From Idea to Object Vol. 2

Building a bathroom vanity when the market refuses to make the one you actually need

Bathrooms are full of fake precision.

Everything looks resolved. Crisp edges. Clean alignments. Stone. Mirror. Faucet. Light. The room presents as finished, even when most of its usefulness is determined by things you barely notice until they go wrong: where the sink lands, how deep the cabinet is, what the drawer clears, whether the material survives water, whether the mirror fogs, whether storage actually matches the routines of the people using it.

That is what makes the bathroom vanity such a good object for this series.

On the surface, it is ordinary.

In reality, it sits at the intersection of ergonomics, plumbing, storage, finish durability, lighting, hardware, electrical coordination, and daily habit.

It is exactly the kind of thing many people assume should be easy to buy.

It usually is not.

The vanity I wanted for my home did not exist as a product.

Or more accurately: the market offered many vanities, but almost none that solved the specific combination of problems I actually had.

I needed it to fit the room precisely. I needed it to work against a curved wall. I needed the proportions to feel architectural rather than generic. I needed the sinks, faucets, drawers, and mirror relationships to resolve cleanly. I needed the finish to survive a bathroom without looking like a landlord special. I needed the whole thing to feel calm, integrated, and deliberate.

That is where the project stopped being “find a vanity” and became “design and build an object.”

That is also where AI became useful.

The real problem was not style. It was fit

The easiest way to get lost in a project like this is to think the main decision is aesthetic.

White oak or painted. Pulls or finger pulls. Stone or solid surface. Minimal or expressive.

Those choices matter, but they are not the core of the problem.

The core problem is fit.

Not just whether the cabinet fits between two walls. Whether it fits the life around it.

How deep should the vanity be so it feels generous without crowding the room.
Where should the sink sit relative to the front edge.
Where should the faucet land relative to the drain.
How wide should the drawers be before proportions start to feel wrong.
How do narrow pullouts relate visually to wider drawer banks.
What finish gives the calm, pale look I want without introducing paint failure at edges.
What happens where a curved wall meets a rigid cabinet.
How do I get water resistance into plywood construction without making future refinishing miserable.

These are not catalog questions.

They are design-and-execution questions.

That is the gap I increasingly use AI to help close.

A vanity is not one object. It is a stack of linked decisions

One reason custom physical work is hard is that objects are never only themselves.

A vanity is not just a cabinet.

It is cabinet geometry plus sink placement plus faucet reach plus drain location plus countertop thickness plus backsplash logic plus mirror alignment plus lighting plus user movement plus cleaning behavior plus the uglier realities of water, swelling, sealants, and maintenance.

The moment I started pushing the vanity toward reality, the project split into interdependent layers.

The cabinet box had to be buildable from plywood.
The fronts had to look refined at room scale.
The storage had to make sense for daily use.
The hardware had to feel coherent across different drawer sizes.
The sink placement had to work visually and ergonomically.
The finish had to survive a wet environment.
The mirror and lighting layer had to coordinate with what sat below.
The wall conditions had to stop being “background” and become part of the design.

This is one of the places AI helps most: not by inventing taste, but by helping hold the dependency map in view.

You can ask one narrow question at a time while still keeping the whole system legible.

That matters because physical projects fail when one “small” decision quietly damages three others.

The room forced the object to become more specific

Generic vanities are designed to be sold to many rooms.

Custom vanities earn their existence by becoming specific to one room.

That was true here from the start.

The wall condition was not neutral.
The proportions needed to relate to the room, not just to a product standard.
The finish needed to work with the atmosphere I wanted, not just with what was easiest to manufacture.
The storage had to support real behavior, not imaginary staged behavior.

That pushed the project away from off-the-shelf thinking.

For example, I spent time thinking through vanity depth not as a standard number, but as a relationship between movement, sink position, faucet usability, and visual mass. The same was true of handle sizing and drawer front proportions. A decision that looks minor on paper becomes obvious when repeated across the full width of the piece.

That is one of the hidden truths of cabinetry: repetition amplifies every mistake.

