My students know all too well what AI corporations are “working on.” They feel worked on, indeed. This past term I asked my students to submit their anonymous thoughts about AI in writing. I wish I could tell you the responses thrummed with defiance. Instead, the mood ranged from resignation to despair. Some read like substance-abuse testimonies.
Anecdotally, I find this aligns with what I’m hearing from college students. There is fear everywhere—fear of AI taking over, fear of being left behind, and a fear of being caught lacking by using your basic human intelligence. I have two reactions to this.
First, this being left behind idea is a myth. There’s always time to catch up with any new technology, otherwise people born today are forever lost. That just makes no logical sense. Showing caution and reason with adoption of a technology doesn’t make you a Luddite.
Second, today’s generative AI is not actually intelligent. It’s a parlor trick of text generation. Yes, it’s powerful and you can do some amazing things with it, but don’t be fooled. It’s not intelligent. Your human intelligence is still—and likely will be for some time, maybe even forever—superior. Don’t neglect building it up and learning to trust your own judgement and intelligence. I can guarantee you the people building AI know this all too well themselves.
This is my advice to the 20-something trying to figure out their first job. To that middle aged person considering school and a career change. To the grandparent who wants to write their first novel. To myself.
Do the thing that most excites you and most terrifies you. That’s it. That’s the sweet spot. Do it because it’s calling to you, especially if you’re afraid. That fear is the tell. It’s your Spidey-sense that this thing is so right for you that to fail at it would leave you devastated and embarrassed.
That’s exactly where you want to be. Go do that thing.
It’s not in our nature to read carefully. Not when reading fiction. Not articles posted online. And not when reading code. There’s this truism in software engineering that no one reads code. Joel Spolsky famously wrote that it’s harder to read code than to write it. That’s true for anything, really. Reading is hard, but we’re in a time now with generative AI where reading has become an essential skill.
I’ve never thought reading code was harder than reading books. Reading is hard. Period. Developers have just gotten away with avoiding it for a long time. We like writing new code because writing gives us the illusion of control. Reading asks something else from us. It asks us to enter into a construction of someone else’s mind, asks us to participate as if it were our own.
That’s true in literature. It’s true in software. When I read fiction, I’m connected to ideas, thoughts, and worlds that are not my own. When I read code, I’m introduced to proofs and algorithms that enforce some program behavior or output. Reading is understanding first, but then, the implications that are less obvious there.
I think about this all the time with Flannery O’Connor. Reading her work was both seeing the South exactly as it was and seeing new possibilities. It was embrace and criticism all at once. That’s a difficult posture to hold. It can feel like holding two opposing ideas in one’s mind. That’s the posture reading generated text and code also requires.
AI generated work is often too verbose. It contains too many abstractions. It oversimplifies and overcomplicates all at the same time.
Reading is reviewing. Reading is making judgment calls. There’s a journey with using generative AI. First, be a great reader. Then shift into editor mode. Review the output. Follow its logic. Notice when the tone is wrong, when the argument is too neat, when the code works but doesn’t belong. People who care about meaning, structure, and clarity will have plenty of work to do here. If machines can generate endless text and code, then discernment matters even more. Reading matters more than ever.
A couple days ago, I came across this lovely piece called Accidental anonymity from Tom MacWright (via Simon Willison.) The post is primarily about AI’s effect on the way people represent themselves when only sharing AI generated content. The specific example was people sending AI generated resumes linking to AI generated work.
I love the larger point MacWright makes, which is this idea that the people who do this:
They haven't put themselves out there. They haven't said anything true.
My personal prejudice here is pretty long-standing. I've written and sung songs, and published my artwork and writing on the internet for fifteen years, and a lot of the reaction I get is about how other people couldn't because they don't know how or don't think theirs is good enough or whatever, but let me be blunt for a brief moment: putting your art, writing, expression out to be judged by others is an act of bravery as much as talent, and a lot of people lack bravery. Sorry to say it but if you need your work to be polished and beyond reproach, that's a determination and character problem, not a skill problem.
“A lot of people lack bravery.” That could be said about so many of us working in corporate environments. Maybe it says more about corporate environments than people, but still, it’s a truth that’s hard to look square in the face. It’s profound to think about AI providing cover for that lack of bravery. But also—putting yourself out there, being brave—I’m confronted by that idea myself.
I do both technical and creative work. It’s easy to see the parallel with creative work. I’ve shared some of my short stories here on this site. In fiction, in feels like I’m laid bare in a way that I’m not in my technical work, but having thought more about this recently because of the above post, I think it’s a mistake to not think of technical work the same way. Sharing technical work online is also an act of bravery. Algorithms are an expression of thinking, and also, no one wants to be caught doing something silly in their code. It would be a shame if we lost that as coding agents become more the norm. Let's all work harder to be brave whether it's creative or technical work.
I’ve not written previously about the similarities I see between belief in AI as inevitable and religious faith in God. I have mentioned the idea frequently to friends and colleagues, and several posts here imply the idea. I’ve been meaning to get around to it, but Alberto Romero has beaten me to it with What Apple Knows About AI That Silicon Valley Won't Admit (via Daring Fireball).
AI is like religion. Either you believe it changes everything, or you don’t believe at all. There is no moderate position; nobody believes in AGI “more or less,” just like nobody is “casually religious.” If God exists, the only coherent response is to reorganize your entire life around that fact, as priests do. If you pray sometimes, then you are just an atheist who’s also fearful. When tech companies spend hundreds of billions on capital expenditures to add sparkly AI features to Office, Gmail, and Instagram, I only see fearful atheists—guys who don’t believe in AI but pretend just in case.
