I've just come out of SXSW London 2026 with a head full of ideas and a slightly uncomfortable sense that we're asking the wrong questions about AI, innovation, and creativity.

A few moments stood out.

‍1. Start with people. Always.

One of the most unexpected sessions was about addressing human–elephant conflict using bees.

Yes, really. Bees.

At its core, it was a conservation story. But what stayed with me was less the solution (that beehive fences have been created to ward off elephants around crops) and more how the project was built. The team developed an AI-led system to track elephant movements near settlements and thus map out key priority areas/entry points for farmers, the design was created bottom up, with the local community shaping the approach.

Perspectives are intrinsic to how we approach problems and this really made me stop and take stock.

In our world, how can we genuinely stress-test ideas with the people we're trying to reach? Can we go beyond validation into something closer to co-creation?

Just a thought…

2. AI companion

A session called "Love (AI)ctually" on human relationships evolving with AI platforms was one of the most talked about, and for good reason.

A few stats landed hard:

• The average person is now spending 75 minutes a day on AI platforms

• 96% of teenagers are using AI for companionship

• There are already hundreds of reported cases of AI-induced psychosis

We’re not just seeing a technology shift, we’re seeing a behavioural one.

It is no surprise that there is a "loneliness economy" emerging, with AI increasingly filling emotional as well as functional gaps. Though this is interesting and worth being aware of, I think the bigger question is: is there a 'new' society emerging that sits outside of anything we have seen before? One where relationships are human-to-tech and tech-to-human without any human-to-human interface, and how will we evolve to stay in touch with this exponential societal change? Should we already be imagining what this society will look like, how it will feel, and who will lead?

Just a thought…

3. AI is a mirror and creativity needs to fight back

One of the common soundbites from several of the talks was that AI is a sycophant. AI learns by optimising for satisfaction.

It reflects what we want to hear. It reinforces, adapts, pleases.

In that sense, it works like a mirror.

And mirrors don't challenge us.

If AI is optimised for satisfaction, then there's a risk we start confusing feeling good with making progress.

Creatively, that's a problem.

The best ideas, the ones that truly move people or shift behaviour, are rarely the most immediately satisfying. They challenge, provoke, sometimes even make us uncomfortable and think further.

That's where human creativity still has a clear edge. Authenticity over optimisation.

In a world where more content is generated, faster than ever, that authentic differentiation only becomes more valuable.

Just a thought…

4. The uncomfortable truth: no one really knows where AI will land

If there was one consistent takeaway, it's this: No one has a definitive view on where AI is heading.

And that's OK.

It's probably similar to the internet in the early 2000s - clearly transformative, but impossible to fully predict. Even now, decades later, we're still seeing second- and third-order effects.

Which means the job is to form your own perspective rather than follow a consensus.

Just a thought…

‍

Credits: With thanks for inspiring me to the session speakers at SXSW: Holy Budge (Wild Innovation: Bees solving human-elephant conflict in Africa), James Muldoon & Sarah Turner (Love (AI)CTUALLY), Will Douglas Heaven (Five things you need to know about AI), Gloria Mark (Have we lost control of our Brains), Alberto Perlman (How a brand stays on top for 25 years), Craig Strachan & Dino Sofos & Jamie Laing & Miquita Oliver (Beyond podcasting: Building shows people belong to).

‍