Understanding AI—Really.
The main opposition to AI in general is the fact that generative AI takes without permission.
In experimenting with our new artificial intelligence technology, a lot of people have leaned towards its darker uses. Of course our curiosity and human nature is always going to explore both sides of a new entity. We’ll never limit ourselves to the “good side” of a new experience. Well, most won’t, anyway. Just think about that, however, in a positive light; without the knowledge of the negative side to something, how do we learn to limit its consequences? You have to understand all sides of a thing in order to protect yourself. And in the beginning of any new and massive technological change, there will be a period of chaos. The trick is to learn every angle and become pragmatic about it instead of fearing it.
Let’s get ahead of this fear by learning the basics of how AI works. This is going to keep you competitive, aware, and it will help you leverage AI’s power to help with more positive things pertaining to your artistic career—like marketing. If you know the details about how algorithms work, you can figure out how to use them your way. And algorithms are AI.
There are different types of algorithms. Some basic ones are decision trees, random forests, neural networks, and genetic algorithms. Instead of turning this article into a book with all the information basics, we’ll just focus on one: Decision Trees.
Decision trees are used in machine learning; they’re a graphical representation of a sequence of decisions that help to make predictions or classify information. To build this algorithm, you choose a root node. This is a starting point, and it contains the problem that needs to be solved. The root node is then split into two or more child nodes, which hold decision criteria. This starts the process of differentiating between best possible outcomes to solve the problem. These child nodes are then studied by the machine until possibilities are exhausted, eventually reaching a “terminal node”. (Remember: the decisions are all based on that first root and all its extensions.)
Pruning the tree: “Pruning” occurs once the tree is built and all the nodes are split and a terminal point is reached. Pruning removes unnecessary nodes and branches to improve accuracy and efficiency.
Predictions: This is the end result of the original problem or question. The tree is used to make predictions or classifications based on the data provided.
That’s a very simplistic way of explaining an algorithm, but it’s basically how it works. Knowing this, and knowing that there are hundreds of types of algorithms, you can get an idea of how much knowledge a machine can provide in any area it’s been implemented. We’ve already seen how much of a dent AI based algorithms have made in our daily lives with social media, Google search, face ID on our phones, even simple emails. We’ve been using it in GPS apps, banking fraud prevention, and recommendations on Netflix. Most of us just haven’t realized what a big part of our lives it’s become—until we started seeing the actual words, artificial intelligence, more commonly used in lieu of new artistic generative apps. Now we’re finally paying attention.
AI is transforming the marketing landscape.
It’s actually offering valuable insights into what we can use it for. As we learn more about it, it will become easier to gain a competitive edge in self promotion for independent artists worldwide. But in order to use it (past our experience with gps), we have to get over the fear of it. So what are those fears, exactly?
The Pew Research Center, a nonpartisan think tank, has shared data that shows there’s an increasing fear and distrust of AI being a part of our lives. The first problem is that there isn’t enough familiarity with it. As of March 2023, 42% of Americans alone hadn’t even heard of ChatGPT, and only 18% of those who knew what it was had actually used it.
Secondly, most of us don’t like the idea of being stalked by authority; especially our employers on a daily basis. And AI is being used to make final hiring decisions, track employees while they work, and record everything people are doing on their computers. While it’s understandable, it also feels insulting. Now let me tell you an even more uncomfortable story. This is true, and it is about my brother in law. Let’s call him Jack.
Jack bought a new vehicle with a safety feature that automatically stops and shuts down the car upon impact. This escalated into a real problem on the highway. He hit a pothole in the road a little bit too hard, and his automobile shut down in the middle of a busy highway. He couldn’t turn it back on. The block in traffic caused an angered and impatient man behind him to exit his car and start a physical fight. This did not end well, and police were called in. A tow truck took his new car away, and Jack ended up in the hospital. Not a good situation for a safety feature to be auto-employed by artificial intelligence.
AI has its faults. We have ours, too. Machines cannot take into account the unpredictable psychology of human nature in situations like busy traffic. It can’t choose the best possible scenario for stopping a moving vehicle, because there’s no way to access every angle of a situation when one encounters a bump in the road. We’re much too fickle to gauge. But what we can do is understand how this technology works, so that we might opt out of a car with that sort of safety feature.
We can also learn the logistics of using AI in our own marketing campaigns. If the algorithms are more understood, our marketing plans can be tweaked to work alongside them. For example, if you know Google search now uses AI to generate an answer to a question at the top of the search results, you know you can write your website text in a way that gives you a better chance of being picked up as an answer to that question. Say Jack wants to look for cars that don’t use safety shutdown features. He types in “What new cars do not have automatic shut off?” (In America), and Google comes back with its generative AI answer stating “GM removed start/stop technology from 2021 full-size trucks and SUVs with V8 engines.”
This is because the website Autoblog.com optimized a blog post by using the title “GM removes start/stop from full-size trucks and SUVs with V8 engines”, and started the article out with the sentence “GM is removing start/stop technology from a number of its popular full-size SUVs due to the ongoing chip shortage. This follows the removal of cylinder deactivation technology from some of its trucks, which was also due to the chip shortage.” The writer began the blog and designed the title to look like an answer to a question someone might ask, and they positioned these things first on the page so Google would pick it up.
This is how you get your information on the front lines in 2024.
Another great thing you can learn to do with AI is to use it for mundane, basic tasks that take time away from your art. It’s great at identifying target audiences, and it increases efficiency by doing repetitive tasks like analyzing social media data and helping build followers according to listeners' habits. Think of it as having a personal assistant, and delegate tasks to it as if it were.
As far as generating music is concerned, that doesn’t have to be feared. Yes, AI music generation is a big thing. But it still sounds a bit “off” at this point. That’s because there’s no emotion. It’s simple math and copying data sets of music that already exists—with scientific precision. Here’s a way to use that positively.
If you have a secondary income stream of mass producing jingles and selling to royalty free platforms, you can generate musical ideas. You don’t generate a tune through AI and use it as is. You generate the idea, and you take that idea and turn it into something else. We all run into what I call “creative exhaustion”. I don’t believe in blocks… I believe we simply get tired if overworked. And when you’re on a deadline, you don’t have the time to wait until you’re inspired to put out more music. In that scenario, I don’t see a problem with generating an idea and turning that into a real song. I personally would never use it as it comes out of a machine; that always sounds unnatural and frankly, too mathematical. I don’t listen to music for the math. Most people don’t. But you can treat a 2 minute AI generated tune like a whistle you might hear in the distance, and turn it into something completely different. It’s a faster way of coming up with enough music to upload to Epidemic Sound, for instance, and make the rent on time.
I’ve used AI in writing to give me outlines and screenplay ideas. I’ve used it for research into career management and in my banking choices. Sometimes I talk to it like a secretary, using voice prompts and putting project plans together. It’s much cheaper than an assistant, and most self employed artists could really use this kind of machine based business collaboration.
Try it out. Don’t let yourself become so afraid of this new thing that you get completely taken off guard when the rest of the world is moving faster than you are. As independents and freelancers, we need all the help we can get, and AI is a great help if you really know how to use it.