Thanks for highlighting the current Ferrari Luce discussion, it dovetails nicely into this. I have formatted a response, edited it, and maintained control over it. The full interview which was distilled here is included at the bottom of the post.
There is an aspiration realm that becomes more apparent with image-gen tools, in my mental model, this the the “dollhouse” designs that will never be made and do not have real criteria, they are play, they are therapy.
The gap between aspiration and capability in industrial design has always been there. AI image generation does not create that gap; it blurs it, in the same way CGI started to blur it, except now the dollhouse space gets flooded with convincing output and the noise-to-signal ratio goes exponential.
Capability
In industrial production, one person does not make everything happen. It is a team, a group of people, and a group of decisions, so capability means knowing what one’s frame is, mastering that frame, or remaining an attentive student learning the steps along the way.
That is situated knowledge — technical skill, judgment, process understanding, tool knowledge, and a realistic sense of where you sit inside a making-system — and a finished-looking image is not the same thing.
The blurring is a self-blurring: you are no longer able to clearly see the delta between where your skills end and where the output image begins.
Aspiration
There has always been a pyramid in design aspiration. At the top are the people who actually design the objects, and below them are different levels of ambition, all the way down to the kid sketching sports cars and sneakers and snowboards in a notebook.
The pyramid itself stays the same, but the bottom of it is potentially getting expanded. There are still mid-level designers who imagine they could become Ferrari car designers, and now there is also a much wider group of consumers and aspirants who feel newly within reach of that dream because image generation can carry their intention forward into a persuasive image.
The Dunning-Kruger effect runs through this: the less you know about something, the more capable you may imagine yourself to be, and AI tools make that feeling easier to maintain by handing you images that look like the work you wish you had done.
Productive Path
Meanwhile the productive path stays narrow. There is still a finite circuit by which a product reaches production — frameworks, chosen designers, factories, brands, and methods that move a project from sketch to manufactured object — and most design aspiration still lives outside it.
Around that circuit is the dollhouse space: imaginary car designs, imaginary shoe designs, imaginary products that are aspirational and often going nowhere beyond images. AI does not erase the productive path from the inside, but from the outside it pushes the noise-to-signal ratio up so far that it becomes harder to tell real capability from surface output.
That is why proof of work comes into the picture. To remain legible, the productive path is going to need more visible evolution and provenance — evidence that this came from a real process, not just a prompt and a nice render.
But proof of work is itself a moving target: it can be faked and simulated, and AI is just as capable of hollowing out today’s signals as it was of hollowing out rendering yesterday. The job ahead is to use the same imaginative tools designers apply to products to rethink credibility signals — to build frameworks where decisions, taste, and judgment are documented in ways that can survive the next round of simulation.
Boundary
Through all of this, the dollhouse space is not the enemy. Play, sketching, aspiration, drawing as a hobby or pastime, or living in a fantasy version of being a certain kind of designer are all legitimate expressions and a kind of therapy.
The point is not to shut that down, but to restore and strengthen the definition between that space and the productive path so both can keep existing without being mistaken for each other.
Open Questions
- The boundary between the productive path and the dollhouse space has never been fully stable — the argument assumes more separation than may actually exist in practice.
- Proof of work as a credibility mechanism can itself be simulated; the piece does not yet resolve what a stable, unfakeable signal would look like.
- One open question is what a stable, unfakeable credibility signal would look like inside the productive path as AI simulation improves.
- Another open question is how brands like Nike have commercially institutionalized the dollhouse space, and whether that is a separate problem or part of the same one; that thread was explicitly parked in the interview.
Supporting section: full clean interview
Frame
AI: Is that a fair one-sentence statement of the topic? Who is the audience for this forum post, and what are the stakes?
Shaw: The audience is industrial designers who are in the middle of this change — in the way industrial design is executed, in the environment, in the way it is looked at, and in the way creative people reach solutions.
AI: And the stakes — what do you want a reader to walk away believing or doing differently after reading this?
Shaw: I’m not looking to change anyone’s mind. I’m working through this myself. We know there’s going to be a change, but it isn’t apparent what that change is going to be. There are a lot of pre-existing misconceptions that get amplified by AI. A discussion is necessary, and it’s important to apply the same imaginative solutions we use for finding problems to the actual field of industrial design itself.
AI: What do you already believe is true about this topic — your prior commitments going in?
Shaw: With car design, there is the desire to focus on very high-status objects — a car, shoes, a trend object that is desirable and has status — and to imagine that one can design it. There’s a pyramid. At the top are the people who actually design it. Below that, different levels of aspiration. The dream aspiration is to be at the top — to be a Ferrari car designer. But the number of Ferrari car designers in the world is rather limited. Right down at the base is the kid sketching sports cars in a notebook. My idea going in is that the bottom of the pyramid is now getting expanded. The pyramid remains the same — there are still mid-level designers who imagine they could be a Ferrari car designer, who in actuality have no chance of reaching that goal. But the range somehow feels within reach. That’s the Dunning-Kruger thesis. The less you know about something, the more you think you’re an expert at it. But the image generation tools now blur the final output. Someone can draw a circle on a piece of paper, say “make me a new Ferrari model,” and Vizcom will generate one. That person feels supported the way you used to feel when your dad was holding the back of your bike seat — you thought you could do it, but you still had a ways to go.