A pull that feels slightly wrong on one drawer becomes loud when multiplied across a full elevation. A sink that sits slightly too far forward stops feeling luxurious and starts feeling crowded. A finish that works on a sample can feel dead across a larger built surface.

AI helped here because it let me reason through ratios, edge cases, and alternatives quickly enough to keep refining rather than settling.

AI was most useful where the tradeoffs were annoyingly specific

A lot of the project turned on questions too narrow to be glamorous and too important to ignore.

What is the right pull length relative to a 26-inch drawer front.
What happens visually if narrow pullouts sit next to wider drawers.
What finish gets plywood closer to pale oak without looking fake.
If I do not want to paint the vanity, what are my realistic options for getting a white, soft, hotel-like surface.
If I use plywood, what should happen at the exposed edges.
What coating sequence gives me water resistance now without creating a nightmare later.
Should the protection strategy differ for faces, edges, and inside corners.
What temporary construction choice will create rework later.

This is where AI became less like a search engine and more like a live reasoning partner.

Not because it knew my vanity better than I did.

Because it let me move through tradeoffs faster.

I could compare multiple finishing paths. Stress-test sizing logic. Think through future maintenance. Map the consequences of coating one surface now versus another later. Ask whether a detail was solving the real problem or only postponing it.

That kind of interaction matters in physical work because the biggest costs are often locked in before anything looks finished.

The expensive mistake is rarely the final one.

It is the early decision that quietly makes five later decisions harder.

The finish problem was really an identity problem

One of the deeper lessons in this build was that finish choices are rarely just technical.

They define what the object is allowed to be.

I did not want the vanity to read as standard painted cabinetry. I also did not want it to read as obvious plywood pretending to be something else. I wanted a pale, quiet, architectural result. Something closer to integrated millwork than furniture-store bathroom cabinet.

That opened a bigger question.

When the material under the surface is plywood, what is the most honest path to the final effect you want?

There are a few wrong answers available immediately.

You can chase a visual reference without respecting the substrate.
You can choose the fastest finish and then spend months compensating for it.
You can protect the object in ways that make future refinement harder.
You can solve for color while ignoring touch.
You can make the room look cleaner in photos while making the actual object feel cheaper in person.

This is another place AI helped well.

Not by making the decision for me, but by forcing the decision into explicit criteria:

Durability.
Repairability.
Edge quality.
Moisture resistance.
Tactility.
Visual honesty.
Future flexibility.

That is a more useful frame than “what finish looks best.”

Because physical objects do not just need to look right.

They need to age right.

The vanity kept exposing hidden systems

One of the most useful things about building a real object is that it reveals how many invisible systems sit inside ordinary life.

A bathroom vanity sounds like cabinetry.

In reality, it is cabinetry wrapped around plumbing.
It is lighting wrapped around moisture.
It is a mirror wrapped around electrical constraints.
It is a storage object wrapped around bodies, routines, and mess.

As the vanity evolved, it kept pulling other systems into view.

Mirror defogging.
Perimeter lighting.
Touch controls.
Low-profile drivers.
Removable access panels.
Drain compatibility.
Waterproofing inside drawer boxes.
What kind of pan or lining belongs at the cabinet base.
How to protect interiors now without trapping future problems.

This is one reason I think the physical-world use case for AI is more interesting than most of the discourse around AI-generated content.

When you are building something real, AI does not just help you make the object.

It helps you discover the object’s true perimeter.

What looked like “a vanity” turned out to be a coordinated system of cabinetmaking, moisture management, hardware design, electrical planning, and daily-life ergonomics.

That is not feature creep.

That is reality, finally becoming visible.

Custom work is often just refusal made constructive

There is a point in many projects where the real driver becomes clear.

Not ambition.

Refusal.

Refusal to accept the wrong proportions.
Refusal to buy something almost right and live with the compromise for years.
Refusal to let “standard” stand in for “good enough.”
Refusal to let the room be shaped by what is easiest to source instead of what actually belongs there.

