A panel exploring the intersection of AI and creativity, and how use of LLMs can both challenge and inspire creative minds.

Panel

• Babusi Nyoni
• Pau Garcia
• Elizabeth Valleau
• Moderator: Monique van Dusseldorp
  

---

These are live-blogged notes from a session at NEXT24 in Hamburg.

---

AI in creative action

Pau has been working with over 75s for the last year, looking at how they can use AI to reconstruct lost memories of people and communities. This work has an impact on history, and politics, but also on people’s own lives, particularly those with dementia. They worked in nursing homes all over the world, talking with residents documenting their witness with generative AI images.

There is something powerful in externalising something that you only hold in your head. There is a relief when you share it.

Part of the work is now capacity building, teaching people to use the tools so they can carry on working with people themselves.

Babusi worked on Vosho Fo’Sho, an app for improving your ability to Vosho — but the same underlying technology could be used for tremor detection and part of Parkinson’s diagnosis. And then used to transfer motion into a metaverse space.

Elizabeth: We’ve all experienced the heights of expectation and disappointment in delivery in tech. But as emerging tech comes together, we need to realise that a world of centralised data and infrastructure is no longer acceptable. Imagine Media was a way of rethinking and reimagining what media looks like for that future, a set of provocations.

When has AI proved most useful?

Back in 2016, Babusi learned how useful AI could be, though building an AI football commentator for Heineken.

Understanding how to put AI in a context that is meaningful to people is everything.

They’re using AI to cut down doctors’ triage times, for example. That was 2019, long pre-ChatGPT. He would never use a general model for that sort of contextual use. It’s risky with a specific model, but much more so with a general one.

There’s divergence innovation, says Pau where you’re just exploring ideas, and then converging work, where you pin it down into place. Many of our tools are convergence-based: taking ideas and making them into something. Generative AI is a divergence tool — it gives you things, but not always what you expected, and that can create new ideas.

Elizabeth:

We’re not at the dawn of AI, but the dawn of consumer-facing, Muggle AI.

Your average CEO now has the language to describe what they want with AI. You can train things that are safe, private, and which are incredibly nourished. If you think about how a bank uses your data, it can be quite overwhelming. But if you think of banking systems trained on discreet, personal data, that system could be a banking companion for your life. Nobody knows you as they do. There are human-first ways to use AI, driven by creativity but mainly just by people.

Now, AI seems to be the answer to everything, says Pau.

We even have clients asking if we can “put a bit of AI here”, as if it’s salt!

AI ethics and opportunities

Babusi is still trying to figure out the ethics around image generation on the ways that infringe on artists. He loves it in the context of memory recreation. But he also loves the scale and speed AI can bring to some tasks, and its ability to model relationships and their dynamics.

Pau points out that Generative AI is mainly a tool for solving language problems. But most of our problems are solved with language — we communicate to solve things.

A good tool is not defined by what it produces for us, but what it produces in us.

Elizabeth is passionate about the problems with the exclusion of the body from the arc of innovation. As soon as we started working on laptops and screens, the connection to our spirits and to nature was curtailed, we’ve become creatures curled around our laptops. AI could be in the process of untethering us from these drives that have trapped us indoors. Now, we’re a long way from the phone being something we’re frustrated by because there are easier ways of doing this, but we could get there.

Has AI taken anything away?

For Elizabeth, AI has taken away the power of individual knowledge and bias. It’s no longer just the loudest, most opinionated voices in the room.

Babusi has lost the joy of solving a problem by coding it purely because he turns to the lottery machine of an LLM instead — and ends up arguing it with it occasionally.