The current approach to image generation technologies involves prompt engineering, which requires users to be very precise with their prompts and include negative prompts to guide the model. However, this is not an ideal future where users have to go through these hoops every time. The goal is to guide the model to a high-quality outcome without requiring users to be extremely precise. Currently, users of Google are considered good if they can use the smallest number of words to get their desired result, and Google has made significant progress in making this possible. However, the current approach to prompt engineering for models like stable diffusion is the opposite, requiring users to write paragraphs to achieve their desired outcome. While some degree of precision may always be necessary, relying on hacks like trending on artstation or specifying resolution should not be required. These models are capable of generating both high-quality and garbage outputs, and they can learn to avoid the bad stuff if properly guided. Therefore, the user interface and ease of use in guiding the model to desired results are crucial factors to consider for future development.

Get the Snipd
podcast app

Unlock the knowledge in podcasts with the podcast player of the future.
App store bannerPlay store banner

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode

Save any
moment

Hear something you like? Tap your headphones to save it with AI-generated key takeaways

Share
& Export

Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode