The week at SXSW London | review and session notes (weeknotes 7 June 2026)

In theory, SXSW London’s fusion of tech, film and culture is right up my Venn diagram.

In reality, it hasn’t yet delivered on that potential.

But it’s getting better, some sessions I went to this year had useful takeaways (👇 see notes below), and I might pitch to speak at it next year rather than just moaning about it.

Logistics

My experience last year was so terrible, I got this year’s Platinum Pass for free. This edition ran far smoother, with a smaller campus, priority queues for Platinum holders, daily RSVP tokens to reserve sessions, and stewards clearing all auditoriums so each new session is fair game for all ticket types.

That’s annoying when you want to see consecutive sessions – having to loop around a venue and back through bag checks – but the fairer system is worth the pain.

Working while attending was impossible last year, and easy this year, thanks to a bigger and better-located Platinum lounge with copious power sockets, and free (sponsored) use of Fora co-working spaces. Nice one, Fora.

Vibes

It felt fairly buzzy, with lanyard-wearing festival crowds dominating Shoreditch. Folks were upbeat and friendly: I had lots of chats with interesting creative types in the queues and seats, which didn’t happen so much last year (we were all too angry about the queues). The app enables networking and meetups if that’s your thing, and I had a few chats through it.

It feels opulent. The money is visible everywhere, which is both good and bad. SXSW’s own branding, staging, staffing, AV and wayfinding are on point. Corporate sponsors and brand activations are everywhere, which I found both inspiring (as someone in an org that fundraises and runs festivals) and off-putting (for reasons below).

Conference slop

The worst thing about SXSW London is the patchy content.

There’s a lot of slop, with clickbait titles that promise much and deliver little, presented by people (or worse, waffled about by panels) shilling products under thinly disguised titles like “the future of X” and “building the next era of Y”.

Plus headline-chasing big name bookings (like Michelle Obama, Ant and Dec, Natalie Dormer, Alex Horne, George Osborne and Mike Skinner to name a few). And 1-in-3 session titles mentioning AI. Yawn.

It feels like a festival over-focused on profit and PR – prioritising style, sponsorship deals and aggressive marketing over substance – and that’s only going to work once or twice. Generously, perhaps that’s a necessary strategy to establish a foothold as a new big event in a saturated city. But it adds up to a feeling of inauthenticity, corporate greed, and an invasive takeover of Shoreditch rather than springing organically from it. IMHO it needs to shift more focus to curating real value with quality content if it hopes to grow a loyal audience.

What I saw

I planned to attend all week but, after watching several surface-skimming talks on Monday, I culled the rest of the week’s schedule to focus on BFI work. I went back on Thursday, and did a half day of BFI work from cafes and Fora, and a half day of conference talks.

Here are bulletpoints on all the sessions I went to, with star ratings out of 5 (use those if you want to skip to the better stuff).

⭐ Love (AI)ctually. Discussion panel on AI companions and their impact on human connection, with Sarah Turner from Feeld.co and prof James Muldon

[A random pick because a work call overran and I missed a more relevant session]
We all now trust AI chat with our intimate secrets and are all susceptible to feeling like we’re in a relationship with them [are we?!]
Research by Feeld (AI dating app) shows GenX are the loneliest generation, globally, and romantic AI companions are meeting a real need for them.
The “epidemic of loneliness” has created a “loneliness economy” that is now a fast growing emerging industry. [Jesus Christ]
[I found this so depressing I left. The Internet I fell in love with promised to connect us, collapse distances, tear down institutional walls and stick it to the man, but instead it’s driven us apart into polarised camps and social anxiety while the man lined his pockets. And now the man is exploiting that to sell us addictive apps that erode our ability to form human relationships even further. Er, no thanks!]

