Prompting trends: Inside the minds of AI image creators

As generative AI evolves, understanding how different groups interact with this technology can provide valuable insights. We conducted a survey of 1,041 respondents for a snapshot of the diverse ways people prompt AI to create images, uncovering differences across generations, industries, and experience levels. From adoption rates to prompt length and word choice, our study reveals the approaches various demographics take when engaging with AI image-generation tools.

Key takeaways

Summary/Overview

Levels
1

Image prompting prevalence and learning methods

AI image generation has become increasingly popular, with a significant portion of the population embracing this technology. Our survey aimed to uncover the adoption rates across different age groups and explore how people acquire the skills to prompt AI for image creation.

Prompting preferences

AI can boost your creativity and supercharge your visual and written content, making it a popular tool for many applications. In addition to creating quick, simple visuals for business slideshows and reports, innovative AI tools are also useful for writing compelling product descriptions and enhancing images with added objects or backgrounds.

Across all age groups, more than four in five people have tried their hand at AI image generation. But Gen Z was the generation most likely to do so; 85% of them said they’ve used AI to generate images.

So, where do people learn to prompt AI for image creation? Self-study was the top method for 68% of respondents. But if trial and error is too time-consuming, learning from experienced AI users can help. According to skilled prompters who rated their abilities a four or five out of five, descriptive keywords were the top component of an effective AI image-generation prompt (77%).

The idea AI image prompt length

An AI image-generation prompt’s length can vary depending on what you’re trying to accomplish. But is there such a thing as too long or too short? Next, we uncover the average number of words used per prompt across different demographics and experience levels.

Prompt length

Gen Z had the shortest average prompt length at just 17 words, while Gen X used the most words per prompt (20.1). Professional industries also showed variations: respondents working in the retail sector used the longest prompts on average. Meanwhile, respondents working in business (professions like marketing, finance, and sales) used the shortest prompts to make AI-generated images.

We also had respondents rate their AI-prompting skill level on a one-to-five scale. Highly skilled prompters (self-rated four or five) tended to use more words in their prompts — an average of 19.6. This allows for more detail and context in their prompts to achieve desired results.

The language of AI image prompting

Let’s examine the specific words used in AI image prompts and see which are the most common for image generation.

T

“Image” and “create” emerged as the most common words, with 37% and 32% of prompters using these terms, respectively. Explicitly stating the intent to create an image may not seem necessary, but with how often AI fails to meet users’ expectations, it can’t hurt.

It might also seem unnecessary to say “please” when prompting AI — only 4% of respondents have included this polite term in their prompts. AI prompting can feel so much like talking to a person that it’s easy to forget you’re not.

Diversity in AI image generation

From Gen Z’s high adoption rates to prompts’ word count and detail across industries, it’s clear that a variety of factors influence how people use AI image-generation tools. By recognizing the diverse needs and preferences of different user groups, developers can create more intuitive tools and features that cater to a wide range of creators.

Crafting effective AI image prompts is a skill that people can hone through practice and learning from experienced users. As more people embrace AI image generation, we can expect to see even more creative and innovative applications of this powerful tool across industries.

Methodology

We surveyed 1,041 respondents to explore image-generating prompt habits and uncover generational differences in AI image creation. We analyzed the average number of words and sentences per prompt by generation, gender, and industry of work and also looked at the occurrence of certain keywords. The objective was to understand how different demographics interact with AI technology for image generation. Prompts with fewer than five words were not included. Of the respondents, 5% were baby boomers, 21% were Gen X, 51% were millennials, and 23% were Gen Z. Only industries with 50 or more respondents were used.

Fair use statement

Feel free to share these findings for noncommercial purposes, but please provide a link back to this page when doing so.

Try Adobe Express today