Two years ago, OpenAI GPT-3 brought artificial intelligence out of the valley of tears. In the years before, when neural networks and machine learning were hype, AI wasn't as far along as people wanted it to be. AI was able to:

  • Detect objects in images and videos
  • Analyze texts for sentiments and keywords
  • Digest large sets of data to detect a pattern
  • Use time-series data to learn and generate predictions

What was missing was the creation of things (content, thoughts, ideas, ...). OpenAI GPT-3 made the first step here: it could magically take tasks and generate text based on them. And not just template-based text generation, but truly original and creatively written texts.

You can experiment with OpenAI GPT-3 on https://beta.openai.com/playground/

You would think the next big step until AI can do something new again would take years. But the thing about technology is that it doesn't get linearly from year to year (10% better, then 20% better, and so on) but exponentially (next year twice as good, then another year twice as good, and so on).

And that's why in early 2021, the world was surprised when Dall-E was initially announced. Now, in 2022, week after week, new approaches are being launched by different teams (Dall-E 2, Stable Diffusion, Midjourney) to use AI to generate and manipulate images.

There are already enough articles on the web showing the workings and examples, so in this article, I'd like to briefly mention how you can try it yourself, assuming you have a MacBook with M1 Chip.

https://diffusionbee.com/

With DiffusionBee, you can create Stable Difussion art on your computer.

Excitingly, Stable Diffusion is an approach that doesn't require that much computing power to work. You can even run it locally at your site on your computer. DiffusionBee is a small application that makes it easy for non-techies to try it out. Just go to https://diffusionbee.com/ and click "Download for MacOS".

ℹ️
For non-MacBook users, you can try it on HuggingFaces, but it might sometimes be overloaded.

Here are some Prompts and results I got for your Inspiration:

It is a gloomy autumn day. A woman walks down a street and gradually fades away. Her lover looks longingly after her. (Without the Surrealist suffix of this post's teaser)
A futuristic dark noir style city. A detective stands on a balcony and smokes a cigarette, Digital Art. (Where's my smoking detective? :( )
A futuristic dark noir style city. A detective stands on a balcony and smokes a cigarette. By Greg Rutkowski and Claude Monet, oil on canvas.
A futuristic dark noir-style city imagined in a fantasy world in lord of the rings. By Greg Rutkowski and Claude Monet, oil on canvas. (Wow!)
A fun summer vacation in Norway is a road trip with a lot of mountains. Photorealistic style. (Where's the fun?)
An epic scene of a spaceship falling from the sky. Art Nouveau, Neoclassical, Fantasy, Digital Art.
Art Hub AI - Discover, upload and share AI generated art pieces.
Discover, upload and share AI generated art pieces.
-

Check out https://arthub.ai/ for more inspiration on potential prompts.

Subscribe to my Deliberate Diligence blog for more ideas and thoughts on how to lead a deliberate diligence lifestyle:

Become a Deliberate Diligencist!