As artificial intelligence (AI) becomes more widespread in everyday life, the art industry is increasingly attuned to the creative and market potential of AI-generated artworks. Although many artists fear that machines will hinder the essence of art and may even lead to the replacement of people as a creator, AI artists believe that automatic art is a new opportunity and that this process becomes an art form in itself.
My introduction was written by an algorithm. If I hadn’t told you, would you have been able to tell? It’s a bit abrupt for an opening paragraph, but otherwise it’s not bad—grammatical, logically sound.
Artificial intelligence has been making headlines for a while as it has grown in sophistication, but its applications used to be almost exclusively data-driven. More recently, however, developers have made huge strides in honing AI’s creative side. From generating art to news headlines and chatbot conversations, humans are consuming more computer-generated content than they probably realize.
Not everyone is excited about this, however. There’s a contingent of creative professionals fearing the chopping block should AI develop enough to create better, faster work than they can. We don’t share those feelings at U.Group—we see it as an opportunity.
As technology has advanced, we haven’t seen people’s job be replaced by robots. In fact, the opposite is true. Humans have moved up the food chain. We’re now responsible for owning and optimizing the technology—we hold the keys to the future.
AI’s creative applications hold the possibility to enhance human efforts and save time, allowing us more time and energy to do what only we can do. Today we’re rounding up the reasons why we’re enthusiastic about the development of AI’s creative side, and why you should be too.
3 Reasons Why Creatives Should Embrace Artificial Intelligence
Pattern recognition isn’t the same as creative expression
There are some similarities between the way humans and computers generate content—both process information and output it in an innovative piece of written or visual art—but AI references data whereas humans reference prior works and their surroundings. In order to generate natural-sounding language or believable artwork, the most common method is to train neural networks on a huge amount of data of the type you’d like it to later generate. These neural nets identify patterns from the inputted data and then create a new piece of writing or artwork based on those patterns.
Although AI is also capable of creating entirely new—even unexpected—outputs from those patterns, it’s not the same as human art. Machines don’t quite grasp nuance and shared cultural experiences. They can only evaluate data at face value, and creativity relies heavily on interpreting abstracts and metaphors. Cards Against Humanity, oddly enough, illustrated the difference beautifully during their 2019 Black Friday marketing stunt. In a cutthroat battle against a neural net trained on NSFW cards and writers’ notes, the writers scrambled to produce a better expansion pack than the machine.
Although the neural net produced more cards, the quality of the writers’ pack was better and sold 2% more card sets. The AI-generated cards all hit on the right tone for the offensive brand, but some were nonsensical or fell flat, offensive merely for the sake of being offensive. The writers’ cards, on the other hand, touched on trending current events and other cultural nuances that gave them a more relatable edge. When it comes to creative assets, that human, cultural touch makes all the subtle difference between cognizance and resonance.
AI isn’t fully autonomous—yet
As a still-developing area of tech, AI has a long way to go before it could plausibly create works of visual or written art unassisted—the algorithms are only as good as the data we give them. Human minds are involved on either end of the program, controlling the data that goes in and editing the resulting product.
The intro paragraph above, for example, was generated based on keywords I selected. The neural net then took those parameters and analyzed, combined, and reworded portions of over a dozen published news articles on AI-generated artwork. I needed to go several rounds with the generator finetuning keywords, deselecting unrelated data sources, and selecting the proper balance between coherence and originality. The human effort is even greater during the training process. The level of human involvement also varies by medium, but the bottom line is that humans are still integral to algorithms’ creation process.
AI’s level of success also varies wildly depending on the content form. Algorithms are widely used to generate simple news articles and headlines—the Press Association uses AI to crank out 30,000 local news stories each month—but it’s doubtful that a bestselling AI-generated novel will hit the shelves anytime soon. Take this Lord of the Rings battle scene writing prompt, for example. OpenAI’s GPT-2, which is currently the most coherent text generator available, still struggles with repetitiveness, plot continuity, and worldbuilding (GPT-2 wrote about fires underwater, for example). Even in a future where AI can generate reliable written content, human eyes would still be essential for editing. It’s safe to say, then, that human roles may change and evolve, but they won’t disappear.
Humans and machines can work together
Given that AI still needs humans to curate data sources and polish output, I envision the human-machine relationship being beneficial rather than antagonistic. If AI steps in to perform more menial or research-based tasks, it will free up human resources and allow us to focus on bigger-picture work. For instance, an algorithm can analyze hundreds of websites to provide recommendations on website design based on what competitors are doing far faster than a human’s eyes could perform the same task. The designer could then look at machine recommendations and have more time to elect and execute the design from there. In this scenario, AI helps enhance human capabilities and save time—something we can all benefit from.
In a time where many people fear what the future of technology has in store, U.Group is optimistic…excited even! As strategists, technologists, and designers it’s our job to create solutions that solve complex challenges (and don’t exacerbate them). If you’re interested in the intersection of technology, society, and culture, it’s a topic we take on frequently. Sign up for our newsletter and be on the lookout for our weekly Instagram series where we address the week’s most compelling tech topics on our Stories.