Hello!welcome back Technology: AIa pop-up newsletter that introduces you to artificial intelligence, how it works, and how to use it.
In last week’s newsletter, I shared my golden tips for getting the most helpful answers from chatbots like ChatGPT, Bing, and Bard. Now that you’re familiar with the general principle of building relationships with AI—the more specific and detailed instructions you give, the better results you’ll get—let’s move into a slightly different realm.
Much of the hype and fear surrounding generative AI is about text. But there has also been rapid and remarkable development in systems that can generate images. In many cases, they share a similar structure to text-based generative AI, but they can also be weirder — and lend themselves to some very interesting creative pursuits.
Image generators are trained on billions of images, allowing them to create new works that were once exclusive to painters and other artists. Sometimes experts can’t tell the difference between AI-created images and actual photographs (aside from amusing creations, this situation has fueled dangerous misinformation campaigns). These tools are already changing the way creative professionals work.
Compared with products like ChatGPT, image generation AI tools are not as developed yet. They need to do a few extra hoops, and probably cost a little money. But if you’re interested in learning the ropes, now is the perfect time to get started.
Creators on social networks like TikTok and Instagram have been buzzing about it since Adobe added generative artificial intelligence to a beta version of its signature graphics software Photoshop last week.
I have quite a bit of experience with Photoshop. When I tested the new feature, called Generate Fill, I was impressed by the speed and ability of the AI to perform tasks that would have taken me at least an hour to complete on my own. In less than five minutes, I used the feature to remove objects, add objects, and swap backgrounds with just a few clicks.
(To try these tools out for yourself, first sign up for a free trial Adobe Creative Suite. Then, install the new Adobe Photoshop beta, which includes Generate Fill. )
After installing the Photoshop beta, import your photos and try this trick:
change background, click the Object Selection icon (it has an arrow pointing to a box), and under the Select menu click Invert to select the background. Next click the Generate Fill box and enter a hint—or leave it blank and let Photoshop give you a new background concept.
I edited a photo of my corgi Max using these steps. I typed “kennel” into the prompt and clicked generate to replace the background. This is before (left) and after.
The New York Times photo editors do not enhance or modify photos, nor do they use artificial intelligence to generate images. But my first thought after testing the generated fill was that photo editors who work in other contexts (like marketing) might soon be out of a job.when i share this Theorize with Adobe CTO Ely Greenfield, who says it might make photo editing more accessible, but he’s optimistic that humans will still be needed.
“I can take really nice pictures with it, but frankly, the pictures I take are still pretty boring,” he said. “When I look at what an artist has created, when you put it in their hands, their stuff is more interesting than what I’ve created because they know how to tell a story.”
I’ll admit, what I’ve done with generative stuff isn’t nearly as exciting as what other people are posting on social media. Lorenzo Green, who tweets about AI, Posted a collage of famous album covers, including Michael Jackson’s “Thriller” and Adele’s “21,” which were expanded with generative fill. It turned out to be very interesting.
(Note: If installing Photoshop is daunting, a faster way to test Adobe AI is to visit Adobe Firefly Website. From there, you can open the generate fill tool, upload an image and click the “add” tool to trace an object, such as a dog. Then click “Background” and type in something like “beach.” )
More Image Generators
Tools like DALL-E and Midjourney can create brand new images in seconds. They work like chatbots: you type in text prompts — the more specific the better.
To write high-quality prompts, start with the medium you want to emulate, followed by the theme and any additional details. For example, enter “photo of a cat in a sweater in a bright room” DALL-E prompt box would produce something like this:
Owned by Open AI, the makers of ChatGPT, DALL-E was one of the first widely available AI image generators that was easy for people to use. For $15, you’ll get 115 credits; one credit can be used to generate a set of four images.
Another popular image generator, Midjourney, is under development, so the user experience isn’t quite there yet. The service costs $10 per month, and entering prompts can be a little more complicated, as it requires joining a separate messaging app, Discord. Nonetheless, the project creates high-quality, realistic images.
To use it, Join Discord and Request an invitation to the Midjourney server. After joining the server, enter “/imagine” in the chat box, and a prompt will appear. I typed “/imagine a manga cover of a corgi in a ninja turtle costume” and produced a convincing set of images:
While typing basic requests is fine, some people have found that obscure prompts produce unexpected results (Beebom, a tech blog, has example list). At Columbia University, Lance Weiler is teaching students how to use AI (including Midjourney) to make art.
No matter which tool you use, remember that it is your responsibility to use this technology responsibly. Tech experts warn that image generators could increase the spread of deepfakes and misinformation. But these tools can also be used in positive and constructive ways, like making family photos look better and brainstorming artistic concepts.
Next week, I’ll share some tips on how to use AI to speed up office work, such as drafting talking points and generating presentation slides.
In case you were wondering, the delightfully insane image at the top of this newsletter was created by one person – The Illustrator charles de marais – not artificial intelligence