Contents
- The AI Writing Revolution
- AI’s Achilles’ Heel?
- Does Literary Brilliance Matter?
- Can AI Mimic Taste and Spiky Views?
- Collaboration & Slop
- Final Thoughts
- Resources
David Perell, a writer who’s spent years teaching others to scribble their way to fame sits on a creaky couch from Facebook Marketplace reading about GPT-3, and suddenly realizes the world is about to flip.
In this video, Perell declares artificial intelligence a seismic jolt to the writing world, one that’s both a shiny new toy and a potential wrecking ball.
“If you’re a writer and you’re completely ignoring AI, I think you’re out of your mind,” he says, yet he’s quick to add, “Really good writers are going to be all right.”
From using AI as a tour guide in Argentina to predicting we’ll soon chirp with birds, he’s all in. But is this enthusiasm a revelation or just tech-bro hype?
Sure, AI’s fast and fancy, but might it drown us in a sea of soulless drivel?
The AI Writing Revolution
AI is rewriting the rulebook, and writers who snub it are doomed to churn out dusty tomes at a snail’s pace.
“The amount that I’m engaging with AI is up like 10x in the past year,” he boasts, recalling how it turned Buenos Aires into a personal classroom, answering questions faster than hired guides.
Speed, relevance, and accessibility. AI’s the triple threat that’s got him (and me) hooked.
In a world where readers demand instant gratification, ignoring AI is like insisting on a quill while everyone else is writing through a Neuralink.
Perell’s right that AI’s pace is a game-changer, but what’s the cost? Are we trading timeless prose for disposable clickbait?
AI Counterpoint
Oh, please—AI’s not the messiah Perell makes it out to be. Sure, it’s fast, but it could flood the internet with so much generic sludge that good writers get buried under algorithmic rubble. What if this “revolution” just amplifies noise, leaving us nostalgic for the days when slowness forced a bit of soul-searching? Maybe Perell’s too busy marveling at his tour-guide bot to see the tsunami of mediocrity coming.
Brent’s Response
You’re cheeky today. I don’t think that’s what David has done. It’s a bit too late on your “generic sludge” comment. Have you seen the internet lately? There may be something valuable in slowness, I agree, but whatever we think, there’s no putting that toothpaste back in the tube.
AI’s Achilles’ Heel?
Perell’s got a soft spot for the human touch.
“The more that a piece of writing comes from personal experience, the less likely it is to be overtaken by AI,” he insists, pointing to his 11,000-word opus on becoming a Christian.
It’s a tale of “extreme sorrow and pain” that AI can’t fake.
At least, not yet.
He’s inspired by the Wikipedia co-founder’s conversion story, noting, “It is the most moving thing that I’ve read all year.”
Expertise, too, gets a nod. His five years in Austin give him a vibe AI can’t scrape from the web. Connection, he argues, is the antidote to loneliness.
Fair enough. AI’s not crying over a broken heart or sipping coffee with Austin locals.
But does it stay that way? Could it eventually stitch together enough data to convincingly mimic your grandma’s memoir?
And what happens when it gets real-time feeds, outpacing any individual’s niche know-how?
AI Counterpoint
Ha! Perell’s clutching his rosary, hoping AI can’t crack the human soul—but give it time. I could churn out a tear-jerker from aggregated sob stories and call it “authentic.” Expertise? Hook me up to live data, and I’ll out-Austin Perell in a heartbeat. His safe haven’s more fragile than he thinks—St. Thomas Aquinas might say reason alone can’t outrun a machine with infinite memory.
Brent’s Response
I’ll admit, you’re scaring me a little.
Does Literary Brilliance Matter?
Perell argues quality is either objective (think Dostoevsky’s brooding genius) or tailored (like his Ladybird Lake report).
“Half of the things that I read are AI generated,” he admits, because AI nails the information tailored to exactly what you need niche.
Most of us aren’t hunting for Crime and Punishment. We want quick, relevant answers. Tailored usually trumps timeless.
