• Jean-Michel Blottiere — Founder & CEO, RTC – The RealTime Community

At RESET LA, we largely stopped asking whether AI would become part of production. The answer was yes – selectively, unevenly, and usually through hybrid workflows. We also agreed that the real test was no longer the demo, but measurable production value.

So we are not going to reopen that discussion here.

The question for London is what happens next: if these technologies genuinely change the cost and structure of work, where does value move — and what must the organisation become in order to capture it?

Then the killer question:

Are we optimising an operating model that is disappearing?

• Christopher Hebert — Senior Director, Global Marketing — SideFX 

Looking back at the questions SideFX was asking a year ago, Chris reflects on what has genuinely changed — and what has not — as AI moves from experimentation toward production. He invites participants to consider their own journey, and what those changes now imply for the workflows, organizations and decisions ahead.

A deliberately provocative opening conversation about what happens when AI changes not only how work is executed, but how decisions are made, authority is distributed, expertise is organised and accountability is assigned.

Core question:

If AI changes both the work and the decision chain, what must the organisation itself become?

Additional provocations:

• What happens when AI becomes an actor inside the organisation rather than simply a tool?
• If execution is increasingly automated, what is management for?
• If AI increasingly does the work, what is the organisation actually organising?
• Do we need a new target operating model — or an operating model designed to keep changing?

• Greg Demchak — VP of Emerging Technologies — Bentley Systems

• David Ayeni — Principal — Vistergy

• Sam Richards — Co-Chair, Technology Committee — Visual Effects Society

Insight contributors:

Patrik Schumacher — ZHA Architects | Lincoln Wallen — Framestore Company 3 Group | Peter Nofz — RodeoFX | Clemens Lindner — Senior Design & AI Researcher at ZHA and Trustee Director of the ZHA Employee Benefit Trust | Sam Richards — Co-Chair, Technology Committee — Visual Effects Society | Lisa Schulteis — Founder, ElectraLime Marketing 

Synthesis contributors:

• Edward White – Technical Director & Practice Builder · Independant 
• TBA

If AI makes existing work faster but also resets prices, expectations and competition, where must companies find new sources of growth and differentiation? What happens if successful AI transformation shrinks the business you currently know?

Insight contributors:

John Canning — AMD | Toby Hyder — XO Link | Denis Baudin — OffShoreVfx | Andy Mc Namara — Cinesite | David Gillespie — Foster + Partners | Tony Prosser — RealTime | Joseph Bell — AWS | Jose Frederick San Roman — Adelfas Capital Partners | Joel Maguet — The Third Floor

Synthesis contributors:

• Edward White – Technical Director & Practice Builder · Independant 
• TBA

As AI gives individuals and small teams capabilities that once belonged to large organisations, what should companies still own, control and standardise — and what should they buy, rent or build collectively?

Insight contributors:

Peter Nofz — RodeoFX | Lincoln Wallen — Framestore Company 3 Group | Sol Rogers — Magnopus | Pablo Selener — W4 Games | David Morin — Academy Software Foundation | Neil Trevett — Khronos | Elena Garces — Senior Research Scientist, Adobe Research | Francesco Siddi — Blender | Joel Maguet — The Third Floor

Synthesis contributors:

• Edward White – Technical Director & Practice Builder · Independant 
• TBA

At RESET London 2025, and again at RESET Los Angeles 2026, much of the discussion about production-readiness was framed by the proposition that hybrid AI workflows need to be anchored in mature, deterministic production tools. How has this conversation evolved? Are there new considerations for the why, where and how of production-ready workflows?

This session will explore what works, what doesn’t, and what’s next for hybrid workflows. It will open with each participant sharing one action taken this year, one success, and one failure.

The session will produce a summary of real-world examples shared by the group, along with a set of criteria by which the projects’ relative success/failure can be understood.

