The showroom and the screen: Cars, AI, and social change

I am interested in the auto industry partly because so many different ideas of progress meet in the same product.

A car can be transportation, enjoyment, an identity, a financial obligation, and a public signal at the same time. As a car enthusiast, I find that combination useful for thinking about the relationship between what an industry makes, what it promises, and what people come to want.

Those questions also run through technology and AI. Both industries have to make capabilities understandable to a buyer, but the things that are easiest to compare or display are not always the things that matter most in daily use.

I treat these as parallels, not proof that car sales explain society or predict the next stage of technology.

A purchase with several jobs

Cars occupy an interesting space between everyday consumption and a decision as substantial as buying a home. Changing a car can still be expensive and disruptive, but for many people it is a more plausible way to change the visible shape of their life than moving house.

That makes the purchase worth examining. What is the buyer asking the vehicle to do beyond getting them somewhere?

Consider a new Toyota RAV4 and a used Mercedes. It would be easy to turn that comparison into a story about practicality on one side and vanity on the other. I do not think the badge gives us enough information for that conclusion.

The Mercedes buyer might understand the particular model, have allowed for upkeep, and genuinely value the experience. The RAV4 buyer might be choosing exactly what their circumstances require, or spending more than necessary because a particular version feels socially expected.

The more useful questions concern the whole situation: what the person needs, what they enjoy, what they can sustain, and what they expect the purchase to communicate.

Veblen's discussion of conspicuous consumption is relevant here. It examines how spending can establish or display social standing, rather than serving only a practical purpose. I find that helpful as a question about incentives, not as a way to decide that every expensive preference is insincere.1

Capability and the story around it

An automaker has to decide where to put its effort. Improving a vehicle's everyday behavior, changing its appearance, adding a feature, and strengthening its brand are different ways of making it more attractive. They may reinforce one another, but they can also compete for attention.

I see a similar choice in technology.

A product can become more capable in ways that matter to its users. It can also become easier to present as capable. A striking interface, a recognizable phrase, or a feature that performs well in a demonstration can make the second kind of improvement easier to notice.

This is not an argument against branding. People need ways to understand differences, and a reputation can represent years of real work. The question is whether the distinction being sold corresponds to something the customer will experience.

A feature that photographs well is not necessarily a feature someone will appreciate every morning. An AI feature that makes a demonstration impressive is not necessarily the one that makes a recurring task easier.

I find the parallel useful because it shifts attention from what can be shown at the point of sale to what remains valuable after the introduction is over.

Horsepower and benchmarks

Performance figures matter. So do technical benchmarks. I would not dismiss either just because they do not tell the whole story.

But a question has to come before the number: performance at what, under which conditions, and for whose purpose?

Imagine a car that is exceptionally quick but poorly suited to the roads, distances, or passengers that matter to its owner. The capability is real. The mismatch is real too.

The equivalent in AI might be a model that performs impressively on a benchmark but is too costly, slow, or inconsistent for a particular job. Or a product might contain a sophisticated model while leaving the surrounding process difficult to understand.

My point is not that lower performance is secretly better. It is that choosing the highest number is different from choosing the best fit. The strongest option for one task might be excessive for another and insufficient for a third.

That gives me a more practical way to think about progress. I want to know what the new capability enables, not just where it places the product in a comparison.

What the monthly payment leaves out

The financial side adds another useful parallel.

For a given amount borrowed and interest rate, extending an auto loan can lower the monthly payment while increasing total interest. A borrower can also owe more than the vehicle is worth. The Consumer Financial Protection Bureau distinguishes these issues from the immediate affordability of a payment.2

This is not a claim that financing is inherently irresponsible. It is a reminder that the number closest to the purchase decision may not describe the entire obligation.

Software has its own version of that problem. A modest subscription price does not, by itself, tell me what setup, integration, staff time, support, or eventually changing systems will require. A quick prototype also does not tell me what maintaining it will involve.

The analogy is imperfect. Software and vehicles have different economics. Still, I find the same question useful in both cases: what am I committing to after I say yes?

That question becomes especially important when a product's appeal depends on anticipated future value. I want to distinguish what it can reliably provide now from what I am hoping it will become.

The image of demand

An expensive SUV or an exotic car can carry a large symbolic presence. A photograph communicates the object immediately. It does not communicate the buyer's income, the financing terms, the ownership history, or whether the car is even theirs.

A $200,000 SUV could be a manageable indulgence for one person and a serious source of pressure for another. I cannot tell which by looking at it.

The same limitation applies when I try to infer a market trend from what I see online. A feed full of supercars is not a representative sample of car ownership. It may tell me more about what captures attention than what most people are buying.

This is where the idea of hyperreality becomes a useful analogy for me. A circulating image can help establish an expectation of what an ordinary successful life should contain, even when the circumstances behind that image remain unclear.3

I would want actual evidence about purchases, ownership costs, and financing before treating that expectation as a description of society. Attention and adoption are not the same thing.

What I bring back to building

I do not want these comparisons to turn enthusiasm into something that needs an apology. Enjoyment is a legitimate part of usefulness. A car can matter because someone loves driving it. A digital product can matter because it makes an activity pleasurable, expressive, or social.

What I want to distinguish is an experience someone genuinely values from an obligation they accept mainly to inhabit its image.

For a builder, that distinction creates a responsibility. Am I making a product that fits a person's life, or encouraging them to reshape their life around a promise the product only partly fulfills?

Cars make the question tangible for me. However compelling the launch, the owner eventually has to use the vehicle on an ordinary day. I think technology deserves the same test.

Sources and notes

  1. Thorstein Veblen. The Theory of the Leisure Class, Chapter Four: Conspicuous Consumption. 1899; public-domain text via Project Gutenberg. ↩
  2. Consumer Financial Protection Bureau. Auto loans key terms. ↩
  3. Stanford Encyclopedia of Philosophy. Jean Baudrillard. ↩