lillian gary homelet’s talk
tech · ai · systems · communities · ideas

the blog

tech & aioct 16, 2026

Who Is AI Built For?

Useful help should be within reach of the people who need more capacity. That belief shapes how I think about AI, affordability, and who we're building for.read →
systemsoct 16, 2026

Designing Systems That Include Everyone

A system should be able to meet people in the middle of an ordinary day. Good design takes responsibility for the effort, assumptions, and barriers it creates.
communitiesoct 16, 2026

Tech Access in Our Communities

Access cannot stop at teaching people to use what somebody else built. Our communities need opportunities to develop their ideas, make decisions, and own the work.
systemsoct 16, 2026

Designing Systems That Include Everyone

If a system only works when people have uninterrupted time, remember every instruction and already know who to call, then it isn’t finished, because life rarely gives us those conditions. People get interrupted, miss a meeting, open something on a phone, or arrive without the background everyone else seems to have. Thoughtful designs make room for those realities. I care about this because the effort a system demands can determine whether someone gets to or will participate.

I have very little patience for an app that makes me work to figure out how to start working. When I open a project, I want to see the project, understand where things stand, and get to the details I need without setup screens and technical clutter taking over the experience. That preference comes from a larger belief about design, which is that the person using something should be able to see what it is asking of them and why, because every unnecessary step spends time and attention that belongs to that person.

The same thing happens outside of software, and it often costs people more there. For example, a family trying to complete a school process from a phone, where the instructions are in one email, a required document is somewhere else, and the form uses language that only makes sense to the people who wrote it. Then halfway through, the family realizes they needed information nobody told them to have ready and although each piece might look reasonable on its own, the full experience asks them to figure out how the pieces connect. The designed process or system handed them work that never had to exist.

That kind of hidden effort is something I’ve learned to watch for through my work in school design and my life as a mother and business owner. A process can be perfectly clear to the people who built it and still confusing to someone meeting it for the first time, which means it isn’t enough to ask whether the instructions exist. We have to find out whether people can actually use them. So that means, in the school situation, walking through the whole process on a phone, testing it with a screen reader, making it accessible and asking people what their experience was like. The people closest to the difficulty hold the information we need and they are usually the last ones we think to ask.

Once we know where people struggle, fixing it comes down to a series of decisions, because simplicity doesn’t happen by accident. We decide what someone needs to know first, which steps are truly necessary, and where a person can get help. Then we tell them what to prepare before they begin and build the process so they can step away after an interruption and return without losing their work. When a task needs explanation, that explanation should be easy to find and when a decision has consequences, the important details should be in plain sight before anyone agrees. Clearly designed systems respect a person’s time and their right to make an informed choice.

Even when those decisions are made well, the work doesn’t end once a system is live, because what people do with it tells us what we missed. Repeated questions are telling us something and so is the separate spreadsheet someone keeps because the official process doesn’t give them what they need. We should be curious about those workarounds instead of treating them as people failing to follow directions and we should make sure the feedback we collect reaches someone who can act on it. If we ask people to describe a frustrating experience and then leave the same problem in place, we’ve given them very little reason to offer their time again.

Designing for everyone is an ambitious responsibility, and people will still have needs we didn’t anticipate, but that is a reason to keep listening and improving rather than a reason to stop. I judge a system by what it allows someone to do on an ordinary day, including the day they are tired, new, or interrupted, when they should still be able to understand what is needed, take the next step, and reach a real person if the expected path breaks down. If we ask people to use a system, we owe them the care it takes to make it work.