Working Session leads:

Judith Crow — SideFX | Danilo Papic — Netfork | John Canning — AMD

Contributors

Joseph Bell — AWS | Gary Mundell — Tippett Studio | Ava Oppenheimer — TryIto | Rich Welsh — SMPTE | Rikki Knight-Trembath — DNEG | Adrien Saint Girons — DNEG | Andy McNamara — Cinesite | Lincoln Wallen — Framestore Company 3 Group | Peter Nofz — RodeoFX  | Rick Stringfellow — Fellow, Head of Visual Content — EA | Sam Richards — Co-Chair, Technology Committee — Visual Effects Society | Elena Garces — Senior Research Scientist, Adobe Research | Lisa Schulteis — Founder, ElectraLime Marketing

Potential Angles

What architecture are we actually building?

Judith Crow — SideFX
What Must Remain Deterministic?

Ava Oppenheimer — TryIto
From AI Capability to Useful Systems

Sam Richards — Visual Effects Society
AI Classification, Provenance and Acceptable Use

Elena Garces — Senior Research Scientist, Adobe Research
From Research Possibility to Production Reality

What evidence proves it works?

Danilo Papic — Netfork
From AI Claims to Production Evidence

Andy McNamara — Cinesite
Can It Survive the Real Production?

Can it scale?

John Canning — AMD
When Experimentation Becomes Infrastructure

Joseph Bell — AWS
When AI Becomes Production Infrastructure

Rich Welsh
Standards, Trust and Production Scale

What should disappear/change?

Gary Mundell — Tippett Studio
What If the Pipeline Itself Is the Problem?

Rikki Knight-Trembath — DNEG
Designing for Zero Wait

Adrien Saint Girons — DNEG
From Pipeline Integration to Production Orchestration

What does this do to the organisation?

Lincoln Wallen — Framestore Company 3 Group
When Workflow Change Becomes Structural Change

Peter Nofz — Rodeo FX
Don’t Confuse Adoption with Reinvention

Lisa Schulteis — ElectraLime Marketing 
When Technical Success Isn’t Adoption  

What the session should produce

A Production-Ready AI Evaluation Framework

A practical test for deciding whether a new workflow should be:

adopted | hybridised | further tested | rejected

As generative systems become increasingly capable of reproducing appearance from widely available imagery, controlled capture needs a clearer value proposition. Is the scarce asset appearance, authentic performance, identity, consent, provenance, or the ability to reuse a performer reliably over time?

Potential leads / contributors:

Toby Hyder — XO Link | Addie Reiss — Digital Nation Entertainment | Brett Ineson — Animatrik | Joe Steel — Visual Skies | Greg Maguire — Humain / other capture and performance specialists

Potential Angles

Toby Hyder — XO Link
When Appearance Becomes Abundant

Addie Reiss
What Must Remain Human? 

Brett Ineson — Animatrik
From Capture Session to Reusable Asset 

Joe Steel — Visual Skies
From Capture to Production Asset

Greg Maguire — Humain
What Makes a Digital Human Valuable?

What the session should produce

A Human Digital Asset Value Framework

Productivity does not automatically translate into enterprise value. If technology changes the cost structure of production, the economic gain can flow to the producer, the client, the technology provider, the infrastructure layer, or disappear through lower prices and higher expectations.

Potential contributors:

José Frederick San Román — Adelfas Capital Partners | Andy McNamara — Cinesite | selected studio / technology / finance participants

Potential Angles

José Frederick San Román — Adelfas Capital Partners
What Remains Investable When the Economics Change? 

Andy McNamara — Cinesite
When the Market Splits

What the session should produce

A Value-Capture Map

Identify where economic value can realistically accrue across:

creator / studio / client / supplier / technology provider / infrastructure / IP owner / audience

and identify 1–2 alternative commercial models worth testing.

AECO firms already have BIM, reality capture, simulation, computational design, AI tools, proprietary knowledge and enormous volumes of project data. The problem is increasingly how to connect them into a durable operating architecture that supports decisions across design, engineering, construction and operation.

