Back to all posts

Five Mind Bending Prompts To Test ChatGPT 5 Now

2025-08-27Daniel Ichbiah and Futura Team5 minutes read
AI
ChatGPT
Prompts

OpenAI's latest model, ChatGPT-5, is here, representing what many see as a major leap forward in generative AI. With reasoning capabilities purportedly on par with a PhD, this new version is pushing the boundaries of what we can expect from artificial intelligence. It's more than just an upgrade; it's a glimpse into a future where AI can handle complex, multi-faceted tasks with greater precision and reliability. But what does this power look like in practice? The best way to understand its potential is to see it in action.

The Next Leap in AI: What Can ChatGPT-5 Do?

Before diving into the prompts, it's worth noting the key enhancements ChatGPT-5 brings to the table. It is now fully multimodal, accepting text, image, audio, and video inputs to generate comprehensive outputs. Other significant improvements include:

  • Sharper Reasoning: The model demonstrates more reliable and nuanced logical capabilities.
  • Agent Mode: It features a built-in agent for automating complex, multi-step tasks.
  • Expanded Memory: ChatGPT-5 can now recall information from and work across multiple long documents, making it a powerful tool for research and analysis.

Now, let's explore five mind-blowing prompts that showcase these new abilities.

1. Code a Playable Video Game from Scratch

One of the most impressive feats is its ability to generate functional code for a complete application. We challenged it to build a classic arcade game with a modern twist.

The Prompt: "Make a modern Space Invaders game in one HTML/CSS/JavaScript file. Slick pixel-art visuals, animated background, built-in sound effects. Include score, lives, progressive difficulty levels. Keyboard controls and mobile-friendly buttons. Code should be responsive, clean, and fully commented."

With this single request, ChatGPT-5 produced 520 lines of well-commented code and a playable preview. After a few minor adjustments, the result was a fully functional and addictive game, ready to be embedded on a webpage.

A screenshot of a pixel-art Space Invaders game created by ChatGPT-5. ChatGPT-5 produced a complete Space Invaders, which then simply had to be copied onto a website. © OpenAI

2. Generate Artful Poetry in Visual Form

Beyond code, ChatGPT-5 exhibits a sophisticated creative flair. We asked it to merge poetry and visual art by creating a calligram, a poem where the text is arranged to form a thematically related image.

The Prompt: "Write a nature-themed poem shaped like a tree (calligram). Visual layout should form the branches and leaves. Provide a 16:9 illustrated version."

It returned an elegant visual where the verses of the poem beautifully formed the foliage and branches of a tree, demonstrating a deep understanding of both literary and visual design.

A poem whose text is arranged in the shape of a tree, with branches and leaves made of words. A poem whose verses draw the foliage of a tree. © OpenAI

3. Create a Detailed Sci-Fi Movie Storyboard

With its enhanced visual consistency, ChatGPT-5 is now a viable tool for filmmakers and storytellers. We tested its ability to generate a complete storyboard for a movie trailer concept.

The Prompt: "Make a 10-frame storyboard for a sci-fi movie trailer. Plot: a scientist discovers how to travel through time. Each frame should include: Visual description (camera angle, color, lighting), Key action, Soundtrack or audio cues, One line of dialogue or narration"

The resulting storyboard provided a clear and visually detailed structure for the trailer. While some frames required refinement, the overall consistency of characters and scenes was impressive.

A vertically oriented storyboard with multiple frames depicting a sci-fi scene. ChatGPT-5 maintains visual consistency of characters. It still struggled to produce this storyboard. © OpenAI

4. Analyze and Critique Video Content

ChatGPT-5's multimodal capabilities extend to analyzing video content. You can provide a link or upload a file and ask for a detailed breakdown. We tested this feature with a music video.

The Prompt: "Can you review this video: https://youtu.be/44Bf3_J140U. Which shots need improvement? What alternate backgrounds could work better?"

It delivered a thorough analysis covering visual quality, lighting, and composition. It even provided creative suggestions for improvement, such as using a "blurred urban background with purple and blue neon to reinforce the Orwellian tone of the music."

A screenshot of ChatGPT-5's interface showing its analysis of a video. ChatGPT-5 can analyze a video and suggest improvements. © OpenAI

5. Synthesize Long Documents into a Presentation

Finally, we tested its ability to process and synthesize large amounts of text. We gave it a collection of articles and asked it to create a presentation.

The Prompt: "I’m giving a talk on AI tomorrow. Here are 10 texts I’ve written. Can you analyze them and draft a 20-minute presentation for beginners? Include key topics, questions, and engaging hooks."

The AI quickly produced a well-structured, public-ready presentation outline. It suggested a title, an engaging introduction, a brief history of AI, and a clear explanation of complex concepts like the difference between generative AI and AI agents, all based on the provided documents. This highlights its potential as an invaluable research and productivity assistant.

Read Original Post
ImaginePro newsletter

Subscribe to our newsletter!

Subscribe to our newsletter to get the latest news and designs.