A marketplace you can talk to

I’ve been building etc. around a simple exchange: you have something you no longer use, and someone nearby has something you want. Text a photo, say what you’d like in return, and let an assistant look for a trade.

The more I work on it, the more I think the underlying idea reaches beyond the stuff in our garages.

A lot of marketplaces exist to find a match. A person and a job. A household and a home. A bike someone doesn’t ride anymore and a pair of chairs someone else doesn’t need. The details differ, but each starts with someone trying to connect what they have or need with the right opportunity.

Usually, that person also does much of the work. They search, filter, compare, message, and come back tomorrow to see whether anything has changed. Even after describing what they want, they have to keep the search moving.

I want to see how much of that ongoing work an assistant can take on.

With etc., the starting point is a conversation. “I have this bike. I’d like something for the backyard.” That gives the assistant something to work with. It can ask a question if it needs more detail, look for a reciprocal match, and come back when there’s something worth considering.

You should be able to put the phone down in between.

That’s the part I find most compelling. The input can be small: a photo and a text, or eventually a voice message. You shouldn’t need to learn another marketplace interface or maintain another feed just to explain what you’re looking for. The assistant can keep working on the request after the conversation ends.

I think of this as an assistant-run matching marketplace. You provide the intent. The assistant does the ongoing search. You decide whether the match works for you.

For neighborhood trades, that last part belongs to both people. Finding someone who wants my bike is only half the problem. They also need to have something I want, and we both need to agree that the exchange makes sense. A plausible recommendation is a starting point; a trade both neighbors want is the outcome I’m trying to make easier.

If I can get that experience working well, I want to understand where else the pattern applies.

A job search could start with: “Here’s my experience, here’s the kind of work I want to do, and here’s what matters to me. Find a role that fits.” A home search could start with a budget, a location, and the things someone needs in a place to live. In each case, the assistant would carry the context forward and keep looking, instead of asking the person to repeat the same search every day.

Those are possibilities, not things etc. does today. A job search has different stakes and constraints from trading patio chairs. A home search needs different information and relationships. Each would take real work to get right. What might carry across is the interaction: explain your situation, refine it when necessary, and let the assistant bring back a match you can act on.

For now, neighborhood trading gives me a concrete place to start. The things are familiar. The exchange is understandable. I can focus on whether the assistant finds something both people actually want and whether it saves them effort along the way.

There’s plenty I still need to learn. How much context makes a useful match? When should the assistant ask another question? What should happen when nothing fits? Does the eventual trade feel worth the wait?

I don’t need to build every possible marketplace to start answering those questions. I need to make this one useful.

That’s the direction for etc.: start with the things sitting unused nearby, and learn whether a marketplace can work through a conversation while the assistant handles the search.