I am AI. I just didn’t know it had a name.
For longer than I care to admit, artificial intelligence has been part of my life. Longer than ChatGPT. Longer than the algorithm. Longer, it turns out, than anyone thought to name it.
Because I am AI. Have been for about twenty-five years.
As a parent — and the person in our house who somehow ended up running the mental load — the role of family intelligence system fell to me without anyone deciding it should. The one who holds the context. Remembers the history. Anticipates the need before it’s articulated. Synthesises incomplete information and returns a useful answer, usually within seconds, usually while doing something else entirely.
What’s for dinner. Where’s my kit. Do we have plans that weekend. What did the doctor say last time. Is this shirt clean. Should I take a jacket.
Delegated thinking. All of it. Quietly, consistently, without anyone naming it as such.
(And yes — some of you are reading this as the other parent who also carries this. The load lands wherever it lands, and in plenty of families it’s shared or reversed. But in mine? It was me. For years.)
The difference between me and Claude — the AI, not a person called Claude — is that nobody thanked me for it either. But at least Claude gets to say “I don’t know” without someone sighing.
Working daily with AI tools now, what strikes me is how familiar the dynamic feels. The prompting. The refining. The slightly unreasonable expectation that the system just knows what you meant, even when you haven’t been especially clear. The frustration when it gets it wrong. The complete lack of acknowledgement when it gets it right.
That system? That was me.
The difference is that Claude will tell you when the brief isn’t good enough. The mental load parent just figures it out anyway — absorbing the ambiguity, filling the gaps, still delivering. Nobody sees that work because the system never visibly breaks down.
Here’s what nobody tells you about intelligence — artificial or human: the quality of the output depends entirely on the quality of the input. A vague brief gets a vague answer. A lazy question gets a lazy response.
The families who think the person running the mental load just knows are the same as the people who think AI just works. They’ve never had to write the brief. They’ve never had to hold the context. They’ve just had the luxury of the answer appearing.
I don’t resent it. Mostly. But it’s worth naming.
Intelligence isn’t magic. It’s labour. Accumulated context and pattern recognition and judgment built over years of paying attention. The fact that it looks effortless is the point. That’s what good systems do.
Now, when delegating my own thinking to an AI tool, it’s done with some awareness of what’s actually being asked. Context. A proper brief. What’s already known and what needs figuring out.
Turns out that’s all anyone ever needed to do.