The Wombo Dream Realistic style is a game-changer!

The speaker introduces an app called 'Dream from Wombo,' a text-to-image tool that uses AI to generate art based on text prompts. The app offers various styles, including a new 'realistic' style and a diffusion process. The speaker shares his experience with the app, noting that it can sometimes produce very surreal and interesting art, and other times the results may be less accurate. The app is presented as a creative and productivity tool that allows users to feel a sense of authorship over the generated artwork.

The speaker explores the app's ability to create surreal images by combining unusual objects like shoes, eggs, alligators, sandcastles, and starfish. He discusses the app's generative process, noting how it provides multiple options before settling on a final image. The speaker also mentions the app's speed, its addition of depth of field, and the viewer's engagement in guessing which image the AI will choose. The session includes live audience interaction, with the speaker encouraging viewers to submit their own prompts.

The speaker appreciates the app's bias against generating inappropriate content, including weapons, pornography, or anything sexual. He demonstrates how to save the generated images in different formats and discusses creating a video of the image generation process. The speaker also experiments with different styles, such as Picasso and Matisse, to see how the app interprets famous artistic styles when generating images of subjects like a crying Bambi at breakfast.

The speaker continues to experiment with various styles, including a Dali style, and attempts to generate an image of Bambi crying at breakfast in a black and white Escher style. He also tries to generate images based on phrases like 'the chat is ready to display messages,' resulting in unexpected outcomes like a cat in a phone chat. The speaker then moves on to creating images of Santa Claus scaling a skyscraper and discusses the app's limitations with certain prompts.

The speaker generates an image of a dog family picnic being invaded by a Santa Claus army. He expresses a desire to capture the unique moment and decides to create a video of the process. The speaker also attempts to generate an image of a praying mantis eating pizza and, after a few tries, successfully gets the app to include pizza in the scene when viewed from above.

The speaker creates images of snow-capped mountains at an orange sunset with heavy rain. He enjoys the process and the results, finding the generated landscapes fascinating. He then adds giant gorillas to the scene, which the app incorporates into the landscape. The speaker also experiments with the term 'photorealistic' to see if it changes the level of detail in the generated images.

The speaker attempts to generate images of a UFO and a dog's reaction to it. He plays with different prompts to get the AI to create the desired scene, including 'dog watching a UFO in the sky.' The speaker also tries to generate an image of an alien watching a dog in the sky, resulting in a surreal image of an alien dog. He concludes with an attempt to create an image related to 'bad breath,' which leads to a humorous and unexpected outcome.

In the final segment, the speaker creates images based on prompts like 'bad breath at breakfast' and 'Pee-wee Herman bad breath at breakfast,' leading to a bizarre and entertaining set of generated images. He expresses a desire to animate the results and tries to get the app to generate an image of a donut breakfast involving Pee-wee Herman. The session concludes with the speaker reflecting on the fun and creative potential of the app.