AI’s so hot right now…

…so what…

Now I assume you’ve all had a play with it, AI I mean. But have you actually used it and the outcome. There are multiple ways to have done this — be that a simple search and copywriting request via ChatGPT or the inevitable play with DALL-E 2 or Midjourney for visuals. I’d assume the answer is yes — but have you considered the implications.

What is the output, and can you own it… really… and can you then control it?

Limitations…

They say they’ve added safeguards to stop DALL-E 2 from creating images that are ‘violent, hate, or adult’ by removing explicit content from training data. But how can that be when AI is learning, combining, and applying to create another outcome.

Consider the idea of removing hate and violence in creativity and flip it. Take Banksy as an example. Without violence, where would his particularly brush of anti-establishmentarianism be, regulated perhaps? How many wars, oppression and other have inspired artistic or musical genres and movements (as well as positive moments of course!). Freedom of expression is creativity — creativity without limitation.

Copyright…

And then there’s copyright infringement. A recent challenge in the US meant that the copywrite of images for a graphic novel (created using Midjourney) were revoked, leaving just the text covered. By its very nature AI is trained on what exists and therefore it stands to reason that the output will be ‘very similar’ to what it has been trained on (at least right now, based on my limited knowledge). And yet, OpenAI says you can generate and then own them (with conditions).

And so — is the output really ownable and due to the very nature of AI training and safeguards — is their creativity limited. Arguably yes. BUT enough of the fearmongering.

Opportunities…

There are obvious opportunities with AI — whether you’re keen on the concept or not. Like many of us, you will adapt and you should.

So, what affect does this have on the creative and design industry — another layer of subjectivity? From our perspective, only if we let it — it’s a choice — never a prerequisite.

My challenge is in both ‘creativity’ and ‘worth’. Is having a machine develop imagery based on parameters creativity? Arguably yes. Does this mean creativity is cheaper? No. You still need to know what to tell it to do, refine it — but even then, do you own it?

For us these AI platforms are limited — only by the wielder’s creativity. They are a tool, another to help express an idea, sense check thought or provide another point of view. In our opinion, never the strategy or idea. We should embrace this as another weapon in an arsenal, another searchable aspect with guidance.

The opportunity is endless — with variations from Artbreeder that allows modification of landscapes, faces, painting and more — literally breeding images together to DALL-E 2 outputting realistic images based on written parameters, to Let’s Enhance, which allows image enhancement and more (although explicitly; as long as there are no 3rd party licenses). This is great if you want to increase the resolution of an image, or simply sense checking sustainable options — like a Google on steroids. It’s a complex world in itself!

But again, they are unlikely to deliver a unique brand positioning or creative / design idea — only an expression of it. They are creators of efficiencies — but first there must be creative thought.

Use it, but remember what it lacks…

From our perspective it’s a way to bring something to life to check if it will work. It’s a way to review an insight or options, collect more angles of data on a demographic — mindset, behaviour or others. BUT remember the limitations. It’s unlikely to be the end result and be mindful of who owns that end, as well as what might have been without it!

Don’t get me wrong — they’re incredible. Just also remember how incredible the human brain is and that it takes time and practice for brilliance.

#brand #aitools #branddesign #design #molasses