What if AI could spot circular opportunities before we do?

It started with a house in France.

My partner and I have our eye on a property there, but we're not quite ready to make a move. So every now and then I found myself going back to the listing, checking whether it was still there and whether the asking price had changed.

Until I casually asked ChatGPT whether it could keep an eye on it for me and alert me if the price dropped somewhere online.

Apparently, it could. It could keep checking at regular intervals and only come back to me when something actually changed.

For some reason, that completely blew my mind.

Not because checking the price of a house is particularly revolutionary, but because I suddenly realised I had been thinking about AI almost entirely as something you actively ask to do things. I ask a question, it gives me an answer. I ask it to research something, it does the research. I ask it to help me write something, it helps me write it.

But this was different. I could essentially tell it: keep your eyes on this for me, because I don't have the time or attention to do so myself.

And that immediately made me think about May Again and a question I keep coming back to: why is keeping products and materials in circulation still so incredibly difficult?

What if we're not short of circular opportunities, but simply missing them?

Buying new has become remarkably frictionless. When we need a chair, a table, a light or a material, we search for it at exactly the moment we need it. We find specifications, quantities, prices and delivery times, place an order and somebody takes responsibility for getting it where it needs to go.

Keeping existing objects in circulation requires something very different. We need to know what already exists, where it is, who owns it, what condition it is in, whether it can be repaired, whether somebody else could use it, whether that person needs it at roughly the same moment and whether moving it from one place to another makes practical and environmental sense.

I've written before about the friction this creates. But what if part of the problem isn't simply friction?

What if circularity also has an attention problem?

A hotel announces a renovation today, but its furniture might only become available eight months from now. An upcycling studio might be looking for fifty identical chair frames six months later. Somewhere nearby, another hospitality project may be preparing to open and need exactly that type of furniture a few months after that.

Those events probably won't happen on the same platform, they won't be described using the same language and they certainly won't arrive neatly in someone's inbox at exactly the right moment.

To us, they look like unrelated pieces of information spread across time and place. Yet put them together at the right moment and they could become a circular chain.

The opportunity was there all along. Nobody saw all the pieces at the same time.

What if something could keep watching?

That is where my ridiculously simple house-price alert suddenly became much more interesting.

What if AI could continuously look for the signals that tend to precede circular opportunities? Not just objects that are already being offered for reuse, but hotel renovations, office moves, refurbishments, closures, new hospitality projects, procurement calls, designers looking for specific materials, companies dealing with surplus furniture or organisations announcing circular ambitions.

A hotel announcing a major refurbishment could be a signal that a significant amount of furniture may become available several months later. An upcycling studio looking for a particular type of frame or material could be matched against those future streams. A new hospitality project opening nearby might become a possible next context for objects that haven't even left their current one yet.

And that last part is particularly interesting to me.

Most circular systems start looking for solutions when something has already become surplus. The renovation has happened, the furniture has been removed and suddenly someone has a warehouse full of objects that need somewhere to go.

But by then we have already lost something incredibly valuable: time.

There are often signals long before an object becomes waste. Renovations are announced, architects are appointed, planning applications appear, companies communicate about relocations and hotels talk publicly about upcoming refurbishments. At the other end of the chain, new projects are being announced, designers are looking for materials and organisations are publishing circular procurement ambitions.

What if we could start connecting those signals before the physical objects ever have to move?

Instead of trying to find a destination for waste, perhaps we could start identifying potential future availability.

Imagine knowing that a hotel is likely to renovate in eight months and that its current interior contains furniture that may become available. At the same time, a system has noticed three organisations within a reasonable distance that are planning projects where similar objects could potentially be used.

That doesn't make anything circular by itself. Someone still needs to inspect the furniture, decide what is worth keeping, talk to the owners, work out the logistics, repair or redesign pieces where necessary and take responsibility for what happens next.

But suddenly we know where to look, and we know before it is too late.

And then I started looking differently at our May Again pilot

With May Again, I'm currently exploring a small part of this puzzle together with a small group of independent designers. We're experimenting with giving physical objects a persistent digital identity, inspired by the Digital Product Passport but deliberately looking beyond compliance.

The idea is that information about an object doesn't disappear every time its context changes. Its origin, materials, previous locations, repairs and experiences can travel with it, while future owners or users can add new chapters to its story.

Until now, I have mostly thought about that identity as a way of preserving context. If we stop losing information about objects, perhaps we can make better decisions about them later.

But suddenly I started wondering whether preserving that information is only the first step.

Because once an object has a persistent identity, that information could potentially start helping us determine what happens next. Its dimensions, materials, condition, location and history aren't only useful as a record of where it has been; they could become signals that help identify where it could go.

An object's identity could eventually help connect it to a repairer, an upcycler, a material need or another location where it might fit. It could signal when the object becomes available. The system could start looking for possible next contexts before anyone has even decided to throw it away.

That takes the question behind a Digital Product Passport somewhere I find much more exciting.

What if an object's identity didn't just document its life, but actively helped extend it?

AI could become part of the missing infrastructure

One of the biggest barriers to a functioning circular economy is that the processes required to make it work simply aren't there yet. We talk about marketplaces, logistics, traceability, product information and reverse supply chains, and all of those things matter.

But perhaps another part of that infrastructure is much less visible: attention.

Thousands of small signals are appearing every day: an object becoming available, a building preparing for renovation, a material being needed, a designer looking for a particular waste stream, a procurement brief appearing, a repair service opening nearby or a product reaching the end of one context without reaching the end of its useful life.

No person can keep watching all of that, especially when the connection between two pieces of information may only become relevant months after the first one appeared.

AI can.

Not because it knows what deserves to be saved or because it should decide what happens to our objects, but because it can keep looking, comparing and remembering until two seemingly unrelated signals suddenly become relevant to each other.

And then, essentially, tap us on the shoulder and say: you might want to look at this.

So I'm going to test it

All of this came from discovering that I could ask AI to keep checking whether the price of a house in France changes.

Now I want to see what happens if I put that same ability to work for May Again.

Could it keep watch for upcoming renovations that might create future furniture streams? Could it find designers and upcyclers looking for particular materials? Could it spot new projects that might become next contexts for existing objects? Could it monitor research, regulation and procurement opportunities and tell me only when something appears that is relevant to what we're testing?

And eventually, could the information attached to an object itself become part of that search?

I don't know yet.

That's exactly why I want to test it.

Because perhaps one of the things the circular economy needs isn't another place where we list things after they have become surplus.

Perhaps it needs a way of noticing the connections before we lose them.

Image credits: Photo by Cam Ballard on Unsplash