Not Faster AI - Now It’s Trying to Be More Creative

AI keeps getting smarter, but now it’s getting… creative? OpenAI has released an AI model designed to write like humans, generating stories, poetry, and narratives that sound more natural than ever.

But before we get excited (or concerned), let’s break down what’s actually happening here.

What Is an LLM? And Why Does It Matter Here?

To understand OpenAI’s new model, we need to talk about what a Large Language Model (LLM) is.

These AI systems work by analysing and predicting patterns in text. They don’t think, feel, or imagine - they recognise patterns based on frequently seen data.

🔹 LLMs are trained on billions of words from books, articles, and the internet.
🔹 They predict the most likely next word in a sentence based on what they’ve seen before.
🔹 They are not creative in the human sense - they replicate existing structures.

So, when OpenAI claims its model is proficient in creative writing, what does that actually mean?

The New “Creative” AI Ability

OpenAI’s latest model isn’t about answering questions; it’s trying to write like an author. This means:

✅ Better story structure - beginnings, middles, and endings that actually make sense.
✅ More consistent style - it can mimic or predict different tones without losing coherence.
✅ More engaging narratives - dialogue flows better, and emotions feel more real.
✅ Improved humour, irony, and metaphor use - it recognises and generates more nuanced writing.

Sounds great, right? But here’s the thing…

What Does “Creative” Actually Mean?

The way AI generates text is based on what humans have already written and what humans have defined as “creative” versus, for example, how you write an email.

Different training data define different outputs, and the word “creative” serves as a command that signals the LLM to rely on specific training when generating text.

AI doesn’t create something entirely new; it rearranges known elements from its training in ways that feel fresh. This raises a big question:

👉 Is AI actually creative, or is it remixing human creativity?

If creativity is about inventing new ideas, AI isn’t really doing that. It’s producing variations of what already exists.

This means that:

🔹 AI is “creative” only within the boundaries set by its training data.
🔹 AI doesn’t have personal inspiration or experiences—it’s pattern-based.
🔹 AI-generated work is a reflection of human knowledge, not a breakthrough in creativity.

And here’s where things start getting messy… we could say the same about humans. But humans make the rules, so we need to stand by them.

Ethics: The Copyright Collision

This new AI raises serious ethical concerns, especially regarding copyright and intellectual property.

❗ AI is trained on human-created content.
❗ It generates work that sounds like human writing.
❗ But who owns the final product?

Imagine an AI writing a poem in the style of a famous author. It’s not an exact copy, but it’s also not original.

So:

🔹 Did AI plagiarise?
🔹 Should the original writers be credited (or compensated)?
🔹 How will we prove who wrote it and credit them? Per word?
🔹 What happens when AI-generated books start competing with human authors?

The legal world is already struggling to keep up, and there’s no clear answer yet.

Which brings us to another question—what is the real breakthrough here? Because so far, old dilemmas are being masked as new AI abilities.

What Is the True Breakthrough Here?

🔹 The biggest advancement isn’t creativity itself—it’s AI’s improved ability to remember more.
Unlike human brains, which naturally rewire and adapt, AI needs infrastructure to do this. Expanding memory capabilities means AI can retain context better, generate more cohesive long-form content, and mimic continuity in storytelling.

🔹 This isn’t about AI gaining true creative insight—it’s about engineering better memory structures that allow it to build on previous inputs instead of resetting each time.
This is a major shift in how AI processes information, but it’s still a technical leap rather than a conceptual one.

My Thoughts

AI writing is getting better, but it’s still built on human ideas.

OpenAI’s new model is and isn’t truly creative in the way humans are. In this discussion, we tend to forget that humans are also inspired by different artists and authors. The difference is that humans follow rules that define what counts as inspiration and what is considered theft.

We are asking big questions because we don’t fully understand what AI is and what it can do.

AI is a concept, not a tool, and the more we shift our thinking to see AI as a change in how we use our minds, the more we will achieve two things:

1. We will better understand ourselves and the power of human thinking.
2. We will learn how to use AI effectively to achieve our goals.

   Need new ways to think about AI? Let’s talk!