⭐⭐ Five Things You Need to Know About AI. Will Douglas Heaven, senior editor of MIT Technology Review

You needn’t have come to this talk. With one AI prompt you could have generated your own talk, infographic, AI podcast or song on this topic. A lot of confusion about AI in the news: good and bad impacts. Here’s the truth of what you need to know…
Thing 1. The jobpocalypse is exaggerated, but jobs are definitely changing. Coding agents are now heavily used, changing software dev. Multi-agent systems could do for office work what Henry Ford did for assembly lines in theory. But no evidence it is yet impacting jobs in the economy. All speculation. Impact likely to come slower than predictions. Use tools as much as poss and experiment but don’t despair, yet.
Thing 2. AI is getting scary for real now. Deep fakes, abuse, political misinformation, cyber threats. AI psychosis, excess trust in its advice. Romantic companions and intimacy serve a purpose but risks harmful advice, dependency and negative impact on human connections. Military use is scary too, being used to select targets.
Thing 3. A lot of people really hate AI right now. People go nuts about the use of it in film and TV. Activism against data centres. Copyright court cases and activism.
Thing 4: AI for science is a very big deal. Brainstorming, literature reviews, actual discoveries coming through in medicine. Maths unsolved problems being solved.
Thing 5: AI is everywhere all at once. Inescapable. It only came out 3.5 years ago. Not even its makers know how it works. Race to the top or the bottom? Unclear. A sense of inevitability. But it’s still just a technology, just like other waves of change, will be a marathon not a sprint. A lot of AI talks this week but be sceptical of anyone who claims to know what is really coming.

⭐⭐ The Reinvention Window: How Orgs are Embedding AI at scale. Marc Warner Accenture/Faculty applied AI interviewed by. Kai Nicol-Schwarz CNBC.

AI requires you to reinvent your organisation to get the value from it.
Through all of economic history, with every new general purpose technology, its value can’t be realised until the process also changes with it. You can’t just take a set of tasks out with extra software and leave everything else untouched.
Without clear focus, incentivising AI adoption can create high cost, consuming tokens
Many ways to screw it up. Leadership is the key one. Trying to change a bureaucracy is hard – by definition any org has curated processes that are designed to stay the same. Determination and understanding of leadership to disrupt those is key.
Thinking of AI as a technology looking for a problem is entirely wrong. Start with problems and user needs. Some of those can be helped with AI. Focus there.
Leaders should ALL now be playing with building things in Claude code.
Word processing didn’t kill writing. Excel didn’t kill accounting. When technology creates exponential productivity gains, consumption increases rather than decreases.

⭐ Venture Philanthropy: Extending Reach Beyond a Charitable Gift. Frank Rimalovski from NYU Entrepreneurial Institute, James Kofi Annan from Challenging Heights.

[I thought this would be useful from a BFI fundraising context, but it wasn’t. James is incredibly inspiring as a survivor of trafficking who escaped, got himself an education, now helping others escape the same. But the discussion was not transferable].

⭐⭐⭐ How we Need to Change Organisational Change in the AI Era. By Andrew McAfee, research scientist at MIT Sloan School of Management and co-founder of Workhelix.

Definition of AI led transformation: “a complicated, tech-heavy undertaking in an environment of high uncertainty and rapid change”
That is what Silicon Valley does for a living, with a proven blueprint for success (the most valuable companies in the world all clustered in the west coast of US)
Which means all companies need to adopt the methods and practices of the software industry now (and this is necessary for the world: it’s not good that so much power and wealth is being accumulated in the west coast US).
William Gibson quote: the future is already here, it’s just not evenly distributed.
So how do the silicon valley geeks do it? These 3 principles always:
1. PLAN A LITTLE AND ITERATE A LOT. Legacy companies have this totally backwards. Waterfall sucks. Clay Shirky quote: “the waterfall method amounts to a pledge by all parties not to learn anything while doing the actual work”. 75% failure rate. Agile manifesto 2000 changed everything, legacy companies need to catch up. Forget about adopting AI if you haven’t adopted agile as your core OS yet.
2. ENCOURAGE LEAD-USER, BOTTOM-UP INNOVATION. The modern AI toolkit gives you superpowers, you can build and scale without coding. New mantra for your teams: “you can just do things”. Let them play, scale what works.
3. STOP FLYING BLIND. AI is like early aviation. Experiments and failure. Try stuff. But you need good information on how your journey is going. Instrument your AI transformation and make them more observable”. Example metrics: average prompt quality per user. Average task complexity (are we using it for more complex things).
[Guess what… Workhelix is the solution, Andrew was a shill all along].

