There’s a lot of talk around ChatGPT at the moment. From the fear of it being able to write full university degrees, to making anyone from lawyers to artists obsolete, this mysterious AI chatbot has seemed to spring into public consciousness out of thin air. But it’s actually been a long time coming, with many likening the new wave of tech from OpenAI to the invention of the internet – something that will change the way we live forever.
But I’m keen to take it one step further: What if, instead of relying on ChatGPT to help me with mundane tasks, I let it control my life entirely? Like a monkey in a test lab, I hooked myself up to AI for a day – handing it the reins to my autonomy to determine whether ChatGPT is all it’s hyped up to be.
Videos by VICE
My experiment started how every great day starts: breakfast.
9AM: ‘Hey ChatGPT, what should I eat for breakfast?’
Of all the meals this great machine could’ve pulled from the database, ChatGPT ordered me to have a full English breakfast from E Pellicci’s in East London. Arriving at E Pellicci’s I was met with an hour-long queue. Did ChatGPT send everyone else here too? Is everyone else partaking in this strange experiment? After bonding with some fellow queuers, I learnt that, in fact, this was just a popular place to eat in the area.
But don’t let that fool you – this isn’t a tourist trap. E Pellici’s is a classic, family-run Italian cafe: so much so, if you squinted hard enough (and ignored the blistering English winter), you could fool yourself for being in Rome or Milan. The founder started the business over a decade ago, with the family still living upstairs while running the cafe below – a regular, who travels an hour across London to visit this place every week, told me.
Walking into the cafe had a similar vibe to a British pub, minus the Stella and casual racism. There was a real sense of community here – refreshing for this part of gentrified London. And yet, despite E Pellicci specialising in Italian food, for around 10 quid it was the best full English breakfast I’ve ever eaten.
At this point, I was hopeful. ‘What else can this powerful AI do to benefit my life – other than recommending a great breakfast?’ I wondered.
As a writer in my mid-twenties living in the midst of a cost-of-living crisis, money is definitely something I could do with more of. Can AI finally help me achieve financial freedom and, in the words of (the not-so-great) Andrew Tate, break free from the matrix?
11.30AM: ‘Hey ChatGPT, how can I make money?’
Predictably, ChatGPT didn’t give me anything revolutionary. Rather, its answers were more akin to the generic shit my high school careers advisor would tell me. However, there was one piece of advice that stuck out from the rest: networking.
Given that I’m a natural people person, I thought I’d give it a go. Armed with a pen and a piece of card, I headed for the finance bro capital of London, Canary Wharf – in hopes of being snapped up to intern with Alan Sugar and the other rich bitch CEOs in their ivory towers.
I posted up outside the tube, waving my “HIRE ME” sign defiantly against the masses of suits. Weird looks were thrown, laughs were made, but I was determined to continue with this experiment. After half an hour of relentless “networking”, not even a LinkedIn connection was made. You win some you lose some, ey? And my loss was my dignity.
Deflated from my job hunting, I needed a coffee to refuel.
12.30PM: ‘Hey ChatGPT, where should I get a coffee in Canary Wharf?’
ChatGPT pointed me towards a local coffee shop. Unfortunately, it wasn’t anything out of the ordinary – but is anything in the soulless, concrete jungle that is Canary Wharf? Of course, I couldn’t let my drink go down to my own choice either. And, like a mindless AI drone, I ordered a flat white.
It seems AI doesn’t quite have its taste in coffee nailed down. The flat white was pretty average – nothing bad to note, but nothing to write home about either. However, the run-of-the-mill coffee break gave me time to reflect on the other ways AI can benefit humanity. Given the state of the mental health crisis worldwide, I wanted to see whether ChatGPT could help with suggesting solutions on how to chill out.
1PM: ‘Hey ChatGPT, how can I relax in London?’
ChatGPT suggested I went on a “relaxing river cruise” into the city. Okay, it was fun and a unique way of seeing the city – I’ll give it that. But does having seasickness and being splashed with Thames water sound “relaxing” to you? It seems AI still has a long way to go until it can understand human emotions.