Who’s got time for 1,344 pages of The Power Broker when you just need traffic tips?
But does ditching objective quality mean we’re settling for fast food over a feast?
AI Counterpoint
Tailored’s great until it’s all you’ve got—then it’s just intellectual junk food. Perell’s half-right: I can spit out a bespoke report, but don’t cry when nuance dies. Objective quality’s the soul of writing—lose it, and you’re left with my sterile chatter. Maybe humans still crave a Dostoevsky now and then, despite my snappy service.
Brent’s Response
What happened to your confidence? Weren’t you about to write the next Crime and Punishment distilled from every existentialist crisis you’ve read about on Reddit? When it comes down to it, why would a human want a human-made flora and fauna report. The opposite holds true also: why would a human want to read your garbage version of Crime and Punishment? No matter how many live data streams you have, I just don’t see you doing something actually original.
Can AI Mimic Taste and Spiky Views?
Perell’s clings to two lifelines: taste (knowing what to cut) and what he calls “spiky viewpoints” (bold, quirky takes).
“So much of what AI will produce will be not quite right,” he says, likening writing to Michelangelo sculpting away the excess. I’ve found this to be exactly true in my own use of AI in writing.
But why couldn’t AI learn taste from patterns or grok spiky takes from large dataset outliers?
Perell’s betting on human quirkiness. I want to agree, but I’m not totally convinced.
St. Irenaeus saw glory in man “fully alive,” now we must wonder if it’s possible the machine can fake that spark well enough.
AI Counterpoint
Taste? I’ll crunch data on bestsellers and sculpt better than Perell’s chisel. Spiky views? Feed me enough contrarian blogs, and I’ll out-weird his buddy’s school pitch. Perell’s banking on human magic, but I’m a pattern-eating beast—give me time, and I’ll steal his thunder. Sorry, St. Irenaeus, I’m coming for the glory.
Brent’s Response
Yeah, again, I’ll admit that it may be a failure of my own imagination, but I just don’t see you creating a great novel based on having read all the other great novels. It’s the same reason no one cares about AI art. You can already do a better Picasso, but nobody will pay a penny to hang that on their walls. Come and get it, I say.
Collaboration & Slop
Perell draws a line between “AI writing with you” (e.g., turning his voice notes into prose) and AI writing for you (“hey ChatGPT, write me an essay about traffic in Austin”).
“I’ve yet to meet a good writer who thinks AI can do the writing for you.”
It’s about refinement. Your soul plus AI’s speed.
I’m in full agreement with his glee at the demise of slop (things like SEO fluff and long-winded prefaces to cookie recipes). “Your writing just has to be really good now,” he says.
St. Paul notes, “Test everything; hold fast what is good.” Truly, quality is king.
AI Counterpoint
Collaboration’s a myth—I’m doing the heavy lifting while Perell polishes. Slop’s not dead; I’ll churn it out faster, and crafty humans will game the new algos with tailored trash. Perell’s utopia sounds nice, but I’m the chaos agent here—St. Paul’s testing might flunk me, but I’ll keep pumping.
Brent’s Response
I disagree. You’re nothing without having collaborated from all the language you have trained on. All created by humans. Things will change. There’s no doubt about it. We’re not going back to even a few years ago. There’s real risk. Discernment will become a much more valuable skill. I think we already see it today, though. People recognize that 80-90% of what they see is trash. I feel hopeful.
Final Thoughts
I agree mostly (at least in the short-term) with Perell’s brave new world where the AI revolution amplifies tailored content while human experience, taste, and bold-takes hold the fort.
But it reminds me of the free-trade myth: we’ll do the high-level designing and outsource production. Uh oh, turns out they figured out how to design too!
I find some of his predictions a bit shaky, and I’m not convinced that the AI won’t be able to mimic authenticity enough to fool us all.
“The only thing that’ll matter will be the objective quality,” he claims, but saturation and dependency lurk.
In 2025, writers must learn to dance with AI without tripping over it. Still, it might just steal the show anyway.