Sure, AI chatbot can write fiction and conjure images. But can a chatbot, like a human author, relates and intimates on the events in life, and the ideas that drive and inspire the story? We look at the fear and loathing directed at a web scraping function now known as AI – and also optimism that comes with it.
Currently, there’s plenty of anxiety about the implications of AI within the creative industry. Machine learning chatbots and text-to-media AI technology such as ChatGPT, Lensa and Stable Diffusion have created a massive disruption to the creator’s economy landscape. They manifested as either problem-solving opportunities or employment killers, depending on where you stand as a creator.
Like most tech disruption stories, the burning issues are copyright theft and the moral debate of making workers redundant in favour of the AI. Reading up about the uses and abuses of AI is daunting, especially after scrolling on countless articles on the copyright infringements.
The copyright law doesn’t apply to an art style
The creatives are sceptical for a good reason. There are plenty of grey area that puts creators off from exploring the full potential of the AI tools.
Take for example, the Lensa AI app that could generate dozens of fantastic avatars through its contemporary art filters. To do that, the machine learning ‘borrows’ freely and heavily from individual artists styles. Unfortunately, at the moment, there’s nothing that the artists could do about financial compensation because “while individual artworks are subject to copyright, the stylistic elements and ideas behind them are not.”
But the image makers aren’t taking it lying down.
At the time of writing, Getty Images is suing Stable Diffusion for a whopping US$1.8 trillion (£1.5 trillion) for allegedly stealing millions of their images to train the AI model. In some of the cases, the Getty Images watermark was clearly visible on the images created by Stable Diffusion.
Plagiarism and copyright theft aren’t a new phenomenon in visual arts. The introduction of the AI image synthesizer models exacerbates the problem further by scaling up plagiarism at an industrial level.
We need to redefine by law on what constitutes “stealing”, “cheating” and “fair use” since replicating become normalised through using AI.
ChatGPT is heaven-sent, depending on who’s asking
Academia isn’t ready to accept ChatGPT as ghost writers. Currently, college professors could detect essays written by the chatbot in students’ work. Darren Hick, a philosophy professor from Furman University in South Carolina was one of the first tutors to publicise catching a student cheating with ChatGPT.
He claimed that the chatbot’s style is recognisable and “writes like a very smart 12th-grader.”
In the age where influencers thrive through aping and faking a charmed life through social media platforms, critical thinking and telling an authentic story are rare skills to have.
A quick query for “orchids by the seaside” on the Stable Diffusion app produced beautiful images within seconds. Nowhere on our planet is a place this perfect. We also wondered whose copyrighted orchid images that Stable Diffusion scraped from.
It’s not all doom and gloom. ChatGPT is received more positively by designers to help solve technical problems.
For instance, ChatGPT could come up with an answer to a coding problem in seconds, saving designers days of diving into tech forums or waiting around for customer support to troubleshoot issues.
If the technology is used correctly, we could finally realise the Keynesian dream of working 15 hours per week. A designer’s perfect companion is a smart machine that doesn’t need to sleep, eat, and could render possible solutions round the clock.
However, what Keynes under-estimated is the society’s capacity for the devious, as well as the good. It’s been reported that the ChatGPT bot was tricked into giving bomb-making instructions.
So, the juries are still out on machine learning technology.
“ChatGPT and Stable Diffusion are tools that can help us create more products, but they can’t offer us an authentic experience without the human input.”
There’s no tears in the rain yet
We’re still at an early stage of the AI-assisted life. Results are chaotic and still up for debates. Whilst there’s no elegant soliloquy from a replicant yet, we’re hopeful about our future with robots.
For book fans, no chatbots could replace the magic of George RR Martin, Stephen King, Mary Shelley or Neil Gaiman yet. An author’s journey behind the making of each book is as important as the story.
A human author writes a book because he or she goes through an authentic human experience in life that prompted a revelation that needed to be expressed through cathartic storytelling. The audience then would form a bond with the story characters to modulate their emotions in real life.
ChatGPT and Stable Diffusion are tools that can help us create more products, but they can’t offer us an authentic experience without the human input.
The AI may come up with a perfect story, but it still will need to borrow or learn from an actual human experience to come up with a machine-generated performance.
We came across a TikToker who suggested that anyone now could write their own eBooks using ChatGPT and produce the cover art in Canva. The responses are hilarious, if not brutal.
“If you need ChatGPT to write a book and Canva to design the cover, you deserve to be replaced by AI.”
“Theoretically possible. Practically, a lot more work than you make it sound. And it’s going to be a dull ebook…”
“You forgot the part where no one wants to buy it, so you just wasted your weekend when you could have been gaming.”
“When I can use ChatGPT to make my own book, I don’t have to buy yours.”