It’s this belief in a soon coming, all-powerful digital god that I see underpinning every decision companies and some individuals are making in terms of the actions they’re taking related to AI. I would have said I see more true believers than casual, fearful parishioners, but Romero sees the opposite. He makes a compelling argument.
What they’re doing has a more charitable name in theology: Pascal’s Wager. They can’t prove AI will be transformative. But if it is and they don’t invest, they’re dead. Or worse: the laughing stock. However, if it isn’t transformative and they invest, they only lose some cash they will recoup anyway by doubling down on ads. So they tithe. They go to church on Sunday. Then fast for Ramadan, keep Shabbat, leave offerings at the shrine. They even perform hecatombs, and, every few months, bathe in the Ganges.
I really hadn’t considered this, but I think I might now agree. Please read all of Romero’s post. It’s a well reasoned argument. I’m not sure I agree on all points. I still think Google and its leadership are true believers. I’m also not sure I agree Apple is completely pure in its convictions. But as a generally accurate argument, I think I’m now convinced. I definitely see the fear Romero describes in my friends and colleagues. I remain convinced that one day the tech industry will come to its collective senses and start behaving rationally again. Until then, I remain committed to reading and writing here in the hope that good arguments eventually win out.
I’m writing this while sitting in the Atlanta airport. We’re headed to London for a family vacation. My younger daughter has been studying abroad in Italy, and we’re all meeting up in London. It’s only 3/4th of a family vacation since my older daughter is on a work trip and couldn’t go. Don’t feel bad for her, though. Her work trip is at a fantastic vacation locale. We’re sort of all vacationing, just in different ways.
All that to say, I’m planning on being mostly offline for the next several days. I’ll post if the mood strikes me or if I do some writing during downtime at a pub.
I’m currently reading Ta-Nehisi Coates’s Between the World and Me. I was looking for a short but deep read after recently finishing Transcription. I wanted a book I could finish in a few days, before I leave for a long trip to London.
I’ve had the book on Kindle for a while but recently spotted the paperback in a local bookstore. I’m glad I picked it up. I no longer like using a Kindle—I’m trying to focus more when I read, one book at a time—and also, it’s such a beautiful paperback to hold in hand.
I’m about halfway into the book now. I knew enough about Coates's work and this book to know what to expect in terms of its examination of American history and what it’s like, as the jacket blurb phrases it, “to inhabit a black body and find a way to live within it.” I didn’t, however, expect the book to also be the story of a man becoming a writer.
The writer, and that was what I was becoming, must be wary of every Dream and every nation, even his own nation. Perhaps his own nation more than any other, precisely because it was his own.
This passage snuck up on me. I’ve stalled out here, savoring these words and thinking through their implications. I hope I finish the book now before I leave tomorrow.
I’m back in an Electric Literature phase. In a bad way actually. I discovered this site not too long ago, when I started back to grad school. The sites available now for creative writing are so much better than when I was an undergrad English major.
We’ve been driving through Pennsylvania for almost two months. People give us stunned looks when we tell them: “How long?” When they ask what brings us through, we tell them we’re traveling children’s entertainers, or we’re land surveyors scoping out sites for a new mall, or we’re searching for our troubled niece Paula, she’s been missing since December, have you seen her? Sometimes they offer to buy us vodka sodas, flagging down the kid slinging drinks at the Best Western Plus.
Short stories come in all shapes and sizes. This one has such a distinctive voice. It’s what I really liked about the story. Also, what’s in the trunk? I think I get it, but I’m not entirely sure. That feeling, persisting all the way to the ending, is what makes this story memorable. It’s a story that is somehow of the present day and literary. It’s one time where not being able to put down my phone was a good thing. I was lost in this story. Highly recommend!
I found this book via a link to it on kottke.org where Jason Kottke also noted its beautiful cover. The way the book was described—literary and “a brilliant meditation on those technologies that enrich or impoverish our connection to one another”—I thought, “How have I never heard of Ben Lerner?” After reading it now, I understand how. The book is as much poetry as story, maybe more so even. It’s the kind of book—the kind of writer in Lerner—that I don’t normally like, but this book has stayed with me many moments after I finished it.
It’s still sitting with me, here as I try to make sense of it, read pages over and over again, trying to under what I just read. It’s a thing to behold and wrestle with comprehending. And that’s the best compliment I can pay it.
I have this habit—let’s call it a bad habit—of regularly changing my social media bio. And I mean regularly. I think I updated it like 3 times just yesterday. Partly, it’s that I don’t know how best to represent myself. I like to think I’m not so easily described. That’s a nice thought. The reality is that my interests and expertise are in constant war with themselves. I’m one part software engineer, one part creative artist, and the two are never at peace with each other. My social media profile is the battle field on which that war is waged.
Sometimes I’m a “software developer working on web and cloud.” At other times, I take a more writerly approach. There was a brief period where the bio was “ I am large, I contain multitudes.” See, I told you I like that idea of myself. Thankfully, I realized the pretentiousness of that and deleted it.
Things would be simpler if I wasn’t like this. If I were purely a software engineer and put this writer’s brain to bed, my career would thank me, probably some colleagues too. It’s hard working for large corporations carrying around Hemingway’s shit-detector. If being a writer was enough, that would make for a simpler description. I’m so envious of those 3 sentence writer bios at the back of a book. My engineering side is way too practical to allow it, though. Immediately, my mind starts rationing my bank balance.
After about the 3rd or 4th change yesterday, I settled on this simple form: “software developer and writer.” It’s direct. It’s true. It’s also the form of myself I’m most at peace with. I really do like being both of these things, fully and completely, and conflicted enough to make life interesting. Now, let’s see how long this version of my social profile lasts.
