Loading video...

Video Failed to Load

Go Home

Using ChatGPT to make infinite image prompts for Midjourney In <30 seconds you'll be able to generate an infinite amount of food photos. Hint it works on anything else too 🤯 No need to know anything about photography. Let's make infinite hamburger prompts 🍔 🧵 A thread

924,427 views • 3 years ago •via X (Twitter)

13 Comments

Linus Ekenstam – eu/acc's profile picture
Linus Ekenstam – eu/acc3 years ago

First our initial prompt, this will set context and help the AI move forward. "Food photography" can be anything First prompt: 💬 Generate an elaborate list the key elements used in an food photograph separate by comma, explain in short below the bullet list.

Linus Ekenstam – eu/acc's profile picture
Linus Ekenstam – eu/acc3 years ago

Second prompt: 💬 "using these elements, tabulate and populate 5 rows of example data, where food is hamburger" We can just populate as much as we want here.

Linus Ekenstam – eu/acc's profile picture
Linus Ekenstam – eu/acc3 years ago

Third prompt: 💬 "Take row 1 and write it out separate each col with a ," You can ask for any of the rows, or copy past yourself, this just makes it faster.

Linus Ekenstam – eu/acc's profile picture
Linus Ekenstam – eu/acc3 years ago

Our first collection of burger shots: We got a red checkered tablecloth, natural light

Linus Ekenstam – eu/acc's profile picture
Linus Ekenstam – eu/acc3 years ago

Our second collection of burger shots: Negative space, spotlight, rough and dramatic.

Linus Ekenstam – eu/acc's profile picture
Linus Ekenstam – eu/acc3 years ago

Our third collection of burger shots: Artificial light, softbox diffuser, asymmetrical position

Linus Ekenstam – eu/acc's profile picture
Linus Ekenstam – eu/acc3 years ago

Our forth collection of burger shots: Rustic wooden table, natural light from window, minimalist burger

Linus Ekenstam – eu/acc's profile picture
Linus Ekenstam – eu/acc3 years ago

Our fifth collection of burger shots: A stack of meat, jalapeños for everyone, cluttered

Linus Ekenstam – eu/acc's profile picture
Linus Ekenstam – eu/acc3 years ago

This simply illustrates how fast and easy it will be to generate anything visual without knowing anything about visuals. You know what you want? type it out, and the AI will do the rest, no need to even know exactly how you want it. AI will change a lot of things.

Linus Ekenstam – eu/acc's profile picture
Linus Ekenstam – eu/acc3 years ago

Learn more about Dynamic Prompting here, including tutorial for how to structure your prompts:

Linus Ekenstam – eu/acc's profile picture
Linus Ekenstam – eu/acc3 years ago

Want early access to a simple GUI tool I've been working on to make this even more simple? Join +2800 thats already on the early access list.

Linus Ekenstam – eu/acc's profile picture
Linus Ekenstam – eu/acc3 years ago

Thank you for reading this far, reply with what you create with this little hack, looking forward to see your creations. Ps. it works in Dalle2, StableDiffusion, Leonardo, BlueWillow, Scenario not just Midjourney.

Linus Ekenstam – eu/acc's profile picture
Linus Ekenstam – eu/acc3 years ago

I hope you've found this thread insightful. Follow me @LinusEkenstam for more AI content Like/Retweet the first tweet below to share it with your audience.

Related Videos