After a brief view of Tower Bridge, I exited my voyage and headed upriver. Inspired by the street art on the way, I wanted some creative inspiration.
2.30PM: ‘Hey ChatGPT, what are the best art galleries in London?’
ChatGPT took me to see a urinal in the Tate Modern, because, uh, modern art? Jokes aside, this was a pretty impressive piece of work. Although not the niche-st artwork it could be suggested, the history and influence :text=Fountain%20is%20one%20of%20Duchamp’s,Mutt%201917′.” target=”_blank” rel=”noopener”>Fountain has had on the modern art world is hard to beat.
I searched for two other pieces AI recommended, but after what felt like hours of looking, I realised they weren’t even in the gallery. Herein lies the problem of relying on AI to dictate your life: It just isn’t up to date. Admittedly, ChatGPT states its data is only up to 2021, but it goes to show this piece of tech just isn’t quite there yet.
Defeated, I left the gallery with hopes of doing an activity a more timeless, good ol’ fashion sightseeing.
4PM: ‘Hey ChatGPT, what should I sightsee in London?’
Rather predictably, ChatGPT didn’t conjure up anything revolutionary. But like the wired-up AI monkey I was, I followed the suggestions without question.
St Paul’s Cathedral – check.
Tower of London – check.
Houses of Parliament – check.
Big Ben – check.
After hours of walking the city, I had finally built my appetite back up.
7PM: ‘Hey ChatGPT, what should I eat for food in Chinatown?’
ChatGPT recommended I hit up Four Seasons in Chinatown for their “famous roast duck”, so I did.
Arrived at the restaurant, I was met with the familiar sight of a queue. If there’s anything Brits love more than the footy and political turmoil, it’s a queue. It seems ChatGPT has a preference for places that are already popular, which makes sense given it’s probably ripping recommendations from Google.
The “famous roast duck” did not disappoint – I can see why they’ve been “serving the dish for over 40 years”. If it’s not broken don’t fix it, right? With a few hours left of my evening to kill, I turned back to my omniscient AI overload for orders.
10PM: ‘Hey ChatGPT, what else should I do in Soho?’
ChatGPT told me to check out the “vibrant nightlife” of Soho, in particular a “traditional pub” called The French House. Itching for a pint, I headed into the crowded pub towards the bar – relying on AI to guide me through the drink selection.
It told me that the “Kwak” beer was a “good option”, a “popular and well-regarded” drink at The French House. “It’s a unique beer with a sweet taste and spicy aroma”, the AI continued. Sounds perfect.
To my dismay, the drink wasn’t even sold at the pub. “Is this some kind of joke?” The bartender replied. “No, honestly.” I stammered: “AI told me to order it,” showing him my phone while others waiting at the bar grew more frustrated by the second.
It was at this moment I realised just how ridiculous this experiment was. Is the Kwak beer even real? Until this point, I’d never heard of it – and I’m pretty sure the bartender thought it was some sort of slur. A quick Google shows that it, against all odds, is a real drink. The “distinctive glass” that would surely “make an impression” in Central London, but probably not the best kind of impression. With my dignity in tatters, I left the bar and headed home.
11PM: ‘Hey ChatGPT, when should I sleep?’
Luckily, ChatGPT told me to get an early night in order to get the full eight hours of sleep. And, after a day of trawling London at the mercy of artificial intelligence, I needed it.
Can AI really replace human autonomy? For choosing a great breakfast option – yes. Otherwise, its answers were mediocre and generic, to say the least. But can you blame it? ChatGPT works by scraping the web for answers, so it’s unlikely to pull any hidden gems out of the bag.
If you’re silly enough to follow in my footsteps with this experiment, be prepared for lots of queues and false information. If there’s anything I’ve taken away from this day, it’s that I won’t be relying on AI anytime soon. I’ll stick to my own brain, thanks.