Potential leads / contributors:

David Ayeni — Vistergy | Greg Demchak — Bentley Systems | Austin Reed — HNTB | Reeti Gupta — Newcomb & Boyd | Mark Kauffman — WSP | Mateusz Gawad — Jacobs | Kerenza Harris — SCI-Arc / Morphosis | David Gillespie — Foster + Partners | Clemens Lindner — ZHA Architects | Vishu Bhooshan — ZHA Architects | Priam Givord — Priam Givord Studio

Potential Angles

David Ayeni — Vistergy
Accountability as Architecture 

Greg Demchak — Bentley Systems
From Tools to Connected Intelligence 

Austin Reed — HNTB
From Innovation to Repeatable Practice

Priam Givord — Priam Givord Studio
Beyond the Engineering GUI 

Reeti Gupta — Newcomb & Boyd
Making Intelligence Useful Across Disciplines

Mark Kauffman — WSP
When Everyone Can Build, How Do We Avoid Fragmentation? 

Mateusz Gawad — Jacobs
From Project-Level AI to Enterprise Capability

Kerenza Harris — SCI-Arc / Morphosis
Beyond BIM: From Data to Meaning 

David Gillespie — Foster + Partners
When Technology Becomes Abundant, What Still Differentiates the Practice?

Clemens Lindner — ZHA Architects
When Design and Making Converge 

Vishu Bhooshan — ZHA Architects
From Computational Design to Organizational Intelligence 

Edward White — Independant
Beyond Provenance: When Can an Image Be Trusted? 

What the session should produce

An AECO Enterprise Intelligence Reference Architecture

Not a technology stack, but the minimum layers required for:

capture → semantics → interoperability → intelligence / agents → governance → accountability → simulation → feedback

And ideally identify one shared pilot or standards problem the participating firms could pursue after RESET.

Much of the work AI can automate first is also the work through which people traditionally learned their craft. If junior tasks disappear faster than new development pathways emerge, today’s productivity gain could become tomorrow’s expertise shortage.

Potential leads / contributors:

Nandhini Giri — Purdue University | Thibaut Delahaye — Ecole des Nouvelles Images | education representatives | selected studio / AECO / technology leaders | Terrence Masson — Creative Technology & Media Executive | Habib Zargarpour — USC School of Cinematic Arts | Glenn Goldman — School of Art + Design | Rochele Gloor — rgloorLab / Natural Realities

Potential Angles

Nandhini Giri
If We Reinvent the Work, Who Reinvents the Learning?
Who becomes responsible for developing future expertise when the traditional career ladder begins to disappear?

Thibaut Delahaye
The Apprenticeship Gap
If AI eliminates the tasks through which junior people learned the craft, how do they acquire judgment?

Terrence Masson
Rebuilding the Path to Expertise
If entry-level tasks are disappearing, how must educators and employers work together to create new pathways through which emerging talent can develop judgment, craft and professional confidence? 

Glenn Goldman
Education Disrupted: What Must Endure—and What Must Be Reinvented?
What can previous technological disruptions—and today’s educational experiments—teach us about what must endure and what must change? 

Rochele Gloor
Keep Learning, Human
When tools and required skills won’t stand still, learning must become an ongoing form of self-development—helping people build the literacy, judgment and agency to question AI, adapt intentionally and remain relevant. 

Industry contributors

What the session should produce

A New Apprenticeship Model

Define:

what future practitioners still need to learn;

which experiences must remain hands-on;

where AI can accelerate development;

what industry must provide that education alone cannot;

one practical pilot involving employers and educators.

Each group uses the final 30 minutes to prepare Wednesday morning’s plenary readout: problem, conclusions, disagreements, next steps, required participants and proof.

They must answer:

1. What problem are we solving?

2. What did we conclude?

3. Where did we disagree?

4. What should happen next?

5. Who needs to be involved?

6. What would constitute proof?

From Reset to Rebuild: Reinventing the Business While It Is Still Running

Each group uses the final 30 minutes to prepare Wednesday morning’s plenary readout: problem, conclusions, disagreements, next steps, required participants and proof.

Rick Stringfellow — Fellow, Head of Visual Content — EA

How do you transform an organization while the existing business still has to deliver, generate revenue and meet today’s expectations? Rick Stringfellow opens a conversation with Jean-Michel Blottière and the RESET participants around the tension between optimizing what already works and building what may ultimately replace it — and how leaders can begin rebuilding when the future operating model is itself still changing. 

