the current energy crisis and inflation have a tremendous impact on the viability of tours, and this in a world where many artists are still hungry to make good on their pre-covid touring promises (and many players are still trying to make up for financial losses from that period). Borders may close, visas may get denied, shows canceled, fuel prices go up, ticket sales come in differently than expected,... Never a boring moment!
But this also once again means that the expertise that artist teams hold is ever changing and improving, and hence very hard to “codify” for a tool like ChatGPT.

So we don’t see a direct impact of this impressive technology on our industry.
That’s not to say that it isn’t a mindblowing innovation, and one that we’re very impressed with...

ChatGPT, do you agree with us?

So, after coming up with the 3 reasons described above, there was only one thing left to do… See if ChatGPT itself agrees with our assessment.
So we asked it:

“Write a blogpost about why ChatGPT is not useful for tour managers in the live music industry”

And this is the answer it came up with:

While ChatGPT is a powerful tool for many applications, it is not particularly useful for tour managers in the live music industry. Here are a few reasons why:
Lack of domain-specific knowledge: Tour managers in the live music industry have to deal with a wide range of tasks, including booking shows, coordinating logistics, managing budgets, and dealing with artist contracts. ChatGPT, like most language models, is not tailored to any specific domain and lacks the specialized knowledge that a tour manager would need to perform these tasks effectively.
Lack of real-world context: Tour managers often have to make decisions based on real-world factors such as venue capacities, local regulations, and artist availability. ChatGPT does not have access to this kind of information and would not be able to provide accurate or relevant recommendations in these situations.
Limited ability to interact: While ChatGPT is good at generating responses to specific prompts, it is not designed to carry on a sustained conversation or to interact with users in a flexible way. This means that it would not be able to handle the complex and dynamic interactions that are common in the live music industry.
In summary, while ChatGPT may be a useful tool in some contexts, it is not well-suited to the specific needs of tour managers in the live music industry. They would be better served by tools and resources that are tailored to their specific needs and that have access to the kind of real-world knowledge and context that is essential to their work.

We’ll leave it to you to decide how spot-on the answer is…
… and we hope to see you again in part II of the ChatGPT series next week!