The Software We Consume

The Collapse and Rebirth of Art + Commerce

What Advertising Knew

Advertising was built on a simple premise — art serving commerce. Turns out what drives our economy is also an effective lens into who we are, which is one of the most fascinating synergies. The one thing that I think some of the best ads achieve is that they surprise us in some way. They stick with us because they introduce us to something new; a new way of thinking about a product, of using language, a perspective of culture or our society, or even a new way of thinking about what an ad can be. They’re often reflective of ourselves; what we expect, assume or think we know, and challenge those things. As a whole, creating good ads, like creating art, is a very human exercise.

If I were to ask you what your top 3 favorite (or most memorable) ads were, chances are the ones you’d choose are multiple decades old. Creating these insightful, human communication concepts is not very efficient nor repeatable; two attributes that conflict enormously with how we approach business today. And perhaps the most incongruent aspect is that they often don't create a trove of data to analyze, meaning even if you hit on a great idea you may never know exactly why it was successful.

Much has already been written about the ‘blanding’ of corporate creative so I won’t go into that too much. Needless to say there’s been an effort to de-risk and optimize products and communications; to let the data that iterative development produces guide us towards our next actions. We rely on ‘best practice’, because that’s what we have the most data to support. This new process favors intentionally safe changes that can be tweaked without disruption until made effective enough to make a difference. If you think about it, we’ve been letting computers dictate our decision making for a while now, long before ChatGPT. It’s just that humans used to be needed to make the deck explaining the rationale behind the decisions, but now the computer can make the deck too.

Now, another important thing happened that went hand-in-hand with the blanding movement, and that is we moved everything onto platforms. Shopping, family/friends, education, work, etc. Just about everything was platformed and mountains of data was captured from users; who they were, what they do, etc. The more we know about users, the more efficient and optimized the product can be to deliver as much ‘value’ as possible. Platforming’s most consequential impact was that, unlike advertising where a human insight was needed to connect an idea with a desired behavior, platforms didn’t have to understand what customers were thinking because they create a steady stream of data on what they were actually doing.

This was a significant collapse in how we thought about the relationship of art and commerce. The art side of that tandem, where insights about people was how products could connect on a human/cultural level and hopefully drive behavior, became less and less recognizable. Sure there was still some image-making or copywriting involved, and it was generally created by people with a B.F.A using design tools, but we decided in order to best serve commerce, the objective of these communications wasn’t to influence behavior or empower a consumer to make a choice, but to create quality signals. The real behavioral shifts came from ‘nudges’, like serving consumers product and copy in algorithmically determined spaces or fabricating senses of urgency. The platform became the mechanism for influencing behavior via a feedback loop designed to reinforce existing behaviors. The platform serves as both the product and the ad, and value moved from being exchanged to being extracted.

The insights that made products human became less and less part of the equation as platforms prioritized collecting and optimizing human, but machine-readable, behavior like clicks, data, efficiency, ironically at the expense of actually understanding people.

Human-Shaped Software

I think a lot about where AI fits into our everyday, and I keep coming back to the idea that a lot of the AI anxiety that we collectively have is because, for a long time now, we’ve actually been building software for machines and selling them to people; and we’re finally at a time now where machines can be the users. Sure, as product designers we spend a lot of time trying to make software human-shaped and usable, but in reality a computer is much better suited to completing many of the tasks our software is created to facilitate.

This doesn’t mean that software isn’t for humans at all. However it does mean that we need to reconsider what using software means to people. Humans no longer need to be users of software, instead they can be consumers of it; and that distinction is important. Using software often means that the value comes from the output or result, but consuming software means the value comes from the experience, how it makes you think and feel, and how those feelings impact your behavior. For example, we use an instruction manual, but we consume a magazine. Different in purpose but equally valuable to a human.

Most of our modern design approach stems from usability practices. We design to reduce friction, for more intuitive UI, collect accurate data, etc. We operate under the assumption that everything is a task and that software must reduce the time and effort necessary to complete those tasks. We’ve determined that successful products are those that are easy to use, whereby easy to use often means ‘frictionless’. We’ve even built a whole industry worth of tools and methodologies dedicated to measuring usability so we know exactly where and what to design; which happens to be the engine that AI can run on to perform design tasks. It’s very possible that large swaths of designing software products, as we’ve thought about it over the past 20 years, is a solved problem.

And that’s a really great thing.

Software For Humans

In a world where AI is the user, and potentially a designer of software, we need to take the idea that digital products are meant to be consumed rather than used a lot more seriously. This means breaking away from the task-based ‘jobs to be done’ style of designing products, where efficiency equates to effectiveness. It means we need to think beyond software that simply smooths out a problem, and toward software that satisfies a human need. Consuming software doesn’t mean that the user has a passive relationship with the product, such as streaming a movie; it also doesn’t mean that the product can’t be a tool for some task. After all software is software and designing software to be ‘consumable’ doesn’t change what it is. Consumable software is rather a design philosophy where the purpose of a product is shaped by knowing something about people. Rather than creating solutions to problems we want to design experiences for insights.

For example there’s a lot of problems with living abroad and connecting with friends and family. Aside from the time differences and lack of physical presence, not having the ability to be present leads to feeling guilty and stressed about not maintaining relationships the way we used to. These are all tangible, human problems that software might be able to help us solve. But they are problems without a unifying insight or something that attaches a product to a human. An AI agent can reach out to our friends, while we’re asleep in another timezone, and check in to help relieve the stress of unmaintained relationships. Sure the problem that we’re looking to solve is rooted in the human experience, but the solution does not carry that humanity through.

However what could a product experience look like if we instead looked at designing for the user to spend time with the experience, rather than tactically solving a problem. There’s this phenomenon where even though you may not have seen a friend for a long time when you finally catch up with them it feels like no time has passed. It’s a distinctly human insight, one that evokes a feeling, that we could design a solution experience around. If instead of our product trying to bridge the gap in communication between friends under a given circumstance, what if our product’s primary objective was to create it’s own version of the feels-like-yesterday phenomenon. This might involve contextualizing communications around memories, shared interests, times of year or upcoming events; things that distract and obscure the things that might weaken a relationship like distance or the time between your last phone call and instead celebrate the history and rituals of a friendship.

Pointing Ourselves Out To Us

As humans, we’ll always want to buy products, consume services and accomplish goals, and in the world we live in all of those things involve interacting with software in some way; software remains an effective communication layer, especially in an agentic world. If the agentic vision comes to fruition, agents will understand our intent and execute actions on that intent. Humans will likely perform a single iteration for every 100 interaction an agent performs on their behalf. Many of the commercial spaces that humans inhabit online today will either be inhabited by agents, or just cease to exist. Human intent, how we think about the world will carry much, much more weight given that agentic actions (compute) will be spent on executing on it. In a world where humans are spending less time participating in the value extraction loop of current platforms, how will they be reached, and more importantly how will their behavior be understood, influenced and shaped? What will resonate with people?

I began this article talking about how great advertising’s ability to point ourselves out to us is what makes it memorable, and I think it’s possible that people will come to expect something similar with software. It’s becoming increasingly clear that as agents perform many of the rote, taxing, unpleasant digital tasks, we as humans may have more energy and time for experiences that are less efficient, but more emotionally and intellectually stimulating. It’s possible that the rules of product design, things like friction is failure and speed is success, will become usurped by rules that prioritize treating the user like a human rather than a machine. When AI handles the "using" of products, humans will be left with the "experiencing" piece, and that creates both a necessity and an opportunity to design software that connects with people on a human level the way the best creative always has.