Computer says yes

A family walks into your arrangement room with a settled view of what they want. They seem confident. Detailed decisions appear to have been made. You ask the usual questions, but their answers sound ever so slightly rehearsed. And that's because yours is not the first conversation they've had about this death, it may not even be the second.

Increasingly, those earlier conversations weren't with a person.

People have been turning up to appointments with professionals pre-loaded with conclusions based on internet research for decades. Ask any GP about the amount of time they've had to spend unpicking Dr Google's diagnoses over the last 20 years, and you'll be met with a weary sigh. But people once weighed up different search results and made their own judgements, whereas now, thanks to AI, the paradigm has shifted from search engine to oracle.

When you ask an AI assistant a question you don't receive a list of competing sources to evaluate. You get a single, confidently written answer, delivered in a reassuring tone, like a close friend offering their best advice.

Like a doctor responding to Dr Google, you find yourself gently explaining the reality.

AI doesn't understand grief or emotion, and it doesn't, and can't, care. It produces responses by recognising patterns and navigating to a logical answer, then wrapping it in language that sounds warm and caring. So, when a bereaved person asks a question that's partly practical and emotional, what comes back is a tidy, rational answer wearing a soft jumper.

By the time that family rings you, an emotional situation has been gently rationalised for them. Decisions have been suggested and costs have been estimated, often using generic figures that don't match the reality. Terms like direct cremation or celebrant have been defined in ways that may not match how you use them. The result is that expectations have already been set by an invisible conversation that ultimately does not serve the family well.

Like a doctor responding to Dr Google, you find yourself gently explaining the reality. Your instinct may be to push back hard against misinformation, but families don't want to be told they were wrong to ask a machine, particularly when a machine was the only thing they could face 'talking' to at 3am.

Saying something like: "Before we go through the options, can I ask whether you have already looked into any of this?" lets a family hand you the prior conversation, if they want to. From there you can affirm what they got right, quietly correct what they did not, and uncover the emotional thread beneath the rationalised one.

That's what AI cannot do but you can – offer genuine empathy, understanding, human connection and support.

Pre-loaded families are not a problem to be solved; they're the new normal. Many arrangements now begin before a family walks through your door. The first conversation may have happened with AI, but the most important one is the one they have with you.