Creating with AI

From producer to creative director

Generative AI fundamentally reshapes the creative time equation. Before, 80 % of the time went into execution (Photoshop, retouching, derivatives) and 20 % into thinking. That's now reversing.

In this module, we look at what it concretely means: mass visual production, self-generating interfaces, and concrete examples of Arneo projects delivered in 2024-2025.

Industrialised visual production

The principle: the image factory

Before AI, a product photo shoot easily cost €10,000 to €50,000. And each derivative (season, mood, context) required a new shoot.

Today, from one well-executed foundational shoot, you can generate thousands of derivatives while staying faithful to the brand identity. This is the so-called "image factory", and it changes the economics of communication.

The three pillars of successful AI visual production

1. Strong brand anchor. You train the model (or prompt it) with identity references: palette, mood, composition.
2. An iterative workflow. Generate in batches of 4-8, keep the best, refine.
3. Human quality control. The human eye remains the final filter. Always.

Generative UX & hyper-personalisation

We move from "directed UX" (linear journeys designed in advance) to framed UX: you set a frame, and the interface itself self-generates in real time according to the visitor.

Directed UX vs Generative UX

Hyper-personalisation in practice

Hyper-personalisation is no longer "Hello Sophie" at the top of a page. It's:

• The hero banner that changes based on your prior journey (article read, product viewed).
• The recommendations at page bottom computed against your last 5 actions.
• Contextual internal linking: "read next" links rephrased on the fly based on what interests you.
• The summary of a product sheet adapted to your usage context (sporty vs beginner, pro vs consumer).

Arneo case studies

🥽 Optic 2000 · Virtual Try-On

Virtual fitting module: the user uploads their photo, the AI superimposes any pair of glasses from the catalogue in real time, with realistic rendering (size, curve, shadows).

Impact: higher conversion rate on online sales, fewer in-store fittings, market expansion in areas without physical stores.

🌿 Fermob · Garden Styler

The user takes a photo of their garden. The AI detects zones and proposes a layout with Fermob catalogue furniture, in the right style and colour.

Impact: turns a product page into a personal project simulator. The customer visualises at home, not in a generic showroom.

👁 Optic 2000 (bis) · Product image factory

Industrial production of lifestyle visuals from one foundational shoot. Several thousand seasonal derivatives generated, brand-faithful, ready for retail and social.

🪻 Fermob · Social media visuals

Generation of lifestyle inspiration visuals that preserve the real product details. The rule: you can generate the setting, mood and light, but the furniture stays pixel-faithful to the catalogue.

💊 Pierre Fabre · Generative UX journey

An interface that adapts in real time to visitor intent. A healthcare professional sees a prescription/clinical-data oriented UI. A consumer sees a benefit/support oriented UI.

🚗 Suzuki · Voice inspection reports

Dealership inspectors speak into their phone → the AI transcribes, structures and generates the full report automatically. Attached photos → automatic detection of scratches or damage.

Impact: 60-70 % time saved on report writing. The inspector can do 5 dealerships per day instead of 3.

📋 Aptitud · HR dossier generation

From a raw CV, automatic generation of a competency dossier in Mantu's brand format, ready to send to a client. The sales rep saves 30 minutes per profile.

💼 Kota · Fast quoting and pricing

Internal pricing tool that combines the company's rate cards with the client brief to produce a structured quote in minutes.

🧠 Waldo · Become an "expert" in 5 minutes

Ultra-fast synthesis to get operational on a complex topic (environmental regulation, vertical market, jurisprudence) before a client meeting. Structured research + sources + ready-to-brief deliverable.

📄 RFP chatbots

Rather than sending a 60-page PDF proposal to the client, you give them a chatbot that knows the recommendation. The client "chats" with the proposal: "Why this stack?", "How many days for phase 2?", etc.

Impact on creative professions

1. From production to reflection

Creatives no longer spend 80 % of their time on execution. That freed-up time goes to art direction, strategy, and the quality of the idea.

2. Augmentation, not replacement

A designer can now test 50 visual directions in a day, where they tested 3-5 before. It's a "human on steroids", not a human replaced. Selection, critique and finishing remain human.

3. Designing for machines too

Today, your site is no longer only read by humans: it's also crawled by AI agents that synthesise for their users. That's the whole point of Module 5 (GEO).

4. Accessibility & entry barrier

AI dramatically lowers the barrier to complex tools. Someone without Photoshop can produce a pro visual. Someone without Premiere can produce a video. It's a democratisation of creation that makes the expert eye and taste even more valuable.