To use AI wisely, humans must summon morality

The Z Factor

Last month, OpenAI’s ChatGPT was trending for its ability to generate animated images in an artistic style copied from Japan’s famous Studio Ghibli. The trend alarmed artists and writers worldwide. If AI can so easily replicate a style of art that took years to perfect, it threatens the intellectual property rights of creative people everywhere.

Yet, this is what AI is effectively designed to do: industrialise creativity. AI is the knowledge economy’s equivalent of the 18th century Industrial Revolution. Old inventions such as the steam engine and the spinning jenny mimicked the physical movement of human muscles to haul water or spin yarn – only much faster than any human could manage. AI does the same thing with creative work. By mimicking the mental processes of human brains, AI can write, draw, or design in a few short minutes what the ordinary human might create over several hours or days.

There’s no reason why we shouldn’t be as enamoured with AI as the premodern world was with the spinning jenny. By designing content so rapidly, AI can mass-produce creativity and make it accessible to more people at a far lower cost. Yet, on the flip side, this could leave creative humans with little incentive to create – no different from how the spinning jenny left humans with little incentive to spin yarn by hand.

Young and upcoming artists, writers, and musicians are more vulnerable to AI than established ones. Established creators already enjoy sufficient brand value and no AI-generated content can compete with that. ChatGPT might learn how to mimic Studio Ghibli’s style, but it won’t diminish Studio Ghibli’s legacy appeal. It might learn to write and sing songs like Taylor Swift, but it won’t take away Taylor’s ability to sell music to her fan base.

But undiscovered talent is already under threat. If AI can copy your style of art, writing, and music – and produce all of those things at a faster clip than you can – it leaves you far less space and opportunity to establish yourself as a creative artist.

As terrible as this might sound, however, AI’s real threat to creativity is not in its ability to copy and mimic but in its ability to disinform and discredit. In recent years, AI-generated images have been widely used to spread disinformation. In February, one photo falsely suggested that Israeli Prime Minister Benjamin Netanyahu had met Elon Musk in Washington. Another one showed a Muslim man dressed as a Hindu monk in an apparent bid to attack the Kumbh Mela this January.

Given these risks, many have argued that the use of AI should be regulated. But that’s easier said than done. It’s easy to license and regulate the use of firearms or motor vehicles because access to those machines can be controlled and their uses are limited. But AI can be designed and used by anyone with a computer and an internet connection for everything from building financial models to writing an email. And since its processes are unpredictable – even to its designers – accountability for its misuse will be hard to establish under any law.

All of this might sound like an argument to put an end to the proliferation of AI, but I’m no Luddite. Technologies once invented are never reversed. That is why despite several decades of diplomacy, the world has barely been able to prevent the proliferation of nuclear weapons to newer countries, let alone get rid of existing stockpiles. But more than any technology ever invented, AI will rely more on the ethics of human beings than it does on regulation to ensure its proper use.