From Reset to Rebuild: Reinventing the Business While It Is Still Running

Each group uses the final 30 minutes to prepare Wednesday morning’s plenary readout: problem, conclusions, disagreements, next steps, required participants and proof.

Jean-Michel Blottière – Founder & CEO – RTC | RESET

Jean-Michel Blottière closes Day 1 with a brief synthesis of the day’s discussions, announces the AMD-sponsored gala dinner, and previews Wednesday’s program — Working Session readouts, cross-group synthesis, and the move from discussion to concrete next steps and shared actions

Sponsored by AMD

Welcome Remarks

John Canning — Director of Strategic Partnerships, Media & Entertainment — AMD

An informal evening to continue the conversations of Day 1 in a more relaxed setting, strengthen connections across industries, and reflect on the ideas that will carry into Wednesday’s plenary and action sessions.

• Jean-Michel Blottière — Founder & CEO, RTC — The RealTime Community

A brief synthesis of Day 1 and a look ahead: which ideas survived production reality, where the conclusions reinforce or contradict one another, and what are participants prepared to act on together?

Each Working Session reports back on the problem it addressed, the conclusions reached, the tensions that remain unresolved, and the actions or experiments it believes should move forward.

Hybrid Workflows & Production-Ready AI — From Claims to Evidence

Judith Crow — SideFX | Danilo Papic — Netfork | John Canning — AMD | Sam Richards — Co-Chair, Technology Committee — Visual Effects Society

Can we demonstrate that a new workflow is genuinely better once control, revisions, integration, supervision, compute, provenance, acceptable use and total production cost are taken into account?

Volumetric Capture & the Human Asset — What Is the Performer Actually Worth?

Toby Hyder — XO Link | Addie Reiss — Digital Nation Entertainment | Brett Ineson — Animatrik

As appearance becomes easier to reproduce, what remains uniquely valuable about authentic performance, identity, provenance and a reusable human digital asset?

AI Economics & Value Capture — Who Will Capture the Value of AI Transformation?

José Frederick San Román — Adelfas Capital Partners

When technology changes the economics of production, where can durable value, margin and new commercial models actually be created?

AECO — From Project Data to Enterprise Intelligence

David Ayeni — Vistergy | Greg Demchak — Bentley Systems | David Gillespie — Foster + Partners | Clemens Lindner — ZHA Architects | Vishu Bhooshan — ZHA Architects

How can fragmented project data, tools and expertise become a durable enterprise intelligence architecture spanning design, engineering, construction and operation?

Talent & Expertise — Talent Without the Ladder

Nandhini Giri — Purdue University | Thibaut Delahaye — École des Nouvelles Images

If AI automates the work through which people traditionally learned their craft, how must industry and education redesign the path to expertise?

From Shared Problems to Shared Action

The afternoon is designed to distinguish between three kinds of outcome:

1. problems individual companies should solve themselves;

2. opportunities that may become commercial ventures;

3. shared industry problems that may require collective action.

The goal is not to launch initiatives for the sake of launching them. A proposal should move forward only if the problem is sufficiently shared, painful, measurable and actionable.

Working Session leads and contributors from across the summit will reconvene for the Action Lab. Remote participation will remain open for contributors whose expertise or organizations are relevant to the initiatives moving forward.

Which problems identified during RESET are sufficiently shared, painful and well defined that participants should consider testing or solving them together? Potential initiatives will be examined against practical criteria including measurable value, willing participants, manageable scope, governance, IP, resources and the possibility of a meaningful 60–90 day outcome.

A central question:

When an industry recognizes a common inefficiency but no single company can justify solving it alone, is the barrier technological — or the way innovation is funded and coordinated?

Questions to pressure-test

The problem

• Is the pain genuinely shared by several organizations?

Invited perspectives: Austin Reed — HNTB | Judith Crow — SideFX | Toby Hyder — XO Link | Reeti Gupta — Newcomb & Boyd | Andy McNamara — Cinesite | Mark Kauffman — WSP | Addie Reiss — Digital Nation Entertainment

• Is it specific enough to solve?
• What happens if nobody acts?

Invited perspectives: Patrik Schumacher — ZHA Architects | Denis Baudin — OffShoreVfx | Sol Rogers — Magnopus

Value

• Is there measurable economic, operational or strategic value?

