Being the Human in a World of AI
How to Do Group Work with Machines
Have you seen the Lensa AI app? It sucks up some of your selfies and then, in a matter of fifteen minutes or so, churns out half a hundred aestheticized renderings of your face. Unfortunately, Lensa’s artificial intelligence learns its style from artists who go unacknowledged and unpaid. Also unfortunately, the app sometimes sexualizes its clients. In my case, it simply suggested that I stop shaving my head.
I probably wouldn’t have paid $3.99 to try out this app, if AI hadn’t been so much a topic of digital chatter over the past couple of weeks. Folks are especially worked up about how AI creates images and produces prose. It’s called “generative AI,” and Atlantic writer Derek Thompson labels it one of the “10 Breakthroughs of the Year.”
Before I suggest what generative AI might mean for rising professionals, let me acknowledge the two questions I’m ignoring in this week’s Mode/Switch.
Will artificial intelligence kill humanity? It’s a dystopian query beloved by Gen Zs, but it’s a question for which I have little to say. This robot, though, has thoughts.
Will AI kill your job? It’s a live query for people doing work right out of college. The bots are likely to show up at the low end of the labor pool. Again, I’ve got nothing to say that hasn’t already been said a hundred times. (Here’s an aspirational answer I respect, as you’ll see below.)
But, here, finally, is the question I do address this week: given that rising professionals are doing a lot of group work with machines these days, how can they collaborate with the tech—without becoming tech themselves? Recent AI-related journalism focuses answers on two communicative practices: reading and writing.
Reading at Work
Most of the 47 early careerists I’ve interviewed for my latest book do a lot of reading. They have to, working as they do across the knowledge economy. But for all of us, and not just rising professionals, reading at work entails skimming and scrolling. The skim-scroll conducts Kathryn Hayles calls hyper attention: “a cognitive mode that has a low threshold for boredom, alternates flexibly between different information streams, and prefers a high level of stimulation” (12).
Sometimes the skim-scroll generates silly ideas like Sam Bankman-Fried’s claim, “I would never read a book.” (Is it just me, or does that sentence sound like clickbait generated by a frat-bot?) Admittedly, book-reading probably wouldn’t have kept SBF’s crypto-fortune intact. But understanding how he lost so many billions in a day requires more than the six-paragraph blog post.
How can you make reading a vital part of your work? In the early 2000s, Richard Lanham wrote The Economics of Attention, which commended “oscillatory attention.” But quickly shifting attention isn’t productive if you have 32 tabs open on your browser. Instead, you should cultivate what Maryanne Wolf calls the “biliterate brain.” Bi-literacy means, in part, that you read six-paragraph blog posts and that you “bookend” your day by reading 20 minutes morning and night.
An early-career professional told me just this week, “I read so slowly.” Most of us do, because as Wolfe says, there’s nothing natural about reading. But practicing the slow, deep work of reading can be what L. M. Sacasas calls a “counter-practice.” Counter to what? To the algorithmically driven, artificially intelligent media streams that burn out your brain. (Anne Helen Peterson has suggested recently that one good way to treat burnout is to read books.)
Writing at Work
AI has also made the news this week for generating not-terrible prose. In response, Stephen Marche pounded out an Atlantic piece about how the arrival of ChatGPT means that “The College Essay Is Dead.” I logged onto ChatGPT to check out this evil bot for myself. If you’re hoping for Ultron, you’ll be a little disappointed to find that bot looks like this:
I’ll admit to some sub-human snark: I asked what the bot finds hard about professional writing. The reply came in candid, if gaseous prose:
One of the problems with chatbots that generate professional writing is that they often lack the ability to truly understand the context and nuances of the topic they are writing about. This can result in writing that is superficial and lacks the depth and insights that are typically expected in professional writing….
I’ll stop the quote there. But as it turns out, the chatbot had a lot to write about why chatbots aren’t good writers. Things were getting too meta for the Mode/Switch, so I pulled the plug on my collaboration with ChatGPT. But still, I’m not sure that we can avoid writing with generative AI—or that we have to. The trick is not to compete with robot-writers; the trick is cooperate with machines in a manner that maintains good human prose. Generative AI will improve its writing skills. The question is, will you? By which I mean, will you learn to write in your own reliably personable voice?
For starters, you might try writing for some minutes everyday without a keyboard. This week, I was in a marketing meeting, fighting against the voyeuristic urge to stop looking at the screen up front and to stare at the page my colleague was scribbling on next to me. The carbon pencil was mesmerizing. After the meeting, I talked with him about his practice. He’s no Luddite. He’s a media producer. He loves digital tech. But he’s writing longhand to keep being who he is amidst all the AI.
The Moleskine in the meeting has it downsides. For one thing, you can fill notebooks with scribbles that you can’t find again. For another, your hand hurts. After a day of handwriting, I find myself finger-flexing like Mr. Darcy. But David’s advice for your aching hand is rather like Wolf’s for your distracted brain: write more slowly.
Artificial intelligence is too generative, too resourceful, too pervasive for us humans to keep it out of the big group project that is the late-modern workplace. But keep your wits about you. Thompson notes that AI’s risk to humans is not how closely it mimics people, but how constantly it tempts people to mimic it. That insight suggests to me that rising professionals in a world of AI should find ways to practice authentic intelligence in cahoots with artificial intelligence. Much of what you do today will be done by agents that aren’t human. Don’t shy away from that partnership, but do insist on being you in the thick of it.