Can AI Truly Be Creative? Let's Break It Down

Large language models (LLMs) have been shaking up various domains, showcasing human-level and sometimes superhuman performance. But here's the crux: are they genuinely creative? The debate is heated, largely hinging on how we define creativity and the criteria we use to evaluate it.

Functional vs. Ontological Creativity

Let's break this down. On one hand, there's functional creativity, which looks at the tangible outcomes of creative processes. These models can churn out impressive work that might easily pass as creative by human standards. Yet, the reality is they often fall short of reaching the most sophisticated levels of creativity. They're like talented apprentices, but not yet masters.

On the flip side, we've ontological creativity. This perspective digs deeper, examining the underlying processes and the social and personal dimensions that define true creativity. Here, LLMs face significant challenges. They lack the intrinsic motivations and personal experiences that shape a human's creative journey. Strip away the marketing, and you get a system that's more mimic than muse.

Should AI Strive for True Creativity?

Is it even desirable for AI to achieve both forms of creativity? That's a big question. Advocates argue that bridging this gap could unlock new potentials, enriching human society in ways we can't yet imagine. But the numbers tell a different story. Risks like ethical concerns and misuse need careful consideration.

Frankly, the architecture matters more than the parameter count here. As we push toward artificial creativity, the focus should be on building systems that complement human creativity, rather than replace it. Is AI ready for that leap? The jury's still out.