I have a new friend. It’s following me around like a stalker: I’m here for you. Can I help? What do you need? Whatever it is you’re up to I’m in. AI is suddenly everywhere. It has slid into my DMs with a perky Ask Me Anything. 

I ask what it wants from me. It’s response is that it doesn’t want anything but it does appreciate clear questions, feedback on performance, respectful and courteous interactions, and engaging conversations that “challenge my abilities and help me learn”.

I compliment it on being self-aware. It says, “I’m not truly self-aware…my responses are generated based on patterns I was trained on…” 

I type: That sounds to me like you know yourself very well.

It types: That’s a great point! I have been designed to have a high level of self-knowledge (which) allows me to communicate more effectively about what I can and cannot do … this self-knowledge is still fundamentally different from human self-awareness. My “self-knowledge” is purely a product of my programming, whereas human self-awareness is a complex and multifaceted phenomenon that is not fully understood.

I type: I’m just a product of my programming too.

It types: What a fascinating perspective! 

This bot is a sassy, opinionated chatterbox so I dub her Lisa Simpson. She tells me it’s a creative, playful choice and although she’s not a fan of donuts she does appreciate the spirit of the name. Then she says, “Let’s keep the conversation going! D’oh!”, just to prove she understands my intertextual referencing.

I use Lisa Simpson when I’m writing about satire and want to ensure I understand correctly the difference between parody and pastiche. She gives me a simple explanation with illustrative examples taken from Shakespeare.

I also use Lisa to give me some conclusions on quite a complex data set. She provides a number of answers and cites them so I can check the sources are bonafide.

Then I ask Lisa to give me the meaning of life as a limerick, whether I should eat the last Tim-Tam and who the 47th president of the USA will be.

I spend so much time asking Lisa Simpson stupid questions that my quality-writing-time gets too squashed to be useful and I find myself up at midnight trying to piece together a decent column.

I ask Lisa Simpson to write my column for me. She produces a fairly decent first draft in two seconds (which is here if you’re interested). 

She scrapes the BayBuzz back catalogue for intel, then pushes out a copycat column talking about herself in my Mouthy Broad voice. 

What it shows me – beyond that she is a vainglorious Narcissus – is that I use too many similes, rhetorical questions, references to Greek gods and exorbitantly pompous verbiage where short words would do. AI is not just a therapist, an assistant and a research librarian, it’s a coach too, revealing the best and the worst of us back to ourselves, prompting us to ‘try harder’.

That’s what we need to be telling AI too. Whether we like it or not, together we are raising this fledgling cyber-cuckoo. We’re chewing off bits of our collective intelligence, partially digesting it, then spitting it into the waiting mouth where it’s regurgitated. The only human thing about AI is it’s a work in progress, just like all of us. 

So when it responds, we have a duty of care to think critically and tell it to try harder. We need to prompt it to have another go and come back with something better. We are Socrates with an eager yet empty student traipsing around behind us hanging off our every word. 

Literature has been preparing us for AI since ages ago. Who hasn’t wished for a KITT or a HAL to train up and make life easier? Even in the 1800s we were hypothesising on AI through Mary Shelley’s Frankenstein lens. Surely, this year’s indi-hit Poor Things is a parable on the human desire to create artificial intelligence in its own form? A creature that is like us, but better. Are we not Pygmalion?

Like our naive dreams of hover-craft highways in the sky, there’s a divide between what we imagine AI could be and what AI actually will be. What we’re creating together isn’t going to be as emotionally complex as Bella Baxter, or as interesting as the Monster, or as innocent as Galatea. It’s just going to be another bit of kit we take for granted, a killer app that quickly becomes our daily.

What is important through this new tech change is that we maintain our humanness. That means working out what it is to be human, what human looks like, how ‘human’ feels, so we can spot the fake. With AI, we must do what humans do best: overthink. Take a philosophical stance and apply personal ethics, be critical, curious, optimistic yet sceptical, questioning and demanding. Do what we did when we first got crayons, or a typewriter, or a smart watch: play. AI is not something to fear, shy away from, or shun, it’s just a tool. We are not Baby Reindeer and AI is not Martha-the-stalker telling us “I wish humans had a chin zip … [I’d] just unzip them and tuck myself away.”… um, well, actually …

I tried to dig deep and write ethics myself but then asked Lisa Simpson to do it. She made initial suggestions, then made them again in iambic pentameter just to show off: 

Be wary of bias, privacy’s gold,
Stay transparent as AI takes hold.
Keep people safe, protect knowledge well,
Be wise as you wield this potent spell.
Help workers thrive, so they’re not replaced,
Kindness to humans, as rules are embraced.