Pitch · Original Series · Google DeepMind

Living
Proof.

The story of AI, told from the front line.

Pitching AI through motion is like trying to sell a laptop by making you read the manual. We too often frame it in sci-fi tropes or through dense science papers and code.

Enter Living Proof. A premium documentary series that brings DeepMind's breakthroughs — GraphCast, GNoME, AlphaFold — out of the lab and into the trenches of human experience.

We aren't filming algorithms. We're filming the exhausted meteorologist using GraphCast to outsmart a super typhoon. It's emotive, awe‑inspiring, and tangible. In this series, AI isn't the hero; it is the invisible scaffolding elevating human grit.

Cinematic photo of the ITER fusion reactor workshop, raw and unedited.
The same ITER workshop overlaid with the mint-green DeepMind Lens — structural wireframe and telemetry boxes projected onto the scene.
Raw Footage DeepMind Lens

⟵ Drag to reveal · Episode 02 visual reference

01 — Human Momentum

Filmed on the move.

Contributors filmed in the field, in the workshop, travelling through their environments. Interviews used primarily as fast-paced voiceover. No sterile talking-head interviews.

02 — Visual Scale & Mixed Media

Drone scale meets WhatsApp grit.

High-end cinematic drone footage of cavernous workshops, cut with raw mixed media — field photos, phone footage, WhatsApp notes — for intimate documentary-journalism grit.

03 — Balanced Optimism

Hopeful, never naive.

Anchored in active problems DeepMind is solving. We don't shy from challenges, costs, or skepticism — leaving audiences inspired and curious, rather than anxious.

The Visual Signature

Painting data on the real world.

Cutting away to a sterile, floating 3D graphic kills emotional momentum, takes us away from the people, and is expensive. So, we don't cut away.

Using simple 2D motion tracking, we project sleek, minimalist data directly onto our cinematic footage. The science lives natively in the human environment.

As our engineer walks the Swiss Tokamak workshop gangway, the 'DeepMind Lens' projects a structural wireframe onto the component above her, translating complex engineering algorithms into live, visible metrics right there in the room. Relatable visual metaphors break down the densest topics — sophisticated, and highly scalable for a lean production team.

Target audience · Channel strategy

Protect the rigour.
Expand the reach.

Architecting an ecosystem, not just a show.

DeepMind is founded on scientific rigour. If we dilute the science, we alienate our core audience. However, creators, business leaders, and people curious about AI are fascinated by its real-world impact — but they aren't reading technical whitepapers. We should translate the science to them to affect the cultural conversation. This is something Anthropic currently excels at.

How do we compete? We don't just launch a show; we architect an ecosystem. We deliver awe-inspiring human stories to the masses, while protecting the science with practical tutorials and leaning into pre-existing content across the wider Google video ecosystem (e.g., Google for Developers, the Google DeepMind podcast). This ecosystem means repeatable, format-specific pipelines that don't burn out the internal team.

Hero Strategy

Living Proof

FormatDocumentary Series · YouTube long-form
AudienceThe Curious Mass Market.
ContentPremium, emotive human storytelling.
GoalDemystify AI. Prove our models are tools for tangible human progress.
Hygiene Strategy

DeepMind Tutorials

FormatShort, tactical how-to videos
AudienceBuilders, developers, and professionals.
ContentSimple, tactical tutorials (e.g., "Building automated research workflows with Gemini").
GoalDrive hands-on adoption and practical utility.
Hub Strategy

Content Curation

FormatCurated sessions, vodcasts, talks
AudienceEngaged Technologists & Researchers (data scientists, ML engineers, academics).
ContentStrategic curation of event sessions, vodcasts, and deep-dive technical talks.
GoalEncourage further learning and guide them deeper into the Google ecosystem without creating a new production bottleneck.
Episodes · Talent · Longevity

The production
in action.

Proof of concept, character archetypes, and evergreen longevity.

Skip the celebrity host. Let the science do the talking. Our characters are the hosts. This keeps the show cost-effective and operationally scalable. We drive every episode with character duality:

Internal · Owns the "How"

The DeepMind Engineer

They translate complex models into relatable visual metaphors. The voice of the science.

External · Owns the "Why"

The Field Pioneer

The real-world practitioner. They provide the emotional weight, the human stakes, and the stakes-on-the-ground urgency.

Representation — behind & in front of the lens

Science and AI have a historical representation gap. If Living Proof is going to truly resonate with a global mass audience, the people telling the story must reflect that audience. This philosophy extends behind the camera. Hiring diverse local crews enables other voices in shaping the narrative of the future.

