FLORA — Generative AI Canvas for Creative Teams

Product

Solutions

Enterprise

Pricing

Resources

Contact sales

The Creative AI Adoption Curve

The adoption of AI creative tools isn’t a race to automate everything. It's a series of choices, where what matters most is human judgment. Your judgment.

On one end of the curve, craft is the whole point and every asset is made by hand. On the other end, your collective creative taste is encoded into a system, and that system becomes what you and your organization build from. Though neither end of the curve is more advanced than the other, today most work happens at the base of the curve where most everything is hand crafted.

This is a map of where we think creative tools are headed, and we want you, the creative, to know which level fits the task in front of you, and what to look for when the work has outgrown your toolset.

Hand-Made

What it is

Every asset is shot, designed, and edited by hand. AI isn't in the production process. Someone on the team may be experimenting on their own, but nothing it makes reaches a campaign.

What you can use it for
One-off hero work where craft is the whole point and volume is low. A flagship campaign film, a founder portrait, the single image a launch hangs on.

What to watch
Output is capped by shoot budget and headcount while the channel and region list keeps growing. The volume gap is the reason teams leave this level, rarely a problem with the work itself.

When to move up
Move up when the volume of assets you owe each quarter stops fitting inside what your shoots and your team can produce, and you want AI to help with the parts that do not need a shoot.

Assisted

What it is

AI works beside you on isolated tasks while you make the asset. You give it a prompt in a separate tool or inside your editor, and it hands back raw material you react to and finish by hand.

What changes at this level
You stop starting from a blank canvas. The model gives you range and a first pass, and your taste does the rest. Nothing it makes ships as is.

What you can use it for
Mood boards, concept exploration, visual directions, headline and copy variations, quick mockups to get stakeholder buy-in before a shoot.

When to move up
You’re getting real value, but everything here is thinking, not production. Nothing ships. Move up when you want AI to produce assets that actually go out, starting from your real source material.

This is where the work inverts. Above, AI tools assisted you in creative production. From here, AI produces and you direct.

Human-assisted

What happens

AI produces the work and you correct most of it before it ships. The relationship has flipped. The model is the maker now, and you’re the editor. The source is usually a real shoot or a hand-built hero, and AI multiplies it into the variations the campaign needs.

What changes at this level
AI moves from helping you think to producing deliverables. Your job becomes reviewing and fixing rather than making from scratch.

What you can use it for
Resizing and reformatting a hero across channels, recoloring, reposing, background swaps, and generating the variations a campaign needs across aspect ratios and regions.

When to move up
The rework is starting to eat the time you saved, because the model keeps drifting off brand. Most teams stall right here. Move up when you load real brand context so the output stops drifting and you can trust the system inside a defined scope.

Human-directed

What it isAI produces at scale across channels, regions, and formats, and you direct it. You set the brief, the brand, and the standard, and you approve. The unit of work is still the campaign, and you are running each production yourself.

What changes at this level
You stop producing and start directing. Your hours go to the brief, the standard, and the call on whether the work is good rather than making the assets.

What you can use it for
High-volume variation across every channel and region for a campaign, with you steering each run and judging the result.

When to move up
You are directing every production run by hand, and the same setups keep recurring. Move up when you want to package those setups so they run without you. That is the point where the work stops being production and becomes building the system.

The work moves upstream. From here, the deliverable stops being the assets and becomes the system that makes them. Value concentrates in the judgment you encode, and the economics shift from billing for output to licensing a system you keep improving.

Systematized

What it is

You stop producing assets and build the reusable workflows that produce them. In FLORA's language, a Technique: a packaged, repeatable workflow that turns an input into a class of on-brand outputs. The deliverable is the workflow, and other people, even non-designers, can run it without understanding what is underneath.

What changes at this level
The work moves upstream. You are no longer making or directing assets one campaign at a time. You are building the machine that makes them, and your judgment gets encoded into the workflow rather than applied to each output.

What you can use it for
A launch-kit workflow regional teams run themselves, a paid-social variation engine, a product-photography pipeline, any repeatable production pattern you want to hand off and reuse.

When to move up
The workflows run well and others use them, but each one is a separate tool with its own logic. Move up when you want the brand itself, not just individual workflows, to be the system everything generates from.

Encoded

What it is

The whole brand is encoded as a machine-readable system: positioning, voice, audience, color, type, components, and motion, all structured so any agent can read it and build from it. The deliverable is the brand as a living system, what some studios call an atomic kit or a brand world model. Assets, pages, and campaigns generate downstream from it and inherit its logic. You maintain the system, and you’re paid for access to it rather than for any single output.

What changes at this level
The brand stops being a document that drifts after handoff and becomes infrastructure that holds. The economics shift with it, from billing for output to licensing a system you keep improving. The depth of the encoded understanding is the moat, and it compounds as the system runs.

What you can use it for
A brand any team can build from and stay on brand, a system that produces a coherent landing page, UI, or campaign from a single prompt, a living brand that updates everywhere when you change one value at the source.

When to move up
There’s (currently) no higher rung. The system can hold and produce the brand, and it will still converge to the median. The one thing it cannot do is the next section.

2026