That refusal can become perfectionism if you are not careful.

But it can also become authorship.

That is the more useful version.

The vanity mattered because it forced a lot of quiet standards into the open. How much dead space is acceptable. How much maintenance burden is acceptable. How many visual mismatches can exist before calm becomes clutter. Where the object needs to feel soft. Where it needs to feel exact. What should disappear. What should remain legible.

AI helped by making it easier to pressure-test those standards instead of abandoning them at the first sign of complexity.

That is a subtle but important difference.

The value was not that AI made the project easy.

The value was that it made persistence more practical.

Building the vanity changed the room before it was finished

One thing I keep noticing in projects like this is that the object starts changing the room long before the final install.

The moment you stop treating the room as a container and start treating it as a field of relationships, everything sharpens.

The wall matters differently.
The mirror matters differently.
The faucet stops being a fixture and becomes part of a composition.
Drawer widths stop being storage units and become rhythm.
Material transitions stop being details and become evidence of whether the room has integrity.

That is another reason I wanted this article in the series.

A vanity is mundane enough to be underestimated and complex enough to expose the whole thesis.

AI is most useful when it helps translate intention into a chain of executable decisions across multiple domains.

Not just a prettier idea.

A more buildable one.

A more durable one.

A more truthful one.

AI did not replace knowledge. It widened access to it

I still had to decide.

I still had to judge proportion.
I still had to understand what water does to edges.
I still had to think about coatings, clearances, plumbing constraints, and what hands touch every day.
I still had to test whether the elegant answer was also the durable one.

What AI changed was access to momentum.

It made it easier to ask precise questions without waiting until I had a full expert vocabulary. It helped me compare approaches before committing. It helped me spot where a local decision would create a downstream problem. It made hidden tradeoffs visible earlier.

That is a real shift.

Because many people do not fail to make better objects because they lack taste.

They fail because the translation cost is too high.

Too many unknowns.
Too many interdependencies.
Too much friction between idea and next step.

AI lowers that friction.

Not enough to eliminate craft.

Enough to make craft more reachable.

Why this object matters

The vanity matters because it turns a room people use every day into something more precise.

Not more luxurious in the abstract.

More resolved.

The storage can match behavior.
The proportions can match the room.
The surfaces can match the atmosphere.
The mirror, sink, drawer, and wall can stop arguing with each other.
The object can stop feeling purchased and start feeling placed.

That is a different kind of value than convenience.

It is domestic clarity.

And I think that is where some of the most useful AI-assisted making will happen: not in spectacular one-off objects, but in the things that quietly organize daily life.

Storage that fits.
Lighting that behaves.
Cabinetry that acknowledges moisture.
Repairs that do not look improvised.
Objects that feel inevitable once they exist.

The bathroom vanity is one of those objects.

Common enough to be ignored.

Important enough to shape the day.

What I would tell anyone building one

Do not start with style.

Start with constraints and relationships.

Map the room.
Map the users.
Map the daily behaviors.
Map the wet zones.
Map the maintenance burden.
Map the parts that must feel soft and the parts that must feel exact.

Then use AI to help you do four things:

1. Break the object into linked systems.

2. Compare options against real constraints, not imagined ideals.

3. Surface hidden downstream consequences early.

4. Refine the object by asking increasingly narrow questions.

Do not ask for “the best vanity.”

Ask what the room, the routine, and the material are actually asking for.

That is where custom work begins.

Why this belongs in From Idea to Object

This series is about the useful, tangible things that shape daily life.

Furniture. Lighting. Repairs. Plumbing. Storage. All the things that sit below the threshold of spectacle and above the threshold of importance.

The bathroom vanity belongs here because it is one of the clearest examples of the gap I care about.

The gap between wanting a room to feel resolved and knowing how to make the object that will resolve it.

The gap between preference and specification.

The gap between taste and execution.

The vanity started as dissatisfaction with what existed.

Then it became a set of questions.
Then a system.
Then a build path.
Then an object.

That is the movement this series is about.

From idea to object.