Sample slate · Season 01
Cinematic shot of a coastline beneath a high-rise structure, with the wave below — visual benchmark for Episode 1.
Episode 01

Outsmarting the Storm

GraphCast
Tech

GraphCast

Duality

A DeepMind researcher predicting cyclones on a laptop meets a meteorologist relying on that data to save thousands of families on the coast.

Look

Cinematic drone shots of coastlines overlaid with a bold, 2D-tracked pressure grid showing exactly how the AI models weather movements over the ocean.

Longevity

A format built for seasons.

Whilst campaigns expire, Living Proof is an evergreen format. The structure is standardised to ensure endless variety without creative ops needing to change.

Variety: We jump from nuclear workshops to biomedical labs. The visual world refreshes every episode without changing the format.

Season 01
The Elements

GraphCast, GNoME materials, energy.

Season 02
The Code of Life

AlphaFold, healthcare breakthroughs.

As long as DeepMind breaks boundaries and publishes papers, we have a predictable pipeline.

Roles · Budget · AI Pipeline

Scaling without
burning out.

The operational blueprint: hybrid staffing, tiered budgeting, and frontier AI workflows.

Trying to produce a high-end, global documentary series purely in-house is a bottleneck. But handing it entirely to an agency strips Google DeepMind of creative control and a chance to develop its visual language and voice.

My approach is a hybrid production model. Agency logistics meet in-house soul. We outsource the heavy lifting — global logistics, crews — to nimble external partners while retaining creative ownership, format guardianship, and post-production template design in-house.

Team & Roles

Executive Producer (Me) Internal
Creative oversight, multi-platform strategy, format safeguarding, budget management, and workflow automation.
Producer FTE/Contractor
Brings each episode to life based on the strategy, format, and creative plan. Manages the budget per episode and produces the edit.
External Production Company Agency
Manages local fixers, global location permits, cinematography, crew, and initial offline editing (if needed).
Editor Freelance
Pulls together the edits alongside the producer.
Freelance Motion Designer Specialist
Builds the standardized templates for the graphics — the visual signature toolkit.
Scientific Liaison Internal
Ensures scientific truth, fact-checks scripts, and unblocks communication between researchers and producers.

Indicative Budget

~£30K – £50K
Per episode · Global location production
£30K covers approx 3 days on location internationally, cinematic crews, and motion designer. £50K covers the same with a freelance producer and editor included.
3 days
On-location shoot · per episode
Tight, journalistic shooting windows — designed to capture authentic momentum without overstaying.
Hybrid
Production model
External agency owns global logistics and crew; in-house owns creative direction, post-production templates, and brand voice.
Figures reflect premium documentary standards shot on location. Estimates represent 'hard' costs of production and assume the contractor/FTE roles of the Executive Producer, Producer, and Editor are accounted for separately.
Frontier workflows

AI in the production pipeline.

We can scale output without inflating headcount by embedding AI workflows directly into the lifecycle:

01 — Pre-Production
Gemini scripts & treatments.

Synthesises 50-page whitepapers into accessible visual metaphors, script outlines, targeted interview questions, treatments, and graphics ideas. Effortlessly templates briefs.

02 — Storyboarding
Atmospheric storyboards.

Video and image generation tools build rapid, atmospheric storyboards during pre-production — moving the team from idea to visual brief in hours, not weeks.

03 — Localisation
AI dubbing at global scale.

Automated translation and transcription scale Living Proof across global languages without post-production friction.

04 — Operational Automation
Call sheets, logistics, upscaling.

AI models instantly produce call sheets and schedules, manage travel logistics, and upscale lower-resolution field assets.

Creative north stars · Pitch assistant

Creative
north stars.

Proven formats anchoring our production standards.

Google Maps Series · IMakeApps

Benchmark for conversational pacing, personality, and fast-moving format execution.

Square · Dreams Series

The gold standard for premium cinematography, composition, and human-first brand messaging.

Cleo Abram · Huge If True

Blueprint for kinetic mixed-media editing, accessible pacing, and seamless data-on-screen integration.

Google · Search On Series

Definitive proof of premium cinematography executed efficiently by nimble, operationally disciplined production teams.
AI Pitch Assistant

Speak to my Copilot.

To demonstrate how frontier AI tools can reinvent the traditional creative pipeline, I've built a custom Pitch Assistant loaded with this strategic plan and our production roadmaps. Scan to launch — pressure-test the strategy or audit my CV in real-time.

QR code — scan to open the Google DeepMind Pitch Assistant on NotebookLM

Scan to open the GDM Pitch Assistant