Generative AI: The now, and the future
The ‘next big thing’ in AI is equally exciting as it is unsettling. Generative AI, is the catch-all term for widely available tools that enable you to originate images, and short videos, by typing in instructions or guidance. Generative AI uses machine learning to manipulate, merge and create original images from your instructions.
Generative AI is part of the low code/no code future – easy to use and available to many. As well as stand alone tools such as Dall-E2, MidJourney and Stable Diffusion, there are multiple other suppliers springing up.
At Formative Content we’ve started experimenting with these technologies in our visuals creation. How can we use them to innovate client visuals and video, and add to the range of output we create?
A vibrant alternative to stock imagery
We all know those images of laughing perfect people in a meeting, or that man with the beard on his iPhone or his iPad. These stock images are ubiquitous, and are over-used on corporate marketing sites as they are non-copyright images and easy to access and cheap to use.
But, let’s face it, they’re dull, boring and sanitised. Imagine if you could simply and easily create vectors, images and designs that reflect the ‘mood’ of your article, your post or your presentation? That’s what Generative AI will allow you to do.
By typing in your key search words for generation of your images – and then refining them – it’s straightforward to create “original” images that more truly reflect the emotion and/or content of our posts and articles. They might not work every time at this stage, as they are only as good as the range and selection of ‘instruction’ images uploaded, but this will no doubt be a useful, and potentially cheaper and more impactful, option for stock alternatives.
Another potential use case is ‘mock-ups’ to show ideas or concepts at a really early stage. It can be a challenge to easily explain to clients or colleagues what we have ‘in mind’ with an idea for a visual image or video. One person’s ‘vision’ is another person’s gobbledygook (certainly it is when I’m trying to explain it!), and – in an agency world where time is money – sometimes you don’t have the luxury of time to “spend” on multiple execution options.
That’s where generative AI can come in hugely useful – if you’re looking to generate an image of a location, a background, a fixture, it’s possible to create multiple renders quickly and easily that can support your future design.
Immediate use cases
With some creative thinking, and an understanding of the tools available, the output you can create is unlimited. Whilst investigating tools to create header images for my upcoming LinkedIn newsletter (it launches SOON, since you ask!) – called The Funnel (see below for final header). Formative’s Head of Visuals developed the draft versions below using Dall-E. By typing in simple instructions around the word ‘funnel’, Alex created some vector-style images to work from. This is a learning process, and these are just simple first experiments – imagine what we can do as this technology becomes more embedded into our ongoing design work?
Alex is already using these tools in day to day work at Formative:
“AI had me feeling equally curious and cynical about what this could mean for designers and creatives. However, once you experiment and speak the AI language, you immediately realise the staggering potential. The smartest way to see immediate benefits is to hand over some of the tedious, time consuming tasks of mood boards and ideation style frames and allow creatives to focus on the big ideas, the storytelling and the design thinking – where our true value shines. It is impossible not to feel a touch anxious by the impact of AI capabilities, but I am now captivated by what these tools will mean for all creatives… and it’s now not the future.”
And needless to say, in a world dominated by video, it’s not just still images that will ultimately be on offer. Generative AI video is coming down the track, fast. Facebook is an early mover in the AI video space – in September it launched Make-A-Video. Whilst still at experimentation phase, it takes a text prompt and uses that to generate a simple short video; other suppliers like Runway are already in the space and offering solutions. We haven’t started using these, but no doubt it won’t be long before we start experimenting.
The opportunities – and issues – are immense
The issues are legion, and fall into some obvious categories. They include issues with copyright and image manipulation. To train AI you need input material, and in this instance it’s often images that already exist. But what about the originators of content that is manipulated? Not everyone is happy, as you can read in this piece from CNN.
Then there is violence, nudity and aggression issues, as well as the growing issue of misinformation and ‘fakery’ for a cause. The genie is out of the box, but is current legislation enough to deal with the obvious problems? The boss of Stable Diffusion believes a libertarian approach, devoid of corporate oversight, is the right one to deliver the most opportunity from the tech.
But alongside the concerns, there is such opportunity and excitement around this tech, which is already being used by artists, designers and creators worldwide.
Source: Jon Porter
If you want to read more about the issues thrown up by Generative AI, and an interview with the boss of Stability Diffusion, read this article from the New York Times.
Thanks to Mark Schaefer and those on the $RISE Discord community for help, info and inspiration for this post!
About the author: Gay Flashman is the CEO and founder of Formative Content. Connect with her on LinkedIn here.
5 insights from Davos for corporate communicators
The World Economic Forum adopted a newly minted word to frame the debate at its 2023 Annual Meeting in Davos...
Getting the most out of live events like Davos
Events are a goldmine for content creation, enabling companies to tell their stories and share key messages in real-time, while also...
Why being seen to be green is not enough: 5 communication lessons from COP27
Being seen to be green has become a crucial part of reputation management for almost every company around the world.
Generative AI: The now, and the future
The ‘next big thing’ in AI is equally exciting as it is unsettling.