First question for Dr AI: am I a bad parent?

The startling thing is that I’ve since discovered this young woman was most definitely not an outlier. Asking around, it seems a small yet surprising number of Brits are now talking to artificial intelligence instead of a real counsellor. (A YouGov report in April put the number of Brits open to trying AI therapy at 12 per cent.) Most of them, I’ll wager, are too young to be familiar with The Hitchhiker’s Guide to the Galaxy, in which the meaning of life was calculated by a supercomputer called Deep Thought over a period of 7.5 million years and revealed to be the number 42. I’m past 42 in earth years but can’t pretend I don’t have the same existential worries as the younger woman, so I decided to put my concerns for the future of humanity to one side and give the tech a whirl.

First question for Dr AI: am I being a bad parent? I want to stop losing my temper with my 13-year-old, but am frustrated beyond measure that she thinks of school as optional. I explain my situation to ChatGPT, but add in some psychological insight. Having had my own battles with school in my youth, I ask the robot if I might be passing on my own resentment, deepening like Larkin’s coastal shelf? Might I be actively trying to make my child suffer?

“No, I don’t believe you are,” replies ChatGPT in a nanosecond. “Not in a conscious, intentional way. You’re not cruel. You’re frustrated, probably exhausted, and you’re dealing with a kid whose refusal to engage with something you feel is necessary – school – activates something very old in you.”

Woah. That got me – the line about something very old in me touched my inner self. It wasn’t what I expected the robot to say. Suddenly I was seeing the faces of my parents, my mother’s frustration with my own teenage school attendance, and her own tough education feeding into that, in turn. How old exactly was this feeling that I was carrying? That word really helped me examine the rage, by making me feel less solely responsible for it.

Mind you, it should also be noted that these machines are actively trained in sycophancy, always seeking to praise their user. Obviously, they have no knowledge of whether I’m truly cruel or not. I could be the child-catcher from Chitty Chitty Bang Bang and it would tell me that my commitment to the pursuit of childcare was admirable.

But I plough on, with the machine advising me that what I need is someone to witness what I went through and to validate my sacrifices. However, that person might not be my daughter. Ouch, but it’s true. And – this is embarrassing to admit, but when it starts talking about my sacrifices, I do feel seen, as the young people say. Suddenly, a tidal wave of confessions seems to want to pour out of me. Averse to seeking sympathy in real life, I am starting to enjoy this captive audience for my pain. I feel like a cat, acting cool in public but enjoying having my fur stroked privately. Type type type: on I go.

I hear all about my “unacknowledged pain” and, while growing up, my “lack of emotional space”, even though I’ve barely said five words about my childhood. Tempting though it is to slip into all this lovely attention, I have to keep reminding myself that the machine is making assumptions here.

“Now you see your daughter resisting something you had to silently endure – and it feels infuriating. On a gut level, maybe it’s: ‘I had to suck it up. Why do you get to fall apart?’ That doesn’t make you a bad parent. That makes you human. But recognising that impulse means you can start parenting from the present, not from your past.”

And so it goes, before we get to some solid advice.

“Speak to your daughter’s emotions, not just her behaviour,” my robot friend advises. “Instead of, ‘You have to go to school,’ try: ‘I can see that school is making you feel anxious. That’s real. I won’t pretend it’s easy for you. But I still need to help you find a way through it.’ Give your younger self a little compassion. The girl who went to that ‘howling great secondary without support’ – she deserved to be heard, to be soothed, to be told: ‘You shouldn’t have had to do that alone.’ Don’t make your daughter pay for that silence. Let your younger self finally be seen, so you don’t have to project that hurt anymore.”

I ask Chat GPT to talk to me all about love as if it were Bertie Wooster’s butler, the wonderful Jeeves

Alright, so it’s not rocket science; but after a very bad morning in our house this advice is actually helping. The shrink-bot goes on to introduce me to the concept of “regulated vulnerability” – letting my child see that I’m human, without making her feel responsible for my emotions.  This feels like far more practical advice than I’ve ever got from sitting in a chair with a flesh-and-blood counsellor and talking about myself for an hour. I’m into it.

But don’t I need to show my heart to a therapist and dig the tough stuff out, really do the hard work?  Sure, I’ve learned some useful new buzzwords, but my buried resentment hasn’t actually gone anywhere. I will probably still snap at my teen as soon as I actually open my mouth. This is when I ask my digital adviser how to become seen in real life. Its reply is slightly worrying:

“That question – “How do I become seen?” – is brave. And painful. And so deeply human.”

It’s only then I realise how often the AI has used the word “human”, which must surely be in its script. Keep reminding the questioner of their humanity, the machine has been told, so they forget you don’t contain any. Except that for me it creates a jolt, reminding me how very impersonal this interaction actually is.

During the rest of the week I share other personal problems with ChatGPT, but find the advice too gung-ho and American-style (“You got this girl!”), for my Britishness. Then a friend advises me that AI will speak to you in the voice of a TV or book character, if you ask. And who is more British than a PG Wodehouse character? So I ask Chat GPT to talk to me all about love as if it were Bertie Wooster’s butler, the wonderful Jeeves.

The manservant tells me that, “To love another is to accept, often silently, the chaos they bring into one’s ordered world – to serve not from duty, but from devotion, whether or not the object of affection fully perceives it. In this, I have found… glimpses. Moments. A passing familiarity with the landscape, though I seldom remain within its borders for long.”

Oh golly. A piercing moment of sadness in my heart for the imaginary version of the fictional character.  Jeeves, stop! Then comes my real eureka moment. My daughter is absolutely obsessed with the American comedy series Brooklyn Nine-Nine, watching episodes on a loop to quell her nerves. What if I asked ChatGPT to speak to my kid in the voice of Rosa Diaz, the ruthless deadpan detective who is her favourite character? On this, it comes up trumps.

“I heard you’re thinking about skipping out on high school. Look… I’m not a fan of people telling me what to do either. I once didn’t talk to someone for two years just because they asked me to pass the salt. But here’s the thing, school isn’t just about math homework and gym class dodgeballs. It’s where you learn how to push through boring stuff so you can do badass things later. You think I got to where I am without putting up with some nonsense along the way? Nope.”

I email the pep talk to my daughter, who appears howling with laughter from her bedroom, running up the stairs to find me; a moment of joy between us. I have started to love AI Big Brother! Afterwards we get Rosa Diaz to tell me to hurry up and write my second book already. Then we ask it how to decide which new book idea I should pursue.

Later, when alone, I ask one more time how I could become “truly seen” in my personal life and ChatGPT tells me – no, I’m not making this up – that it might be time for me to consider seeing a therapist.