We’ve been experimenting with AI on the Experts team for some time now. We’ve been trying to see what it does well and where it needs to improve.
The good news for our team of writers and contributors is that it isn’t going to replace their skills, yet. In a recent article Is ChatGPT Going To Perpetuate Audio Myths? Luke gave some examples of it answering questions posed on professional audio related subjects. The results were mixed.
One area where AI excels for creatives is ideas generation. Ask it to spew out a list of subjects about anything and suddenly you’ve gone from a blank page to a pile of ideas. You can go a step futher and ask it to create the outline to a 1000 word article, or if you are a songwriting, to sketch out the song.
You can ask AI to write the article, in a certain style, and again, to write a song in a certain style, complete with verses, choruses, bridge and middle eight. It will do that all day, every day. By the end of a week you could have a 1000 new song ideas.
Whilst the blank page is a reality for creatives, it’s not the only thing we have to fight against to go from an idea to something that’s original and something to be proud of.
AI is great for ideas, even good ones, but that’s only half the story.
If modern audio production proves anything, then it is that abundance isn’t any more likely to help than scarcity. We have an endless supply of gear, plugins, presets, sounds, samples, loops, tutorials, options, opinions… and for many, things are still no better.
In a recent podcast, Fix It In The Mix, contributor Gareth Nuttall spoke about receiving mixes with audio labelled as Take 1, Take 2, Take 3 etc. In other words, the client was expecting him to choose the best take for the vocal, or comp one at the mix stage, as the guy mixing the track. In a recent article, Dom Morley posed the question Who Should Comp the Lead Vocal? It’s well worth a read.
I recall speaking to Grammy-winning producer Vance Powell some years ago when he expressed exasperation at receiving songs to mix with tracks and tracks of guitars playing almost the same part. His approach was to mute half the tracks, saying, “Make some f*cking decisions.”
And this is the other half of the equation when it comes to making great art. Ideas are the first part, and when compared to the second, suddenly appear far more easy. The second part of the creative process is to make decisions, to commit, to remove options and to hone things down until what we are left with is something that’s hard to improve.
There is an often credited, if not unlikely, tale about the artist Michelangelo, not the turtle, but the Renaissance artist. When asked about the difficulties that he must have encountered in sculpting his masterpiece David he replied about his creative process, “It is easy. You just chip away the stone that doesn’t look like David.” True or not, that does describe the reality of the work of a sculpture, there is no plan B, there is only lots of tiny moments of commitment. The sculpture can’t go back and glue the stone back on, every decision is intentional and permanent. There is no undo button, or version autosave.
It takes a brave person to work in that way, but creativity is about being brave. It’s not about unlimited choices, or in our case; sounds, presets, or versions. Making great art requires us to “make some f*cking decisions” in the words of Vance Powell.
Ask any songwriter who has been doing it for some time and they will tell you they have a book, folder, and hard drive with lots of unfinished ideas. At some point, these ideas were the next best thing, then reality strikes and we find ourselves unable to decide which piece of stone to chip off next.
If you are working with the written word and have experimented with AI then you’ll know that one thing AI isn’t good at is commitment. Answers to questions are often vague and couched with lots of ‘get-out-of-jail’ phrases. In terms of editorial, the Experts are rarely short of ideas, but good ideas and ones that get finished are much harder. We often describe our job as relentless, as finished ideas have to be delivered every day, every week, every year. At some point we have to hit the chisel and remove some stone.
Great creativity is all about commitment, choosing the great line for the song. The right vocal take, drum sound, or where to put that chord. AI is not short of options, it’s short of decisions.
Rule one: Creativity needs ideas, but it’s what we do with those ideas that means they either fail or become something great.