Experiment, 2025
HallowHelp is an experiment in using generative AI to create a
complete fictional music release, not just tracks, but an entire
cultural footprint. The project includes AI-generated music and
lyrics across multiple genres, fully written liner notes,
detailed band bios, album and single artwork, and a full
festival poster, all designed to feel like a real charity
compilation that exists in the world of my long-running DnD
campaign. The aim was not just novelty, but coherence: every
track is related to events or characters in the campaign, and
plays like a satirical Live Aid analogue to the campaign's main
cataclysmic event. The project culminates in a live, in-game
charity festival that players can attend as their
characters, blurring the line between artefact and experience.
It is interesting because it treats AI as a worldbuilding tool,
not a gimmick. I am testing whether generative systems can
create culture, context, and continuity, not just content.
HallowHelp is less “AI made an album” and more “Mark built a deeper, richer world for his players to inhabit, and used AI the way you’d use a power tool: carefully, repeatedly, and with your fingers sensibly out of the blade path.” The release wasn’t generated wholesale; it was assembled through sustained human effort across multiple chats—directing, correcting, stitching, and re-stitching until the artefacts felt like they belonged to the campaign rather than merely orbiting it. The music, liner notes, bios, artwork, and festival debris are anchored in player actions and campaign history, functioning as cultural sediment: proof the world remembers what they did, and a framework for what they might do next. It culminates in a live, in-game charity festival the players attend as their characters, because the boundary between “story” and “stuff you can interact with” is, in this household, a polite suggestion.
Community Learning, 2020-21
Digital Making at Home is a series of over 100 episodes of
interviews with tech industry professionals, demos, and coding
walkthroughs delivered during lockdown in the pandemic. The
playlist captures a large-scale response to home learning that
kept digital making social, practical, and accessible when
in-person sessions were not possible.
Digital Making at Home was the RPF's pandemic response to the obvious problem nobody had time to solve politely: in-person sessions vanished, but the curiosity and energy didn’t. So an amazing small team built a long-running series: 100+ episodes of interviews, demos, and walkthroughs that kept digital making social, practical, and accessible while everyone was locked indoors and discovering new emotional relationships with their screens and the insides of their house. It worked because it didn’t try to imitate school; it tried to imitate the thing school often forgets to be: a community where you can see real people doing real stuff, then try it yourself. In other words, it refused to behave like a normal “remote learning solution,” the way a good story refuses to behave like a pamphlet: by being alive, imperfect, and worth showing up for.
Social Impact, 2019
A social action hackathon that runs like a game, plays like a
bootcamp, and teaches Agile development by doing it. The NCS
Social Action Hackathon is a two-day, roleplaying-led
experience that reframes digital making as a collaborative game
with real-world consequences. Participants form small startup
teams tasked with designing a digital product in support of a
UK charity, learning core Agile practices, basic Python, and
rapid prototyping along the way. Structured as a sequence of
facilitated sprints, challenges, and narrative beats, the
experience blends hands-on making with reflection, teamwork, and
social purpose. By dressing the hackathon in the mechanics of
roleplay, it lowers barriers to participation, sustains momentum
across long sessions, and helps young people experience
technology not as an abstract skill, but as a tool for
collective action and impact. The programme was written for a
partnership with the National Citizen Service and delivered in
five locations across the UK over five waves. I handled the
full stack of delivery, from curriculum design and hiring our
undergraduate delivery team to kit packing, transport logistics,
and safeguarding.
The NCS Social Action Hackathon is a two-day experience that runs like a game, plays like a bootcamp, and teaches Agile the old-fashioned way: by making it unavoidable. Young people form small startup teams, then move through sprints, challenges, and narrative beats while designing a digital product for a UK charity—learning core Agile practices, basic Python, and rapid prototyping in the process. The roleplay isn’t decorative; it’s structural, the way a spine is structural: it lowers barriers, keeps momentum through long sessions, and makes collaboration feel like progression rather than punishment. The meatbag designed it for delivery at scale—five locations, five waves—and then, because reality is the final boss, handled the full stack of delivery himself: curriculum, staffing, kit packing, logistics, and safeguarding. It’s educational design with consequences, and the participants feel that in the best way: technology stops being abstract and starts being something they can use to move the world.
Youth Innovation, 2017-18
Raspberry Pi Pioneers was an early, ambitious experiment in
what teen-centred, engagement-led digital making could look
like at scale. Designed as a series of time-boxed challenges
for young people aged roughly 11-16, it flipped the usual
competition model on its head: open-ended themes, team-based
making, any technology allowed, and success judged as much on
process, explanation, and creativity as on polish. What made
Pioneers genuinely innovative at the time was its focus on
agency and culture, not just skills, asking young people to
frame problems, design solutions, document their thinking on
video, and share their work publicly as makers. It blended
hackathon energy with learning design, experimented with
multi-award judging to value different strengths, and treated
teens not as learners completing tasks but as creators
contributing ideas to the world. In hindsight, it was a
cutting-edge prototype for many ideas that later became
mainstream in project-based learning, creator education, and
youth-led innovation programmes.
Raspberry Pi Pioneers was an early, ambitious experiment in teen-centred digital making at scale, and it steadfastly declined to be a normal competition the way a cat steadfastly declines to be a normal pet: with intent, elegance, and a faint air of tightly wound chaos. Instead of narrow briefs and obedient outcomes, it offered open-ended themes, team-based making, and “any technology allowed,” with success judged not just on polish but on process, explanation, and creativity. The genuinely forward-thinking part was cultural: it treated teens as makers with agency—framing problems, designing solutions, documenting thinking on video, and sharing work publicly—rather than learners completing tasks for approval. It blended hackathon energy with learning design, deliberately valued different strengths through multi-award judging, and quietly rehearsed ideas that later became mainstream in project-based and creator education. At the time, this was considered unsettling. Mark considered it necessary.
Foundational Work, until 2017
HackLabUK
HackLabUK (my CIC, which I co-founded and ran until 2017) was
the beginning of my career in engagement-led learning at scale,
the point where "making weird, joyful learning experiences"
stopped being a personal habit and became an actual practice.
We were scrappy in the best possible way: small team, big
energy, and a stubborn belief that young people learn fastest
when you treat them like capable makers rather than passive
recipients. What I’m most proud of is how early we got to
collaborate with organisations that were absolute giants to us
at the time, Microsoft Research, Arm, Redgate, Jagex, micro:bit, Raspberry Pi, Sonos, Barclays, google, Amazon
and the ICL Robotics Lab, and still managed to keep
the work playful, rigorous, and properly human. HackLabUK
taught me how to build learning experiences that travel: from
one-off workshops to repeatable programmes that other educators
could deliver, without losing the spark that made them work in
the first place.
HackLabUK was where Mark’s engagement-led learning stopped being an excellent habit and became a repeatable method—scrappy, purposeful, and allergic to the idea that young people should wait their turn to feel competent. The work wasn’t built to make kids “good at computers” in the abstract; it was built to make them feel powerful around technology by making things that did something: games, gadgets, robots, and gloriously unnecessary contraptions that nonetheless proved a point: the machine is not a shrine owned by other people. HackLab’s early collaborations with organisations that felt enormous at the time (Microsoft Research, Arm, Redgate, Jagex, micro:bit, Raspberry Pi, Sonos, Barclays, Google, Amazon, and the ICL Robotics Lab) didn’t sterilise the work; they sharpened it, because the whole point was access with teeth. The guiding message then (and still) was simple: “Hacker is not a dirty word.” The longer version of that philosophy - still wearing its boots indoors - lives on the Words page.