Image Gallery

To create our image gallery, we wanted to show the range of styles generative AI is currently able to produce. But rather than a simple tech demo, we also wanted to use the gallery to prompt consideration regarding the potential impact of AI technologies. This meant that we wanted to create a collection of images that could elicit a range of connections and responses.

We had some early success with Stable Diffusion in rendering more painterly and artistic styles and decided to include some the early generations (July 2023) in the gallery. Some examples that we quite fancied include:

We also wanted to include a range of styles. As such, we opted for images created in DALLE-3/Bing Chat that included graphic and more artistic styles. It is worth noting that generative AI applications have improved significantly over the past six months in terms of generating ‘professional’ looking outputs. This does mean that it becomes more difficult to elicit a more artistic output that strays from a more industry oriented focus. For further context see our ‘Wonky House’ post.

By testing a range of prompts (e.g. “Image depicting the potential positive and negative implications of AI”, we able to achieve some interesting outputs that we hope will invite viewers to consider some of the possible societal outcomes of these technologies. Here are some of our selections in response to a more surrealist style:

For comparison, here are some other examples with a focus on a minimal poster design (also DALL-E3/Bing Chat generated):

We also wanted to test some of the unique features of other AI applications such as Art Breeder. This online tool has the standard text-to-image option, but also allows you to mix images to create a blended output.

The final image gallery includes a mix of styles. We do hope that the selection we included provokes consideration of positives, negatives, things to think about and maybe also just a bit of joy to spark further investigation into the future applications and developments of AI.