Because, let's face it, doesn't AI feel a bit like a circus these days? Replete with mayhem, showmanship, shouting, sleight of hand, trick ponies, and whatnot.

Here's a true story, recounted to me the other day by a (human) friend of mine:

One of my students showed me a letter she'd gotten from a guy she'd broken up with. Ultimately he'd confessed that… ChatGPT had written it: He'd prompted the AI with a few details and asked it to write an apology letter.

When she'd confronted him he'd admitted it had indeed been ChatGPT but that he'd touched it up in a few places…

I'm left wondering whether the take-home message from this anecdote is that AI is awesome — or that humans are lazy (and dumb?).

I recently read a short story, written a few years ago (that is, pre-ChatGPT), about an engineer who built a machine that wrote fiction stories and novels. After the company he worked for started making some decent cash off machine-generated stories, the engineer figured out where the big money lay: The company started buying out well-known authors, meaning, their "big" names got attached to machine-generated books, while they got to count the money (which was split with the company).

Eventually, "at least one half of all the novels and stories published in the English language" were created by said machine. The story ended with (presumably) the author, bemoaning:

This very moment, as I sit here listening to the howling of my nine starving children in the other room, I can feel my own hand creeping closer and closer to that golden contract that lies over on the other side of the desk.

Give us strength, Oh Lord, to let our children starve.

AI taking over the world! (At least of fiction writing.)

Now here comes a fun reveal. When I said just now that this story was written "a few years ago" I was somewhat belying the fact that I meant quite a few years ago — 1954 to be precise. Titled "The Great Automatic Grammatizator", the story was written by none other than Roald Dahl (yup, same Dahl from Charlie and the Chocolate Factory). You can read it for yourself if you like.

1954 — and it reads like it was written yesterday.

Score one for human writing!

Everybody now has policies regarding AI-generated material — from Medium to Science:

AI writing is allowed on Medium. However, fully generated AI writing will not receive enhanced distribution, such as our new program "Boost." … Additionally, the free versions of AI writing that people are mostly using on Medium are really bad, and possibly getting worse as it trains on itself. You get what you pay for. [Medium's Policy on Artificial Intelligence (AI) Writing]

AI-assisted technologies [such as large language models (LLMs), chatbots, and image creators] do not meet the Science journals' criteria for authorship and therefore may not be listed as authors or coauthors, nor may sources cited in Science journal content be authored or coauthored by AI tools. [Science Journals: Editorial Policies]

And, boy, the range of opinions regarding where this is all headed… From Godlike humanity to total annihilation, from artificial general intelligence that will be mind-boggling and completely beyond our ken to "nope, they'll hit a wall".

As an AI researcher I've written about various facets of AI here on Medium, but I'm honest enough to admit that mine are just opinions amongst a shitpile of opinions…

Maybe superintelligence is good — then, again, maybe it's bad.

Maybe AI will be moral. Maybe not.

And just maybe — we'll also improve.

"The noblest art is that of making others happy."

We're coming full circle here since that quote is by P. T. Barnum.

Well, AI, let's see how adept you can become at the noblest art:

Make. Us. Happy.