24 Hours With Nothing but Artificial Intelligence For Company

I decided to take three AI's for a spin - some idiotic, some less so.

Hi,

Whenever I write about Mister Organ I feel like I am taunting anyone who isn’t in New Zealand or Australia. For those not in those two countries — I’m working on it!

In the meantime — if you’re in New Zealand, tell your friends and enemies to see it as an early Christmas present. It’s still in a shitload of theatres, but won’t be for much longer. So go this week or weekend! We open in Australia on December 10 and tickets for virtual Q&A sessions are on sale now.

This review gets it: The realisation that there is nothing special or unique at all about Mister Organ. That’s the point: He is an archetype of a certain type of human we’re stuck with here on earth that we have to navigate — while attempting to stay sane.

With that in mind, this weekend I decided to escape real humans and spend some time with the various artificial intelligences that have been let loose on the internet.

Over the last year more and more of this tech has become available — first for a select few, but increasingly for free as the machines learn and we all start to get sold on this exciting / bleak / terrifying / terrific (you choose) future we have coming.

First up, I thought I’d try out Chat.OpenAI. It launched last week, and the BigTechnology newsletter described it like this:

“A chatbot that meets the hype is finally here. On Thursday, OpenAI released ChatGPT, a bot that converses with humans via cutting-edge artificial intelligence. The bot can help you write code, compose essays, dream up stories, and decorate your living room. And that’s just what people discovered on day one.

ChatGPT does have limits, some quite annoying, but it’s the first chatbot that’s enjoyable enough to speak with and useful enough to ask for information. It can engage in philosophical discussions and help in practical matters. And it’s strikingly good at each. After years of false hype, the real thing is here.”

I decided that ChatGPT would be my friend, and this weekend I wanted my friend to tell me a story.

I’m in orange, ChatGPT is in green:

The AI seemed happy to have been given the opportunity to write for Webworm: “Sure, I’d be happy to help you out with a story for your newsletter. Here’s a story about a bad man doing bad things,” they replied.

In about .5 seconds, ChatGPT had written this unique story about a bad man doing bad things:

It was a dark and stormy night in the city, and the streets were deserted as the rain pounded down relentlessly. In a dingy alleyway, a man named Jack stood over the unconscious body of a young woman, his breath coming in short, ragged gasps.

He had been following the woman for weeks, stalking her every move and waiting for the perfect opportunity to strike. And now, finally, he had her all to himself.

With a wicked grin, Jack reached into his pocket and pulled out a knife. He had planned this moment for so long, and he relished the power and control he felt as he raised the knife above his head, ready to deliver the final blow.

But just as he was about to strike, a loud voice boomed out from behind him.

“Stop right there, Jack! You’re under arrest for the murder of six women!”

Jack spun around, his eyes wide with fear as he saw a team of police officers surrounding him, guns drawn. He had been caught, and there was no way he could escape.

With a snarl of rage, Jack lunged at the nearest officer, but it was too late. The officers were too quick, and they easily overpowered him, handcuffing him and dragging him off to face justice.

And as Jack was led away, the young woman slowly regained consciousness, feeling a sense of relief and safety as she knew that the bad man who had been terrorizing her was finally behind bars.

I suppose the AI had equated “bad man” with someone who is terrible to another gender — which makes sense when you consider the wealth of public information the AI has access to.

So while on one hand it seems like a misogynistic read, I guess if AI is learning off reality then you can’t really blame it.

It was interesting it had put a happy ending — I wondered what the AI would do if I requested a less happy end — one I’d argue is maybe more true to real life when it comes to consequences.

This is what came out in the second version:

“Stop right there, Jack! You’re under arrest for the murder of six women!”

Jack spun around, his eyes wide with fear as he saw a team of police officers surrounding him, guns drawn. He had been caught, and there was no way he could escape.

But as the officers moved in to arrest him, Jack managed to slip away in the confusion. He fled down the alleyway, disappearing into the night as the police searched for him in vain.

And as the young woman slowly regained consciousness, she realized with a sinking feeling that the bad man who had been terrorizing her was still out there, lurking in the shadows and waiting for his next opportunity to strike.

I chipped in:

We talked for hours. You can too, if you get sick of your colleagues or family this week: chat.openai.com. It really is pretty wild when it comes to storytelling — ask it to write you a screenplay, or a play. I think this is probably how James Cameron wrote his Avatar 2 script.

Next I turned to DALL·E 2 — “a new AI system that can create realistic images and art from a description in natural language.” You will have seen these images everywhere, and maybe had a play yourself if you got into the first round of testers.

It’s recently been made completely public — so I dived in with this prompt. For some reason I wrote my prompt as a story, probably because I’d been prompting ChatGPT like that all day:

Journalist David Farrier hard at work in his office writing Webworm, his newsletter that involves deep dives into the internet, and into bad men doing bad things. He writes on this laptop with his stuffed Goose Heather by his side, Nine Inch Nails blasting from the speakers.

Here’s what it spat out after about two minutes:

The goose is a good fit - as is my mouse and keyboard. The demon versions of me in the background are a little unsettling. Also - is my face melting?

A have a very tight fitting shirt. Strong Clark Kent vibes.

I like that an element of podcasting has crept in, but I do worry about the sound gear I’m using. Also - I appear very sweaty and moist.

Pretty great, although I do worry my face appears to be melting away.

For my last trick, I tried out something MyHeritage.com is touting as a “AI time machine.” It’s a pretty cheap parlour trick that I’m pretty sure the Russians did better in an app — but it is entertaining so here you go.

This is me as a guru —

… and here I am as “A Nobleman”:

David Farrier, prehistoric hunter:

…. and “Ancient Greek Philosopher”:

Arctic Explorer:

…and this is somewhat problematic, but this is me as “A female in 18th Century France”:

If you want to escape reality this week, I suggest you have a play around with some of this stuff. I’m curious what comes out for you.

Finally, back in the real world (possibly as an act of revenge for me pointing out earlier this year I’ve won zero awards during my entire life) the New Zealand Skeptics Society has awarded me with Skeptic of the Year:

One independent journalist has had a series of wins this year, uncovering, highlighting and successfully unseating the leadership of Arise church for their abusive behaviour of less senior church members — as well as pointing out serious issues with other religious organisations such as City Impact church, Bethlehem College and Hillsong. He’s also exposed some of the major issues with cryptocurrencies and NFTs, written a great summary of the Peter Ellis case and the craziness of the Satanic Panic, and detailed the daft fake stories about teenagers identifying as cats. Hence, we are pleased to announce that the winner of this year’s Skeptic of the Year is David Farrier.

I love the work NZ Skeptics does in keeping a sane eye on things, so this means a lot. It came with $250 in prize money, which I have passed onto Rainbow Youth. Paying Webworm subscribers keep me afloat in journalism land, and excess money like this will always go to places like Rainbow Youth.

I hope your week is shaping up okay. Talk soon, and share your AI madness below in the comments!

David.

PS: A dose of reality. Maybe it’s time to go back to real people for a bit?