⭐⭐ Memes are Money: Rewiring the Fandom Economy with Gen Alpha. With Hew Moran, content creator, and Nick Walters from SuperAwesome youth audience agency.

Marketing without memes: make content > distribute it > market it > get attention > make money > sell ads > build sales team
Memes collapse this to: get attention > make money
“Steal a Brainrot” is bigger on Roblox than Travis Scott. Fortnite has Italian brainrot characters and a six-seven emote gesture. YouTube searches for brainrot massively outstrip searches for major movie releases like Zootopia. Toys and plushies of these meme characters selling in shops. Memes are worth millions in clicks and ad revenue, if you know how to jump on them and build them into your content/product.
Example from Hew Moran of a video jumping on a trend, that got 50m views. [Ugh this stuff is so vapid and creates no value, just clickbait for whatever’s hot]
Minecraft Movie chicken jockey moment. Organic. Brands can’t easily predict or replicate this success but can try; and be ready to exploit when it happens.
Dogecoin is worth more than Sainsburys. Billionaire Musk so focused on memes he took a sink into Twitter HQ just so he could tweet “let that sink in”. White House propaganda using memes from films to promote the Iran war. Proves this stuff is powerful to get and hold attention, for good or ill …
In conclusion, our advice is: find ways to use memes to drive traffic to get sales. Use them in your marketing well, and you get massive views and sales. Talk to our companies for how. [Boo!]

⭐⭐⭐ The Next Frontier: Audience Building in the Platform Era. Panel discussion with Ant and Dec, with Ed Sleeman and Nadia Afiari from Mitre (their production co).

The media landscape changed massively during A&D’s career. Linear TV reduced, digital, social, fragmented audience. New challenges but also more opportunities to create and reach audiences.
A&D always been curious about mechanics of the business since Byker Grove, asking Qs on set. Passion for good content, and connection with audience.
Now 20M followers across social platforms. Kings of live TV but still felt necessary to build social profile and direct connection to audience. BGT or I’m a Celeb has some interactivity but format and screen are barriers to audience shaping the content.
Created Jungle Club as a show on social media, direct chat, e.g. got audience input to how to deal with contraband issue, took that audience viewpoint into the production meeting room, helped inform the outcome.
Australia feels a long way from home and disconnected from the audience other than votes. Then Twitter enabled them to get real time input and reply in the ad breaks. Got excited about using it for engagement. Dual screen audience behaviour. They as presenters want to be doing the same back, live: is that triple screening?!
Had various incarnations of their own production entity. Ant and Dec Productions won the commission for SMTV and Saturday Night Takeaway. Then Galaga productions, various shows. Now Mitre studios. Small, underdog to Freemantle and Banijay to compete for a finite number of TV contracts, but also focusing on pillars of social platforms and branded content, d2c. Small. Agile and nimble. Independent.
Ant and Dec wrote the company manifesto, Ed Sleeman implemented it. Consolidated Ant and Dec archive IP and repackaged for fan market delivered through YT and insta and FB and TikTok under the brand “Belta box”. New original podcast Hanging out with A&D, and Taste Mates show discussing food. New IP show formats launched by A&D then expanded to use other talent and forms.
Podcasting is a perfect form for creativity, with total freedom to pivot, and direct connection to the audience. A fun, flexible form. Asynchronous: people come up to them in the street talking about episodes recorded ages ago, having just listened.
The A&D formula is stick firmly to core values: entertaining content, audience engagement, family friendly (bring fam together to watch), heart and humour.

⭐⭐⭐⭐ Design Stage Welcome Address by Sabba Keynejad, co-founder of VEED (video creation platform).

