.png){kind=logo}
AI’s creative influence was once behind text suggestions and grammar corrections. Fast-forward to today: AI models make cinematic trailers, design brand identities, write poetry, and even create hyper-realistic avatars of non-existent people. The change is not just technological; it is also cultural.
Early models like OpenAI’s GPT-2 and GPT-3 created the foundation for text generation. However, the modern story has gone a few steps further with the coming of multimodal models. Now, DALL·E, Midjourney, and Stable Diffusion allow people to translate text prompts into highly detailed art and digital images. At the same time, Runway ML and OpenAI’s Sora are breaking new ground in video-making, enabling even beginners to envision entire movie scenes from a few lines of description.
At the centre of these innovations are diffusion models, which construct content from random noise, progressively developing it into meaningful images or sound. This mechanism has opened up avenues of creativity that are new to industries, ranging from advertising to entertainment.
AI's impact isn’t confined to the aesthetic or literary domain; it pulses with new life into science and research processes. Deepmind’s Alphafold has set an illustrious example for the field by solving the half-century challenge of predicting protein structures, an achievement considered a watershed moment in biology.
IBM WatsonX and NVIDIA BioNeMo are equally helping make breakthroughs in healthcare. They use generative models to model molecular structures and enhance drug discovery. In environmental science, AI simulates climate scenarios and optimises clean energy solutions.
Christopher Bishop, head of Microsoft’s AI for Science program, thinks we have only just begun to see AI’s potential as a research tool. “We’re applying AI not only to analyse data but to pose questions we didn’t know how to ask in the first place,” he said.
As more creative generative AI is developed, questions about the essence of creativity itself are being raised. Can an algorithm, based on existing data, create something truly new?
Critics feel that although AI may be able to mimic styles and mix ideas, it lacks intent, emotion, and living experience, the essentials of authentic artistic expression. Take-Two Interactive CEO Strauss Zelnick expressed this concern before GTA 6’s release: “AI is based on reused data. Big hits have to be made from scratch.”
Even the developers of software such as Runway, Luma AI, and Pika Labs, which enable users to edit and create videos with basic commands, recognise that AI’s purpose now is co-creation rather than complete autonomy.
With great power comes great controversy. Generative AI has created a firestorm around ethical use, copyright infringement, and digital ownership.
Artists and illustrators have complained about models such as Artbreeder, Dream by Wombo, and NightCafe for using their work without permission. The internet has also been filled with AI-generated artwork replicating artists’ signature styles, such as Hayao Miyazaki and Banksy, raising serious questions regarding consent, credit, and fair use.
Audio AIs such as ElevenLabs and Voicemod can clone voices with chilling accuracy. Deepfake artists employ them to impersonate celebrities or politicians, producing lifelike but completely fabricated video and audio recordings. The tech is mighty, but so is the capacity for misuse.
Despite the apprehensions, most creators are welcoming AI as a collaborative partner. Musicians employ software such as AIVA, Soundraw, and Amper Music to create game and movie soundtracks. Writers and marketers use Jasper, Copy.ai, and Writesonic to break writer’s block or simplify content creation. In schools, Khanmigo, an AI tutor by Khan Academy and OpenAI, is tailoring learning like never before.
Generative AI is not rendering creatives unnecessary; it’s liberating them to concentrate on vision, emotion, and story, with the machine performing repetition and scale.
Generative AIs have emerged at the watershed between technology and our very conception and distribution of creative labour. As ever-greater humanly creative tasks become amenable to machine acts, one must ethically draw the line regarding generative models and inform the systems fairly regarding attribution and ownership.
Ride the wave; don’t resist!
The question isn’t if AI will transform creativity; it already has. The question is how we, as creators and custodians of innovation, build this future. Do we employ AI to expand human expression, or do we let it homogenise it?
In this next chapter of creativity, the most influential instrument may not be the machine but the human hand that controls it.
