The two-year anniversary of ChatGPT

The Medium Newsletter
The Medium Blog
Published in
Sent as a

Newsletter

4 min read6 hours ago

✌️ Welcome back
Issue #221: your thoughts on AI-generated writing + a tour of Parisian bookstores

ChatGPT is as old as some of my friends’ kids now. It turned two last Saturday, and I remember exactly where I was when it was announced to the world. We were sitting in a friend’s living room in San Francisco talking about the movie M3GAN, which was premiering around the same time (the trailer was everywhere). It stars a ventriloquist doll who develops human intelligence and stages a murder-suicide. (Fun!)

Weird how culture converges around a theme sometimes.

I was writing some marketing copy for Medium, and I was working so hard on it my friend was just like, Why don’t you ask ChatGPT what Medium is? She did and it… did an okay job! I didn’t use the copy because it sounded boring and was wrong in subtle ways, but it was unsettling — and a little awe-inspiring — to see how capable this tool already was. (No, we do not use ChatGPT to write this newsletter.)

Anyway. Whether you’ve used it to code, play D&D, or peer into your fridge and tell you what to cook (or not) it’s arguably the most important invention of the last decade. ChatGPT’s monthly userbase grew to 100M within two months of its release, making it the fastest-growing app in history (TikTok took nine months to reach that milestone; Facebook took four years).

The technology is not as new as it feels. GPT = “generative pre-trained transformer,” a fancy term for a system that converts text into numbers and maps the probability of a character string appearing next in a sequence. OpenAI has been working on these models since 2017. Rudimentary language training models based on the same principles date back to the ’60s. The first one was created in 1966. It was called Eliza and built to sound like a therapist.

So, OpenAI didn’t invent something new — they improved upon something half a century old. “It’s not a fundamentally more capable model than what we had previously,” one OpenAI researcher admitted last year, but the team’s meticulous attention to how it might be used made it take off. The interface is intuitive; it guides you toward using it successfully. GPT aside, I think there’s a lesson here about the value of attending to details.

On Medium, Georgia Tech professor Mark Riedl explains how it works and shares a tip: Don’t anthropomorphize. Humans have an innate tendency to see humanity in all things, but LLMs are not human and you’re not “having a conversation” with them (journalists keep getting confused about this).

Harris Sockel

Your thoughts on writing and AI

On Monday, we sent a newsletter on what AI means for the future of writing. Many of you had strong feelings on this topic! Here are a few responses that stood out to us.

Andrew Wagner invokes an Apple Intelligence ad:

[This newsletter] reminded me of David Ogilvy’s assertion that “People who think well, write well” (wonder if Paul has been reading some Ogilvy : ) Also reminded me of this Apple Intelligence ad which perfectly sums up one of the scariest things about AI — to me anyway. If Ogilvy is correct (and I think he is), that people who think well, write well, then why do we want to create tools that mask bad thinkers (like Warren, in the ad)?

Mollie Cora disagrees with the premise that a world where you don’t need to write = a world where you don’t need to think…

I would strongly disagree with this statement. I know enough people in the STEM fields to know not everyone needs regular reading and writing to think critically.

…and jotisPC believes that human progress depends on skilled storytellers (i.e. writers):

Human progress is crafted by those who can write. Or tell stories. And at the base of that is the improvement of thought that comes from writing or storytelling. […] Imagine if the (US) Constitution had been written by the non-thinkers of the United Colonies instead of debated and crafted by those who were disciples of classical thought. While the masses have been successful revolutionaries at times in history, unless they had thinkers who could communicate at the helm, chaos was waiting not far down the road.

📖 A good sentence I highlighted recently

“A bookstore isn’t like most stores, where you walk in, hand over money, and walk out. It’s more like a contemporary museum, with artifacts in the form of books.” — photographer William Sidnam, via his visual tour of Parisian bookstores

Deepen your understanding every day with the Medium Newsletter. Sign up here.

Edited and produced by Scott Lamb & Carly Rose Gillis

Questions, feedback, or story suggestions? Email us: tips@medium.com

Like what you see in this newsletter but not already a Medium member? Read without limits or ads, fund great writers, and join a community that believes in human storytelling.

--

--

Responses (5)