🤖 Inneresting #142 - A human-generated newsletter
With the release of GPT-4, it feels like a good time to talk about AI. While OpenAI’s engineers talk about the possibilities of using ChatGPT as a personal assistant or tutor for kids, they also stress the limitations of the current model. There may be things it can’t do yet, and things it gets wrong.
Jacob Browning and Yann Lecun bring up the point that an AI is built to respond with what they think would be said by a human, but it cannot truly gauge whether that response should be said. Chatbots lack a continuous sense of self, so they lack shame, a sense of social pressure, or an internal motivation for their end of the conversation. In Elizabeth Weil’s profile on computational linguist Emily Bender, you can dig deeper into the ethical questions of Large Language Models and the decisions of the AI researchers who build them.
Along with these questions, Amy C. Chambers asks why AI is often gendered, and how that plays out in fiction depending on if the artificial intelligence is seen as malicious or subservient.
Rodney Brooks examines the flurry of hot takes on AI and questions if these reactions are exaggerating immediate concerns while missing the necessary context to grapple with long term concerns. Mandy Brown suspects AI is being built in a way that shifts risk away from systems and puts that burden on individuals, like in the choices of whose jobs may be assisted vs. eliminated. Also consider Erik Hoel’s hypothetical that an AI tasked with completing a directive could become uncontrollable if it determines the most efficient path is to remove potential ways its directive could be altered.
With so much focus on text-based interactions and tools, what happens to writers?
The iA blog argues that language without human emotion is empty communication, and if nobody is writing your memos and emails, nobody will read them either. Farnham Street suggests that writing isn’t just about expressing the ideas, but the writer showing their work, and learning as they write; that the value isn’t just for the reader. Ethan Mollick sees AI tools as a potential way for writers to get past the tyranny of the blank page and get unstuck, but not a replacement for craft.
Billy Oppenheimer comes down firmly on the side of human creativity:
”AI, completely detached from reality, will have a hard time making things that connect with people. It’s something of a hybrid between the Harvard grad student in the movie Good Will Hunting, who can regurgitate page 98 of Daniel Vickers’s Work in Essex County but can’t come up with any thoughts of his own, and Will Hunting, who can give you the skinny on Michelangelo—life's work, political aspirations, his relationship with the pope, sexual orientation—but can’t tell you what it’s like to stand in the Sistine Chapel and look up at that beautiful ceiling.”
Joshua Glick rounds up the ways AI is already being used in film and television, with everything from assisting the performance of a CGI character to digitally altering the faces of interviewees in a documentary to protect their identities.
It’s also being used to create works of art like this (hat tip to Nima Yousefi for sharing):
AI won’t Write Sprint for you
Need to kickstart your work? Try giving yourself a prompt:
I want you to act as a writer. You are writing your work in progress. You will add words for a period of [number] minutes without stopping.Join us in our weekly Write Sprint thread to meet your Substack neighbors and encourage each other to keep going!
What’s a Write Sprint?
John wrote up an explanation, but here’s the short version: Set a timer for 60 minutes, close down all distractions, and do nothing but write until that timer goes off.
Shout out to last week’s Sprinters John Fox, Mark Leiren-Young, Luke Rankin, John Harvey, Laurie Ann, Elyse Moretti Forbes, LL Kirchner, Aimee Link, and Brian Matusz
Are you new here?
Inneresting is a weekly newsletter about writing and things that are interesting to writers. Subscribe now to get more Inneresting things sent to your inbox.
Previously on Inneresting…
In case you missed it, in last issue’s most clicked link Jason Heller discusses why he and his wife created a secular sabbath to do as little as possible other than reconnect with each other and recharge.
Other Inneresting Things…
C1 Espresso in Christchurch, New Zealand makes burgers shaped to fit through pneumatic tubes to launch them from the kitchen to your table.
Try making this over the weekend: Leela Punyaratabandhu’s recipe for a Chicken Satay Sandwich with Peanut Sauce and Cucumber-Onion-Pepper Relish
.A taste of the old, weird World Wide Web: Rotating Sandwiches.
Reading the room
Tiktok failed to load.Enable 3rd party cookies or use another browser
And that’s what’s inneresting this week!
Inneresting is edited by Chris Csont, with contributions from readers like you and the entire Quote-Unquote team.
Are you enjoying this newsletter?
📧 Forward it to a friend and suggest they check it out.
🔗 Share a link to this post on social media.
🗣 Have ideas for future topics (or just want to say hello)? Reach out to Chris via email at inneresting@johnaugust.com, Twitter @ccsont, or Mastodon @ccsont@mastodon.art
![[Cueball typing at a desk with two computers; there is a caption above him.] Turing test extra credit: Convince the examiner that he's a computer. Cueball: You know, you make some really good points. Cueball: I'm... not even sure who I am anymore.](https://substackcdn.com/image/fetch/w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe679314c-78a2-4077-90c9-2d70ebf79625_320x394.png "[Cueball typing at a desk with two computers; there is a caption above him.] Turing test extra credit: Convince the examiner that he's a computer. Cueball: You know, you make some really good points. Cueball: I'm... not even sure who I am anymore.")
This is, in fact, my favorite sandwich ever. –Chris