The take
An AI system doesn't need to be biased or inaccurate to reshape competition.
It simply needs to consistently recommend a small subset of perfectly reasonable options. Discovery is moving from search results to AI-generated recommendations – the system isn't ranking an existing list, it's constructing the consideration set itself. You could have businesses that are objectively competitive, perhaps even superior, systematically excluded from consideration without any obvious failure in product, pricing, or reputation. We may be approaching a world where machines become extraordinarily efficient at allocating demand without necessarily becoming better at recognising what makes something valuable.
Editorial bio
Rui Vasconcelos spent twenty years acquiring and running resorts. His second career – Altavella, and inside it Concord Passport – is a bet that the next competitive battleground in hospitality isn't ratings on OTAs. It's who shows up in the small set of options that LLM-based recommenders decide a traveler should even consider.
Concord Passport observes what leading AI systems actually recommend when a consumer asks "where should I stay", across models, prompts and contexts, and investigates the evidence they're drawing on. His broader thesis follows from two decades on the operator side: as machines get better at optimising commercial decisions, the human qualities that make an experience worth paying for become harder for algorithms to recognise – and therefore, over time, harder for businesses to economically justify.
Worth following if you want the version of "AI and competition" that comes from someone who's actually run the hotel.
Key facts
- based in
- Portugal
- current
- Founder, Altavella (and Concord Passport)
- prior
- 20+ years leading hospitality businesses – acquiring and running resorts
- focus
- AI recommenders & consideration sets · economics of human experience · competition in LLM-mediated discovery
Featured work
Concord Passport
Observes how leading AI systems construct hotel recommendations, which signals they use, and what competition looks like in LLM-mediated discovery.
Altavella
Adaptive AI frameworks for hospitality that predict customer lifetime value while optimising for both revenue and authentic human connection.
Research on the economics of human experience
On what AI systems can and cannot recognise as valuable.