AI image generators have greatly improved. You can now use natural language to describe the kind of image you need. You can specify the genre and style you’d like and well at the “materials” that are used.
DALL-E was launched last year and can generate digital images from natural language descriptions called “prompts.” It uses a version of the GPT-3 modified to generate images. After 25 free images, you must pay per image generated.
As an example, I used it to generate these statues and murals of people using laptops:
MidJourney is an AI that also launched last year. MidJourney is probably the best AI image generator currently available. It produced the featured image of this article.
However, MidJourney is still in beta, and it can only be accessed by joining a Discord server.
MidJourney has hundreds of painting and drawing styles available to use. It can produce great results, but it will also result in “uncanny valley” style ones. To truly understand its strengths and weaknesses, it’s best to practice with AI for a few days.
If you pay $30 per month, you can generate unlimited images in “slow mode,” which means your images are generated in around 60 seconds instead of 20 seconds. If you use “fast mode,” instead, the number of images you can produce per month is limited.
Here are some images produced by MidJourney: