So, there is a mouse trapped in the living room of me and my partner’s flat. And until the agency sorts our Rentakil to come this week, we’re restricted to hanging out exclusively in the bedroom, kitchen and bathroom. So last night, we sat on our bed with our homemade udon noodle soup and instead of watching a movie or talking about our day we decided to browse the internet’s AI potential.
We each took it in turns to create visual works of art with the help of an AI generator. My favourite was ‘woman in a field gazing at the moon’ and his was ‘black boy swimming in abstract blue’.
When it comes to our stance on the use of AI generators to produce creative work, we must consider the pros and cons. For example, we wrote a piece yesterday on a graphic designer’s work being plagiarised, openly, by a potential client. The client inputted examples of the designer’s work into an AI generator to create artwork for the album instead of paying, crediting or commissioning the artist.
On the other hand, however, “constantly having to come up with new content can be a struggle and a very serious source of burnout for creators” Lia Haberman, adjunct marketing professor at UCLA Extension has brilliantly acknowledged, adding that AI could help.
In fact, last night I posted a TikTok where I used AI to generate some lyrics for me – and they were pretty crap. The ideal stance for now seems to be one that utilises AI generation for support rather than pure creation.
Ollie Forsyth, global community manager at VC fund Antler, put together a landscape of generative AI startups with his team.
“I think initially we listed about 160 platforms,” he said. “Now there are roughly 200, but it’s actually so much bigger than we initially thought.”
Some of these startups will be more helpful than others for creators to solve the challenges they face on a day-to-day basis, Forsyth said.
For example, text-based AI can support creators with the scriptwriting for videos, and with copywriting for social-media captions and website posts. Video-based AI could help with different video-editing pain points, like cutting up short-form videos from a longer piece of content, or with dubbing existing videos in a variety of languages. And gaming AI startups could allow creators to make their own games or try roleplaying experiences.
Anyway, here is a list of AI generators that my partner and I researched last night and found interesting.
That’s right, Canva has hopped on the AI generator train. The user can add a written prompt into the AI platform, and the platform generates copy for things like website posts, marketing ads, and social-media captions, in seconds. Like regular Canva, the AI version of the platform can be helpful to generate content ideas, and provide creators with support in coming up with creative social captions or blog posts.
If you haven’t heard of this one yet where have you been? Apparently its even helping dating app users generate pickup lines! In more serious situations, ChatGPT can be great at supporting creators with writing scripts or getting past the writer’s block for content, supplementing the work of scriptwriters.
The one coming for my job! Just kidding, this is where the ‘support’ element needs to be reminded. For real, it can be your best friend in writing tweets, Instagram captions and YouTube descriptions. For full length pieces it will lack personality but can aid with SEO.
This one is super interesting as it uses AI to help translate content into different languages and speed up the dubbing process. It also recreates the speaker’s voice, in an effort to make it more similar to the original.
DALL-E 2
Okay, this is the one we stayed on for ages creating the woman in the field and the black boy in abstract blue. Basically, DALL-E 2 uses AI to generate images from descriptions in natural language. It can create all sorts of pictures and replicate styles, concepts, and attributes. YouTuber Zac Alsop described using DALL-E 2 to help visualise thumbnails for his YouTube videos, and how the process helped with making the thumbnails better.
PhotoRoom is a mobile app that uses AI to generate backgrounds for images based on a text prompt, using a licence from text-to-image model Stable Diffusion. It also allows users to easily retouch images by changing backgrounds or removing objects.
Honestly, I think this one is going to be a huge help for content creators. RunwayML helps facilitate and streamline video editing through AI, providing tools like text-to-image generation, image-to-image generation, background and object removal, motion tracking, and custom animation, among others. It also enables real-time, secure collaboration on projects. It is even being used currently by teams at The Late Show with Stephen Colbert, pop band Yacht, and YouTube superstar MrBeast.
