Man in kitchen with Unox equipment

AI won’t replace your GM: The human-first case for automation

Hospitality leaders explain why the smartest use of AI is freeing staff, not replacing them.

Every hospitality operator has now sat through a pitch for some AI tool promising to fix the industry’s staffing crisis, its margins or its guest experience, sometimes all three at once. Most of these pitches skip past an obvious question: what, exactly, is hospitality selling? Not food. Not a table. An experience, built almost entirely on human judgement, timing and warmth. Any AI strategy that ignores this isn’t a strategy, it’s a shortcut, and shortcuts in this industry tend to get noticed by the people paying for dinner.

The more useful conversation, and the one worth having, is not whether AI belongs in hospitality. It clearly does. It’s where the line sits between the tasks worth automating and the moments worth protecting. A handful of operators who have actually implemented AI in their businesses, rather than just fielded sales calls about it, are converging on the same answer: automate the admin, never the relationship.

The line operators are drawing

Rebecca Yazbek, founder and CEO of Edition Hospitality (Nomad, Rene & Lu, and the Florentino group), draws that line as clearly as anyone in the industry.

“Anything that puts something between you and your guest,” she said, when asked what AI “theatre” operators should walk past. “To me, there’s a lot of AI that talks about improving the guest experience. To me, people improve the guest experience. So AI, for me, is very much a back-of-house tool that we’re implementing so that the people that we employ have the time to do what they do best, which is engage with our guests.”

For Yazbek, that means AI is doing the forecasting, the procurement and the systemised admin work, freeing up the floor for actual hospitality. It’s a filter she applies before adopting anything new: does this tool sit behind the scenes supporting the team, or does it insert itself into the guest relationship?

Her philosophy goes further still. “AI is never going to replace an amazing general manager,” she said. “So to me, the biggest asset we have as an industry is the humans within it.”

Freeing people up for what they do best

Adam Baker, managing director of voice AI platform Otto and a former operator himself, framed the same idea from the technology side. He experiments extensively with AI agent tools in his own business, but says the goal was never headcount reduction for its own sake.

“The reason why I do that is not to replace humans, it’s to replace the need for a human in a role,” he explained. “That essentially would then unlock the ability for me to dedicate that resource to another area, and the area that I dedicate those resources to is humans, to build relationships and to essentially support customers.”

Baker’s read on why hospitality workers burn out isn’t complicated. “I don’t think anyone in hospitality got into hospitality because they like answering emails at midnight,” he said. “I think people get into hospitality because they love people and they love making food and they love creating these amazing experiences.” The opportunity, as he sees it, is for AI to absorb the repetitive, energy-draining admin so staff can spend more time on the parts of the job that brought them into the industry in the first place.

Baker pointed to a real example from the trade show floor: Pepe, who opened a pizza shop in Queensland late last year with one simple ambition, to make pizza and make guests smile. Today he runs the entire venue with one staff member. Kiosks, online ordering and voice AI handle everything else, leaving Pepe free to do the two things he actually wants to do himself: make the pizza and greet the guests. “Everything else just runs on its own,” Baker said.

Yazbek picked up on the same example later, framing it as a question of choice rather than a universal template. “Not everyone wants to go to a restaurant and order themselves and not interact with a human until they pick up the pizza,” she said. “But those who do have the option there. Whereas if you would rather go somewhere else, you can do that.” The point wasn’t that every venue should look like Pepe’s. It was that AI now makes different service models genuinely viable, and operators can choose the one that fits their brand.

The energy audit: a simple starting point

For operators wondering where to actually begin, Baker offered a low-tech first step: an audit, done with pen and paper, before any tool gets switched on.

“Grab a piece of paper and a pen, and do a bit of an energy audit,” he said. “Write down everything you do in a day, everything you do in a week, and then figure out: what gives you energy? What contributes positively to the guest experience? What takes away from the guest experience and drains your energy? And anything that’s in the drain-energy category, whether it be something that’s repetitive, whatever it might be, start doing some research on how you might be able to take away that pain, so that you can focus on what you enjoy doing in the business.”

It’s a deliberately unglamorous exercise, and that’s the point. Rather than starting with a tool and looking for a problem to justify it, it starts with the problem and works backwards, which is precisely where most AI rollouts in hospitality tend to go wrong.

Why the humans are the asset

Dr Nici Sweaney, CEO of AI consultancy AI Her Way, put it best. AI is genuinely transformative, but it has limits that hospitality operators shouldn’t lose sight of.

“AI should be used for what is repetitive and mundane, and steals time, money, resources and joy,” she said. “You should reallocate those resources to places that create connection, because AI is coming for us all… but what it can’t replicate is human connection, because we’re biological creatures. We will always crave those social interactions with each other.”

For an industry facing chronic staff shortages and the highest turnover rate of any sector in Australia, that’s not just a philosophical stance, it’s a practical one. The venues on this panel aren’t using AI to need fewer people. They’re using it so the people they do have are doing the work only people can do: reading a table, remembering a regular’s order, making sure a martini lands ice-cold at exactly the right moment.

As Yazbek put it, describing what her business is really selling: “It’s an hour or two of actually connecting with another human. So the people, to support that, if their time is to ensure that martini is ice cold and on that person’s table as soon as they sit down, that’s the customer experience that we want to be nurturing.”

That’s the real test for any AI investment in hospitality, not whether it’s impressive, but whether it buys your team more time to do the one thing software still can’t: make a stranger feel looked after.


Rebecca Yazbek (Edition Hospitality), Adam Baker (Otto AI) and Dr Nici Sweaney (AI Her Way) were speaking at a Business Stage panel, “Hype vs Reality: The Truth About AI in Hospitality,” at Food & Hospitality Week Sydney in May.

Join our growing community of restaurant decision makers

Unox oven

Follow Us


Copyright Restaurant Business