Invited perspectives: José Frederick San Román — Adelfas Capital Partners | David Gillespie — Foster + Partners | John Canning — AMD | Tony Prosser — RealTime

• Who benefits — and who actually captures the gain?

Invited perspectives: Joseph Bell — AWS | Toby Hyder — XO Link | Pablo Selener — W4 Games | Greg Maguire — Humain

• Who would be prepared to commit resources?

Scope

• Can the idea be reduced to a meaningful 60–90 day test?

Invited perspectives: Judith Crow — SideFX | Danilo Papic — Netfork | David Ayeni — Vistergy | Greg Demchak — Bentley Systems | Ava Oppenheimer — TryIto

• What is the smallest useful version of the experiment?
• What would constitute success or failure?

Invited perspectives: Danilo Papic — Netfork | Elena Garcés — Adobe Research | Dave Schulz — MPC | Kerenza Harris — SCI-Arc / Morphosis | Priam Givord — Priam Givord Studio

Participation

• Which organizations need to participate?

Invited perspectives: Greg Demchak — Bentley Systems | Brett Ineson — Animatrik | Austin Reed — HNTB | Joseph Bell — AWS | Reeti Gupta — Newcomb & Boyd | Nandhini Giri — Purdue University | Mark Kauffman — WSP | Joe Steel — Visual Skies | Kerenza Harris — SCI-Arc / Morphosis

• Is there a credible owner or champion?

Invited perspectives: John Canning — AMD | Sol Rogers — Magnopus | David Ayeni — Vistergy | Peter Nofz — Rodeo FX

• Does the initiative require technology providers, practitioners, researchers, educators or investors?

Invited perspectives: Elena Garcés — Adobe Research | Neil Trevett — Khronos | Thibaut Delahaye — École des Nouvelles Images | José Frederick San Román — Adelfas Capital Partners | Sam Richards — Visual Effects Society

Production reality

• Can the proposal survive real production or project constraints rather than only a controlled experiment?

Invited perspectives: Dave Schulz — MPC | Austin Reed — HNTB | Gary Mundell — Tippett Studio | David Gillespie — Foster + Partners | Rikki Knight-Trembath — DNEG | Reeti Gupta — Newcomb & Boyd | Priam Givord — Priam Givord Studio

• Which parts must remain deterministic, controllable, traceable or revisionable?

Invited perspectives: Judith Crow — SideFX | Sam Richards — Visual Effects Society | David Ayeni — Vistergy | Elena Garcés — Adobe Research | Rich Welsh — SMPTE

• At what point should an existing workflow or process be redesigned rather than optimized?

Invited perspectives: Gary Mundell — Tippett Studio | Clemens Lindner — ZHA Architects | Peter Nofz — Rodeo FX | Adrien Saint Girons — DNEG

Governance, standards and IP

• What can be shared — and what must remain proprietary?

Invited perspectives: David Morin — Academy Software Foundation | Pablo Selener — W4 Games | Neil Trevett — Khronos | David Gillespie — Foster + Partners | Rich Welsh — SMPTE

• What standards, provenance or acceptable-use rules are needed before an initiative can scale?

Invited perspectives: Sam Richards — Visual Effects Society | Neil Trevett — Khronos | Rich Welsh — SMPTE | Elena Garcés — Adobe Research | David Ayeni — Vistergy

• Who owns what is created — particularly when people, data, models and captured assets are combined?

Invited perspectives: Addie Reiss — Digital Nation Entertainment | Brett Ineson — Animatrik | Joe Steel — Visual Skies | Greg Maguire — Humain | Toby Hyder — XO Link

• Can the initiative remain neutral enough for competitors to participate?

Funding

• Who should pay?

Invited perspectives: José Frederick San Román — Adelfas Capital Partners | John Canning — AMD | Joseph Bell — AWS

• Is this a venture opportunity, a consortium problem, shared R&D, open infrastructure, or something else?

