5 generative AI tools that caught creators' fancy in 2023

Adobe, the giant company in the creativity software space—well known for software like Photoshop and Illustrator–that’s also gobbling up competitors like Figma, released Firefly earlier this year. Image: Shutterestock

1. Adobe Firefly

If you head over to firefly.adobe.com, right there on the screen is a simple interactive option to see how you can play around with images. Just moving the cursor side to side on the main image on the page slides over one option or the other, depending on which way you move the cursor.

Adobe, the giant company in the creativity software space—well known for software like Photoshop and Illustrator–that’s also gobbling up competitors like Figma, released Firefly earlier this year. It’s now out of beta, barring newer features being added.

Just type natural language inputs into a ‘generate’ box that looks like the familiar search box, for what you’re looking for and Firefly will generate multiple image options and you can dive into one and start tweaking it, by adding more elements.

One way to do that is ‘generative fill’ through which you can select portions of an image and fill them with whatever you want, basically. Ask for an image of a sunrise and then add two moons if you like. Adobe is also attempting to make it clear that the images are AI-generated, by adding ‘generated by AI’ credentials to all such images.
 

2. Canva Magic Studio

Canva’s latest product for creators—read that as anyone who wants to create—is its Magic Studio, and within that Magic Design. And the company’s own description says it best: “With Magic Studio there’s no need to toggle between multiple AI tools or learn lots of different software—all the best of AI is at your fingertips.”

“Created for the 99 percent of the world without complex design skills, Magic Studio is jam-packed with easy-to-use AI-powered features across every part of Canva to help you work smarter,” the company says.

Again, at the heart of Magic Design is a natural language-based way to start something. By now we’re all used to asking ChatGPT to find us recipes or teach us the history of investing in forwards and options, or whatever. Similarly, ask Magic Studio to help you get started with say “please make a poster for my friend’s birthday” or if you’re in HR in an enterprise business, “plan a quick tour for our new intern” as one of the examples on Canva’s website itself suggests.

A big attraction with Magic Studio is, you can upload your own pictures and videos and ask it to combine them to reflect a narrative that you can just type in. You also create short videos based on simple text inputs. Head over to canva.dom/magic to check out all the features available.

Also read: Is Generative AI a disruptive paradigm-shifter? The questions to ask
 

3. YouTube Dream Screen

Speaking of videos, a list like this wouldn’t be complete without an update on what’s up at the biggest video platform there is. In September, YouTube announced it is working on a generative AI-based video maker and editor that it’s named Dream Screen. A release date isn’t known yet.

In the meantime, YouTube has also released YouTube Create, a right-in-the-phone tool to make short videos, especially, and also edit them with a fair number of features. This includes precision editing and trimming, automatic captioning, voiceover capabilities and access to a library of filters, effects, transitions and royalty-free music with beat-matching technology–currently available on Android only, and in some markets.

As to AI, Google plans several additional useful features, as listed in its September blog post on Dream Screen:
Next year, YouTube Studio will help creators tap generative AI to come up with video ideas and draft outlines to help creators brainstorm. These AI-powered insights will be personalised for each channel and based on what audiences are already watching on YouTube.

Also, next year, YouTube is planning an assistive search feature for music for creators, making it easier for them to find soundtracks. Simply type in a description of your content and AI will suggest the right music at the right price.

And then for automatic dubbing, Google is bringing Aloud, a product developed at the company’s incubator Area 120, into -YouTube, giving creators an AI-powered dubbing tool that will help creators open up their content to the world.
 

4. Descript Regenerate

Descript started out as a simple product with features like transcription. Today it’s a sophisticated platform that offers a host of video editing and podcasting features and several ways of correcting, improving, and enhancing your content.

One AI feature is Overdub, where you can train Descript with your own voice and get a surprisingly realistic AI-generated version of your voice. One way this could be useful, in a podcast, for example, is by typing a word or a sentence and having Descript generate the audio version in a speaker’s voice. It is useful in improving a question or replacing a mis-pronounced word and so on.

Another is Regenerate, where one can actually change the tone and intonation of a word or sentence to make the whole more consistent. For example, say in mid-sentence, a guest switches off and sounds a bit distracted, making the rest of the response lacklustre compared with how the guest started out.

In the old way, all one could do was to try and edit, so as to minimise the boring-sounding parts. With Regenerate, one can tell Descript’s AI to match the tone of the voice with the more enthusiastic parts, making the whole response more engaging to listen to.

Also read: Indian IT leaders are ready to embrace generative AI but also have concerns: Salesforce study
 

5. Runway AI Gen-2

Time magazine wrote 2023 might be the year of generative video, in a post describing Runway as the leader in this latest advancement in how generative AI is being applied in the creative space—text to video.

Runway, a New York startup founded in 2018 and backed by investors including Salesforce Ventures, Google and NVIDIA, was named to the Time100 most influential companies of 2023. The startup has raised some $236 million and offers AI-powered video-to-video (Gen-1) and text-to-video (Gen-2) products, which were also used by the team behind the Oscar-winning film Everything Everywhere All at Once.

Runway’s tech is also available as a smartphone app, RunwayML, and has millions of fans. The idea behind this mind-blowingly sophisticated piece of software is so simple that anyone can use the app. Just type to describe the video you want and tap generate. That’s it.

As part of its marketing and community building exercise, Runway has a creator partner programme, offering exclusive access to its tech to selected creators. It has also announced a video competition open to anyone who will run it for 48 hours starting October 21.

The competition, GEN:48, is an entirely online event taking place from 9 am EST (6.30 pm IST) on October 21 to 9 am EST on October 23, 2023. After the competition closes, a selection of finalist films will be put up on the website for a public vote that will decide the winning entry. The theme and creative requirements for all films will be revealed at the start of the competition, according to Runway.