Set myself a challenge to not say “AI” in this talk, you’ve had enough of that.
Design, don’t tinker. Making things a little bit better incrementally is easy. Creating something different is hard.
True design means reinventing things from the ground up and challenging previous assumptions. Remaking. Reinvention. New tech means we all have to do it now.
Think of Netflix, a mail order DVD rental business. When the team decided to build an online VOD platform, their users weren’t asking for that; they were probably asking for features like better late fees process or more catalogue. Huge bet .. paid off. Think also of Dyson. Reinvented vacuums, driers and fans from scratch.
As designers, we need to understand any new tech deeply and not bet against it. Tech is material for us to design with, and that changes what we design.
At VEED, our mission has stayed constant: make video creation accessible to anyone. But tech changes mean the solution changes.
Our customers tell us they want faster, better editing. Been marginally improving this daily. But we need to break away from that vector and reinvent.
Been tough to do a major pivot. It’s hard to get teams pointing in a different direction. People disagreed with the pivot, some chose to leave. But courage of conviction.
About to launch V2, product is utterly changed, exponentially easier and faster editing via simple prompts.
Technology is the medium. Solving the user problem is the craft.

⭐⭐⭐⭐⭐ Building Products That Stick by Luke Des Cotes, CEO of Metalab.

Metalab has been involved in UI design and software dev for products used by billions daily. Slack, Headspace, Zoom, many other apps on your phone.
Learned a lot of key tenets. They are: Build a brand that’s felt throughout. Create systems that drive trust. Help users do the core job. Inject moments of pause and delight. Make complexity feel simple. Sweat the details and finer interactions.
Most teams know the tenets above, but still can’t ship products that stick.
Here’s how to evolve beyond and make apps that cut though.
Theme 1: the best product teams build over and over again. Calibrate your internal meter for greatness. Keep swinging and missing until you knock it out the park. 10k hours Beatles adage. Same for shipping products. 80% is easy it’s the last 20% that makes it great.
Theme 2: the best product teams don’t just ask, they understand. Get to know users and pain points. Talk to customers EVERY WEEK. Get to know them deeper. Build test learn. Not in a loop but all 3 at once. Build something completely different than what you thought you were going to build. Beware users telling you what they think you want to hear, or think they want but won’t use. Like a child wanting something for Xmas then not using it. Dig deeper to find out their real needs.
Theme 3: best teams go one level up. Find the obvious expected solution but then go further and do better. Focus on the user’s job to be done, not the product. Push beyond rivals – examples real time tracking of delivery cars; make it multiplayer; reshape the whole experience and industry. Clayton M Christensen: “New products succeed not because of the features and functionality they offer, but because of the experiences they enable”.
Theme 4: know when to kill your darlings. When it’s not working, stop. Eg Stuart Butterfield of Slack: “For a lot of people it feels better to just keep going until your idea dies of suffocation, than to admit you were wrong and pivot”.
Theme 5: very best teams celebrate the hold-outs. The naysayers. Teams that become iconic usually have a figurehead or leader that holds firm and never gives up. Steve Jobs’ maniacal focus on simplicity. David Holes at Midjourney: “if I could have no design in my product I would”. Without a charismatic figurehead holding the line, how do you create enough tension in your teams that creates radical creativity? A team’s pull towards average and agreement is strong. How do you go somewhere less safe? Listen to the holdouts in the team insisting on raising the bar.
To wrap up: culture contrast. The current zeitgeist and obsession is to go faster, use AI to accelerate. My advice is to do things that inherently slow down the process. Going backwards, killing ideas, talking to users more. Speed is being celebrated with AI accelerated coding but friction and thoughtfulness is key to great products.
Tech is obsessing over killer apps to replace Photoshop and Canva etc. mimicking greatness not creating greatness. Don’t kill effort and discomfort and cost and risk, those are the elements that create good products. Ambitious teams willing to endure more pain are the ones that get to great.

⭐⭐⭐⭐ From Attention to Presence: When Brands Start Behaving, by George Eid, founder of AREA17 consultancy.

