The 3 Best Artificial Intelligence Image Generators for Creators to Use

Original Source Here

The 3 Best Artificial Intelligence Image Generators for Creators to Use

Ditch the stock photos for some truly original digital art

An intrepid researcher uses art and science to fight K’gg’norlathsh, Avatar of Suffering (Image credit: DALL-E)

You’ve just finished writing an article and set out to find the perfect, royalty-free image to accompany your words and reflect your main ideas.

You scour through Unsplash, Pexels, Pixabay, Freepik, Nappy, ShotStash, and even Wikipedia Commons. But you’re unsatisfied with what you find. If only you could bring the image in your mind to life.

The boss has you under a tight turnaround deadline, so you don’t have time to hire an illustrator or graphic designer. Then you remember the buzz around AI image generators and decide to give that a shot. But where to start?

DALL·E 2: Join the Beta Waitlist

Since its inception, OpenAI’s text-to-image art generator DALL·E 2 has generated considerable buzz, provoking both horror and awe. The name is a portmanteau of Dalí (as in the Spanish surrealist Salvador of “melting clocks” fame) and Pixar’s WALL-E.

As you might infer, the current version is the second iteration of the tool, which creates new art based on the 650 million image and caption combinations it has already been fed.

The basic premise is simple: create uniquely original images with 400-character prompts.

A prompt is a sentence, 400 characters or less, that describes the image you want.

If you’re in need of some inspiration or want to learn about how to make the most of DALL·E 2’s abilities, Guy Parsons, creator of the DALL·Ery GALL·Ery, has assembled just the thing for you!

A bonafide 82-page guide on how to get started with prompting DALL·E 2:

In general, the more specific the request, the better. State whether you want a close-up image or a particular angle. Use adjectives that don’t have multiple meanings. Provide enough context for the system to gauge the aesthetic feel.

Parsons points out that even the creators of DALL·E 2 don’t fully know what the system is capable of, and experimentation through trial and error seems to be the best way to find out.

This detailed community blog post on LessWrong provides more information on DALL·E 2’s current strengths and weaknesses.

Over the next few weeks, the brilliant minds behind DALL·E 2 will invite one million people off their beta waitlist to try their AI image-generation tool.

OpenAI’s DALL·E Now Available in Beta

Craiyon (Formerly DALL·E Mini): Free for Personal Use

Created by Boris Dayma, a Houston-based machine learning engineer, Craiyon is open to the public and runs ads to recoup costs for the servers that power the AI system.

The burgeoning popularity of DALL·E 2’s resident protégé has meant that the Craiyon website regularly crashes due to receiving such high amounts of traffic.

Since the team behind Craiyon isn’t connected to Open AI, the latter asked for a name change in order to avoid any confusion.

Craiyon’s creations range from the strange (like people vacuuming in a forest) to the sophisticated (like French Renaissance imitations).

From the Twitter account of digital marketing apprentice Matt Laming, a 19-year-old from the United Kingdom
Frame-worthy, you reckon? (Image credit: Boris Dayma/Craiyon)

Images generated through Craiyon can be used for personal use “whether you want to share them with your friends or print on a T-shirt.”

The Terms of Use describe the conditions around commercial use in more detail:

  • Use of Craiyon-generated images for academic, research, educational, or entertainment purposes is free with proper attribution.
  • If you’re an individual or company that makes less than one million USD in gross annual revenue, you can use images for financial gain with one caveat: Craiyon stakes a 20% royalty claim on any revenue from blockchain transactions, such as selling NFTs.
  • If you’re an individual or entity that makes more than one million USD in gross annual revenue, use of Craiyon’s capabilities is subject to the terms of a separate commercial licensing agreement, for which potential prospects are asked to get in touch.

Midjourney: Join the Discord Channel

Last week, the makers behind the generative AI art tool Midjourney opened their closed beta doors, allowing anyone to join their Discord channel and create stylistic renderings using 25 free queries before being prompted to buy a subscription.

The beta is currently operating entirely through Discord with users typing prompts directly into the chat interface and receiving messages from a bot that shows their image renderings in real time.

According to VICE writer Janus Rose,

Unlike DALL-E, Midjourney seems to particularly excel at creating environments, especially fantasy and dystopian sci-fi scenes with dramatic lighting that look like rendered concept art from a video game.

Images generated by Midjourney with the prompts “Interior of a spaceship filled with lush plants” and “Photorealistic minions inside the Windows XP background”

The cheapest subscription plan allows access to 200 images for $10/month, while the premium tier allows for unlimited generations at $30/month. The terms of service stipulate:

Instead of MJ having ownership over images and licensing back to creators, we’re going to try flipping it around. The creators now have ownership, but grant MJ a unlimited license. To be clear, one of the many things we use the unlimited license for is to continue to allow the remixing and openness the community enjoys.

Like Craiyon, Midjourney originally had a stipulation concerning the minting of NFTs (requiring users to pay a 20% royalty on any revenue over $20,000 per month generated from blockchain technologies), but the creators have since removed this condition in favor of a more permissive terms of service.

Other alternatives like Wombo Art, NightCafé Studio, and StarryAI offer additional — albeit simpler — tools to generate dreamy, fantasy-style images though not with the same precision as the systems described above.

Meta (formerly Facebook) is currently experimenting with their own exploratory research concept called Make-A-Scene, which aims to exemplify AI’s potential to empower anyone to bring their imagination to life:

Outputs from Make-A-Scene

AI art generation has certainly come a long way since the days of DeepDream. Whether or not these systems will succeed in counteracting racial and sexist stereotypes remains to be seen.

Have you used AI-generated images in your work before? Share your thoughts!


Trending AI/ML Article Identified & Digested via Granola by Ramsey Elbasheer; a Machine-Driven RSS Bot

%d bloggers like this: