A study by Aquia for the Think Tank Arts & Métiers
What happens when AI joins the engineering team?
Engineering used to mean choosing what to give up. Now the machine explores, and the engineer decides what deserves to exist. The white paper follows that shift in four trails and twelve trends, with ten engineering fields in depth and more than five hours of conversation to listen to on the way.
Volume I was written from November 2024 to February 2025, two years after ChatGPT. Volume II is written from October 2026, two years later.
In a hurry?
Watch a minute, or listen to eleven.
Film · 1:13
The study, told in a minute and a bit.
Narrated, with captions. The question, the shift, the four moves, the ten fields and the conversations. Also cut for phones: vertical version (MP4).
Listen first · 11 min
The white paper, in one conversation
The whole study at once: four trails, twelve trends, and what they mean for the engineer.
Download the audio (2.0 MB)
Remember progress on this device (optional).
Where it starts · the background chapter
The equation hasn’t changed. But the engineer has.
The white paper opens with how engineering got here: from Turing’s question to generative AI, where it already works, and what comes next. In its own words:
Engineering was once a world of precision and control—a place where equations ruled and iteration carved the path to progress. But that world is changing. Generative AI isn’t just speeding up what engineers already do—it’s flipping the script entirely.
Now, systems don’t just assist. They collaborate—testing thousands of ideas, anticipating failure before it happens, and proposing designs no human would’ve imagined.
Part 1 of the background
The generative shiftWhat if equations could do more than calculate?
Part 2 of the background
From blueprints to living systemsIt began as a blueprint. Now it breathes, learns, and evolves.
Part 3 of the background
Creative intelligence with autonomyWhen AI isn’t just fast, but bold—and asks questions back
Read the background chapter · or listen, 38 min
Four trails, twelve trends
The engineer’s job, rewritten in four moves.
Each trail of the white paper follows one move in the engineer’s work, through three trends. Read them in order, or start with the move closest to your own work.
Fig. 1
Each trail moves the engineer’s work from one word to the next.
The four pairs are the white paper’s summary of its trails: “a journey from compromise to coherence, from rigidity to resilience, from tools to teammates, and from execution to ethics.” The roles are quoted from each trail’s conclusion.
HypothesisYour field
Ten fields, each in depth.
Each field has its own chapter, built from the sources of that discipline—startups, industry and research—and its own conversation to listen to.
Seven core engineering domains
- Electronic, IT and telecom
AI-driven circuit optimization, next-gen telecom networks, and automated code generation.
Conversation, 31 min
- Electrical
Smarter grids, predictive fault detection, and AI-powered energy systems.
Conversation, 24 min
- Materials and nanotechnology
AI-guided material discovery, self-assembling nanostructures, and adaptive smart surfaces.
Conversation, 13 min
- Industrial engineering and systems
AI in digital twins, Industry 4.0, and IoT driving hyper-efficient automation.
Conversation, 18 min
- Biomedical
AI-driven diagnostics, robotic prosthetics, and personalized medicine powered by generative design.
Conversation, 22 min
- Civil
AI-assisted urban planning, self-healing infrastructure, and automated site monitoring.
Conversation, 19 min
- Chemical
AI-enhanced molecular synthesis, predictive modeling, and green chemistry solutions.
Conversation, 40 min
Three supporting domains that shape engineering
- Education and training
AI-driven learning assistants, adaptive skill development, and engineering simulation labs.
Conversation, 17 min
- Engineering management
AI-enhanced project optimization, automation of compliance, and cross-team collaboration.
Conversation, 13 min
- Mathematics and AI
The hidden force behind all engineering, from physics simulations to AI-driven theorem solving.
Conversation, 29 min
Listen
More than five hours of conversation, for when reading is not an option.
Every chapter and every field comes with a podcast-style conversation, generated with NotebookLM from the white paper: two hosts walk through the ideas with examples. Listen on the page, download the files, or follow the feed in a podcast app.
Volume II
The Second Act
Two years after the study, a second volume asks what changes when AI begins to perform the work of engineering, not only assist it: a foundation, four essays, and a register of evidence, outlined here while they are written.
Read the outlines Editorial outline · In development
How it was made
More than 900 sources, explored with machines and steered by people.
The authors built a taxonomy from their own expertise and had several language models challenge it; explored the literature field by field—papers, startups, industry, talks; then drafted with language models from their own outlines and edited every draft. The Blueprint tells the method in full, with its pseudocode.
Authors
Christophe Pennetier, André-Benoît De Jaegere and Xavier Kestelyn
Christophe Pennetier
Chief AI and Optimization Scientist at Quincus
André-Benoît De Jaegere
VP Innovation and Strategy at Capgemini Invent
Xavier Kestelyn
Professor at Arts et Métiers Sciences et Technologies
We would like to extend our deepest gratitude to the Think Tank Arts & Métiers for commissioning this report and supporting us throughout its creation. Their insightful guidance, thoughtful questions, and rigorous feedback have significantly enriched our exploration of generative AI's transformative potential in engineering. Their mission to illuminate how emerging technologies reshape engineering practices continues to inspire meaningful dialogues at the intersection of innovation, responsibility, and human ingenuity.