Brands frequently squander customer trust through mismatch of what they say in their marketing and do in the design of their systems. Example: billing you for underutilised licenses. Emails with “we care about you” messages from no-reply email accounts. That’s incoherence.
Character is built through action. What you do, not what you say. Showing up isn’t being seen and heard through campaigns, it’s being there when it matters.
That’s brand PRESENCE: being genuinely, consequently there when someone needs you. Opposite of marketing, sales, billing.
The support team handles frustration. Who is handling the gap – what a brand stands for vs what actually happens when the customer needs something?
Closing that gap at scale is difficult. Org leaders may genuinely care, but the gap between what the brand stands for and what happens exists at every touchpoint.
The thing that can close it might be AI
Quakers built loyalty through conduct. Medieval guild marks and seals were a proof of value and quality.
As orgs got more complex this value exchange got lost. Focus on profit. Modern orgs reached for a substitute: brand and campaigns. Corporate identity. This is all veneer, presentation, not action.
The early Internet made this behaviour gap visible. Every interaction, every gap between what a brand does and says became felt and seen. A new transparency forced orgs to better align what they said in marketing and did in their systems.
Social media subsequently set that back. Likes instead of trust. Appearance of presence, performative posting, crafting viral messages to interrupt the scroll.
Then came enshittification. Algos control who sees your content, optimising for clicks and money not trust, loyalty and value. The system rewards noise so people make more of it. Most attention wins in culture, commerce and politics. That’s the world we have built. Products getting worse, CSR and EDI in retreat. Extracting more value and returning less in exchange. Brands rolling back purpose and focusing on profit.
Now with AI, the power of marketing is diminished. The customer is asking Qs and getting a single interpreted answer. Already weighed against alternatives. Conclusion formed by the machine. Thinking externalised from the human brain. Marketers cannot influence this external brain and plant emotive claims about their product.
Broadcast rewarded consistency. Social media rewarded attention. AI rewards evidence. Brands used to shape perception, but n the AI era, they will need to establish evidence of value the machine can read coherently. Presence. Coherence.
The one thing left for brands is the thing that mattered all along – behaviour. Accumulated evidence of what the brand stands for through all its actions.
Internet gave a look behind the curtain. AI is removing the curtain altogether.
We may have seen a retreat from the optimism of the early internet (empathy, connection and transparency). But we were right about something fundamental all along. Systems reveal character through behaviour.
Orgs have been able to create distance between what they said and did. That’s becoming harder again because the systems are getting better at evaluating quality.
You have one job: design for being there the moment someone needs something.

⭐ To Collaborate is Not to Siphon. Panel with Maisie Willoughby Timberland CEO. Arthur Yates Saatchi Yates, Slawn (artist)

[This was fluffy art chit-chat, the panel had no clear points prepared, I walked out]

⭐⭐ Cool is Fuel: Why Culture Wins in the Age of AI, by Manon Dave.

[He should have just called this ”I am cool” – it was 99% bragging about his own life and achievements, with 1% useful messages. Tbf he is cool though]
Reflections on AI in the past year from a generalist pov…optimistic! Play play learn.
Open AI bought Johnnie Ives cos you need culture and design and cool
Smart: artificial, automated, efficient, trained. Super: authentic, imaginative, empathetic, inspired
Tech has evolved but needs human taste and empathy to make things of value
Examples: Mercedes sound drive he created with will.i.am. Refugee Olympics team
New product about to launch from his work at BBC R&D: a second screen overlay (like a modern day Ceefax!) called BBC Signals, with real time data insights and interactivity for the audience. [Interesting – I wonder if this has a basis in user research and will actually get used. Red button stuff has never caught on yet…]
Kids he played with on his London housing estate as a kid are now Skepta and Mustafa Suleyman, CEO of Microsoft AI. Culture and cool gets you far..

In conclusion…

There’s some decent takeaways and inspiration from that lot, and my 2 x half days at SXSW London (plus some very productive working time in the change of scene in Shoreditch spaces) feels like time well spent. Hopefully you’ve found something useful or inspiring here too.

So…would I go again next year? Probably, but only if the content is beter. And maybe I’ll also pitch to present. I’ll get my thinking cap on for a talk that offers some real substance and value. While also shilling for BFI at the end, obviously.