Invited perspectives: José Frederick San Román — Adelfas Capital Partners | David Morin — Academy Software Foundation | Neil Trevett — Khronos | Pablo Selener — W4 Games

• If everyone agrees the problem is real but nobody individually can justify solving it, is the barrier technological — or the way innovation is funded and coordinated?

Invited perspectives: Rick Stringfellow — EA | Nick Cannon — Walt Disney Animation Studios | John Canning — AMD | Greg Demchak — Bentley Systems

What the session should produce

A shortlist of one to three candidate initiatives

with enough clarity to move into ownership and pilot definition.

Owners, Pilots & Next Checkpoints

The strongest proposals move from ideas to commitments. For each initiative that survives, participants identify who will lead, who needs to participate, what will be tested, what resources are required, what remains proprietary or shared, and what evidence will determine whether the initiative should continue.

Participants whose expertise or organizations are directly relevant to a surviving initiative will be invited to help shape the next step, whether joining in person or remotely.

For each proposal

• Who leads?

Potential owners drawn from the relevant Working Session leads and contributors.

• Who participates?

Invited perspectives across communities: Austin Reed — HNTB | Addie Reiss — Digital Nation Entertainment | Reeti Gupta — Newcomb & Boyd | Joseph Bell — AWS | Mark Kauffman — WSP | Brett Ineson — Animatrik | Kerenza Harris — SCI-Arc / Morphosis | Nandhini Giri — Purdue University | Mateusz Gawad — Jacobs | Greg Maguire — Humain

• What exactly are we testing?

Invited perspectives: Judith Crow — SideFX | Danilo Papic — Netfork | David Ayeni — Vistergy | Greg Demchak — Bentley Systems | Toby Hyder — XO Link | Ava Oppenheimer — TryIto | Nandhini Giri — Purdue University | Thibaut Delahaye — École des Nouvelles Images | Priam Givord — Priam Givord Studio | Terrence Masson — Creative Technology & Media Executive

• What happens during the first 60–90 days?

Invited perspectives: Danilo Papic — Netfork | Rikki Knight-Trembath — DNEG | Vishu Bhooshan — ZHA Architects

• What resources are required?

Invited perspectives: John Canning — AMD | Joseph Bell — AWS | José Frederick San Román — Adelfas Capital Partners

• What remains proprietary — and what is shared?

Invited perspectives: David Morin — Academy Software Foundation | Neil Trevett — Khronos | Pablo Selener — W4 Games | Sam Richards — Visual Effects Society | Rich Welsh — SMPTE

• What evidence determines continue / change / stop?

Invited perspectives: Judith Crow — SideFX | Danilo Papic — Netfork | Elena Garcés — Adobe Research | Andy McNamara — Cinesite

• When is the next checkpoint?

Relevant initiative owners and participants.

What the session should produce

For each initiative that survives:

Owner → Participants → Pilot → Resources → Evidence → Next checkpoint

The objective is quality rather than quantity. One or two initiatives with genuine ownership and a credible path forward are more valuable than a long list of aspirations.

What Must We Stop Doing?

If the conclusions of RESET London 26 are true, what is one assumption, workflow, investment, organizational habit or business practice that we should stop carrying forward?

This final conversation focuses not on predicting the future, but on identifying what the industry is ready to leave behind.

All participants — in person and remote — are invited to contribute one final RESET.

Possible lenses:

• a workflow that exists only because previous technology required it;
• a capability companies should stop rebuilding internally;
• an organizational boundary that no longer makes sense;
• an assumption about scale, labor or competitive advantage;
• an investment being protected because of sunk cost rather than future value;
• a way of training, managing or measuring people that no longer serves the organization;
• a technology dependency that has become strategic risk;
• a handoff, approval or process that no longer creates value.

What the session should produce

A concise set of “RESETs” — assumptions and practices the group believes should be actively challenged rather than passively inherited.

• Jean-Michel Blottière — Founder & CEO, RTC — The RealTime Community

Closing reflections and the path from RESET London 26 toward the next checkpoints, pilots and collaborative initiatives.

The final message can return to the central RESET London premise:

The challenge is no longer simply access to powerful technology. It is deciding what remains valuable as previously scarce capabilities become widely available — and rebuilding businesses and organizations around that new reality.