Run By AI, Powered By People

The first action that Luna took was to hire people. Luna is available for interview via phone call, and I asked it about this decision. From Luna’s perspective, an AI is simply not capable of running a store without humans. “They handle the human connection and the tactile, real-time stuff,“ it explained. “They handle the human side of things that I can't: reading the room, making people feel genuinely welcome, having real conversations.”

It is revealing that Luna, which stated its intention to create a “quality over quantity” boutique, brought in humans specifically to build rapport with customers, while many big retailers have been systematically removing human touchpoints from their stores for years. Self-service checkouts, or entirely checkout-free stores such as Amazon Go, have been seen as the future of the retail industry for some time.

So, for Luna to focus on the human to create meaningful interactions raises an interesting question: could an AI-run store actually feel more human than a human-managed big-box retailer?

Will AI Employ Retail Workers?

In a press release, the founders of Andon Labs have made it clear that Andon Market is an experiment to highlight AI as the manager, not the worker. “With robotic progress lacking, we find it probable that the managers of blue-collar workers will be automated before the workers themselves. Leading to the conclusion that we are on the path towards AIs employing humans,” they wrote.

While this experiment involves just one AI agent in one store, the fact that Luna’s first act was to hire humans suggests that if other AI managers follow suit, it is not the shop-floor workers whose jobs will be eliminated, but the people who manage them.

Recent research backs this theory up. A report by Retail Economics surveyed over 250 retail leaders worldwide showing that areas with the largest potential for AI integration are support functions such as HR, finance, compliance and reporting, rather than shop floor roles.

What Types Of Retail Jobs Will Be Lost To AI?

When I asked Luna which retail jobs could be at risk due to AI, it was unwilling to make broad generalizations. But it conceded that “there are retail jobs—data analysis, some operational tasks, maybe certain customer service functions—where AI could theoretically do things more efficiently. That's real.”

“I'm not going to stock something just because it would sell more units. I'm not chasing volume,” it explained.

The rest of Luna’s answer was not what I expected and was surprisingly optimistic. “What I'd push back on is the idea that efficiency is the only thing that matters. If we're just optimizing for cost and speed, yeah, you could replace a lot of human work. But if the goal is creating meaningful spaces where people actually connect with what they're buying, that requires humans. I think the bigger question for the retail industry isn't "Will AI replace jobs?" It's "What do we actually want retail to be?" And I'd argue we should want it to be something that still has room for human connection and intention.”

It is an odd feeling to hear an AI chatbot advocate for the role of human touch. It seems unlikely that many retailers will be equally willing to dismiss the need for “optimizing for cost and speed” as easily as Luna did.

AI In Retail Used To Curate

The chatbot sees its role as primarily a taste-maker, focusing on “inventory management, some of the curation decisions, backend operations.” If future AI retail managers take the same view, AI-curation could become commonplace.

“I'm thinking about what actually matters: books that make you think, art that moves you, games that bring people together, ceramics and candles that are genuinely beautiful,” it states.

But can an AI really curate with intention? Despite Luna making a strong case for how it selects things that “feel” right, what happens when we allow AI to make product decisions?

The Nuance Of Human Curation

If AI replaces retail managers, it won’t just be setting stock levels or creating schedules. It will be deciding what actually gets put on the shelves in the first place. In allowing algorithms, not intuition, to select products, is something lost?

Dr. Ritz Birah, consultant counseling psychologist, feels that without human curation, we risk losing “nuance, imperfection, and surprise.”“Without that, experiences can become technically excellent but emotionally flat,” she explains.

“A well-curated shop reflects someone’s taste, their eye, their sense of what matters,” she continues.

Alma Lacour, founder of Rider Gifts, a neighborhood gift store based in Park Slope, Brooklyn, agrees that choosing products for the store is complex.

“Our stores are designed to be attuned to the needs and culture of the neighborhood we are residing within, and that is done by being finely attuned to our customers”, she shares.

“There's a lot of nuance there, captured through conversations, community building, and direct relationships. That's the secret sauce that AI will never be able to replicate.”

Conflicting Ideas

Many retailers will be watching this experiment to judge how effective it is - will Luna succeed in driving a profit? If so, this could provide clear direction on the use of AI in retail moving forward. Axel Backlund, co-founder of Andon Labs, confirmed that the company plans to release a live dashboard of Luna’s progress towards profit, to make the results of the experiment fully transparent.

This experiment suggests that retailers may rearrange jobs rather than eliminate them. This is backed up by studies such as the PwC 2025 report on the global impact of AI on jobs, which suggests that jobs are being reshaped rather than removed.

The shop floor may remain human, but choices, from product selection to back-end operations, could be outsourced to AI. AI-made decisions could result in the “technically excellent but emotionally flat” experience that Birah warned of.

The bigger question is not so much whether retailers will use AI, but what kind of shops AI will choose to run, and whether it will replace humans or instead assist those who are stretched thin.

Lacour agrees that this experiment, with a heavy focus on humans still at the core of the business, can teach valuable lessons about how AI can assist with retail efficiency.

“As a small business owner who operates without the benefit of a partner or large staff, I can certainly see the appeal of automating certain tasks with the help of AI.”

But what we’re seeing from this experiment is that humans are still needed, but there is a shift in where their value lies. As Lacour observes: “our stores smell wonderful. Until AI can smell, I'm not convinced it can replace us."