Shaw: Parked: Nike’s “design your own” model as a possible related thread.
AI: Where do you place yourself in this pyramid?
Shaw: I’m not actively in this pyramid. I don’t imagine myself to be a car designer — I leave that to car designers. I’m in the industrial space. I know there are methods by which a product achieves production, certain frameworks, designers chosen to do that, methods that factories and brands use to bring a product to production. That’s a finite path. Beneath that, in the pyramid, there are large numbers of people who aspire to design cars and shoes and other things but don’t. And there’s a wider group of consumers who imagine they could do that — “I could design a Nike shoe,” “here’s my Nike shoe.” At the same time, as I use the tools I have to be aware of the glazing aspect, and the continuous self challenge that is needed to remain sharp, sort of experience vigilant.
Dialectic opening
AI: In one sentence — what is the central tension you want to work through?
Shaw: There have always been designers who aspire to design something they don’t have the capability to actually design — who fantasize or dream about being a good designer of shoes, cars, washing machines, whatever it might be. The current crop of AI image generation tools is going to further blur that capability, and without careful self-critique and self-challenging, that becomes a delusion you can slip into deeper.
Definition round
AI: What does capability mean to you in this context?
Shaw: Technical skill, judgment, process knowledge, knowledge of the tools, knowledge of one’s own place within the framework of making something happen. One person does not make everything happen — it’s a team, a group of people, a group of decisions. Capability is knowing what one’s frame is and being a master of that frame, or an attentive student learning the steps along the way. That’s the capability.
AI: What does blur mean here — does it hide the gap, shrink it, or make it feel irrelevant?
Shaw: When I say it blurs capability, I mean it blurs the perception of one’s capabilities. It used to be that if you needed to computer model something and have it rendered, there were different sets of skills involved, different needs that had to be met. The image generation tools apply a lot of statistically sampled background information and deliver it, often disguising it within context — loading up the image frame with users and environments and beautiful lighting, that golden-hour lighting that AI imagery is famous for. The blurring is a self-blurring. You’re not able to see the delta between where your skills end and where the output image begins.
AI: Is the disguise happening for the person who made the prompt, for the audience seeing the image, or for both?
Shaw: I’ve seen examples where a person does a very rudimentary sketch of a shoe, then compares it next to a beautifully rendered AI illustration of a fully worked-out science fiction shoe that’s nearly impossible to manufacture, and says, “Ta-da, I did that.” There are also the conversations around the new Ferrari — a little controversial in its design — where people are plugging prompts into image generators saying “generate me a new Ferrari,” and there are now hundreds, if not thousands, of images from wannabe designers. None of them is advancing up the hierarchy to become Ferrari designers. But there’s a lot of Dunning-Kruger going on.
AI: The distinction between claiming the output and claiming the capability that produced it — is that the core of what you’re getting at, or is there something underneath it that’s more uncomfortable than that?
Shaw: The core of what I’m getting at is the imaginary space that people work in — I refer to it mentally as the dollhouse space. The designs are going nowhere. It’s pretend. It’s aspirational. There is a certain productive path and that’s a very narrow path — cleanly defined from within, like an electrical circuit. There are a lot of non-productive paths, imaginary paths, imaginary car designs, imaginary shoe designs. That has the risk of overwhelming the signal when viewed from the outside. But when viewed from within the path, I don’t think that’s the case. The job of the actual participants in the productive path has to change. The signal remains, but the noise-to-signal ratio is going to increase drastically — exponentially.
AI: You said the job of actual participants in the productive path has to change — what specifically has to change for them?
Shaw: What has to change is there has to be a lot more proof of work. Everyone involved in the productive path is now going to be aware of the capabilities of image generation and AI — aware that it can hallucinate, that it can produce work that is unsubstantiated — and the substantiation and provenance of work along the productive path is going to have to be much more documented.
Premise exposure
AI: Is proof of work something you mean in a technical or process sense, or also in a reputational sense?
Shaw: Both. But mostly process. I’m talking about the ability to show the work evolving in a real path toward production, and to show the decisions and constraints that shaped it.
AI: If AI can also fake process artifacts, what makes proof of work strong enough to matter?
Shaw: It may not be strong enough forever. Proof of work is my speculation. It’s part of the thinking on how we maintain the credibility of the production pipeline — how we build in the framework of the decisions and how those frameworks are used to build on the final product. Proof of work is one suggestion in my thinking to define credibility in the middle of the design chaff tsunami that’s about to hit us.
Evaluation
Shaw: I’m not against the concept of play, sketching, aspiration, drawing as a hobby or pastime, or living in a fantasy world of being a certain kind of designer. That’s viable as an expression — a good therapy for expression. It should be celebrated on its own terms. The definitions just need to be structured from the inside in order to be more durable and have a continuing presence.
Shaw: Lock: the dollhouse space has legitimate value as expression, aspiration, and play — it should be celebrated on its own terms. The problem is not its existence but the collapse of the definition between it and the productive path, and the definitions need to be structured from the inside in order to be more durable.
