BoSacks Speaks Out: - What AI Can’t Replace: The Value of Human Judgment in Publishing
By Bob Sacks
Thu, Jul 3, 2025

Redesign or Resign: How Publishing Must Evolve in the Age of AI Replacement
There was a time, not long ago, when artificial intelligence was wrapped in the gauze of optimism. The corporate line was neat: AI was here to help, not replace. It would handle the drudgery and free up humans for higher thinking. The era of collaboration was upon us. Augmentation, not obsolescence. The narrative was smooth, comforting, almost melodic in its predictability.
But that tune has changed. Now, the hushed tones of boardrooms have erupted into full-throated truth-telling at conferences with catered hors d’oeuvres and panoramic views of the Rockies. At the Aspen Ideas Festival, Ford’s CEO Jim Farley pulled no punches. AI, he said plainly, will “replace literally half of all white-collar workers in the U.S.” Not might. Not eventually. Will. And his delivery was that of a man who’s already seen the PowerPoint and signed off on the restructuring memo.
And he’s not the only one letting the mask slip. Marianne Lake at JPMorgan is openly predicting staff reductions. Andy Jassy at Amazon isn’t hiding the trimming of corporate ranks. Dario Amodei of Anthropic says half of all entry-level jobs could evaporate in five years. These are not fringe declarations. These are the architects of the modern economy sounding the alarm from the penthouse.
The era of polite ambiguity is over. The plausible deniability has expired. And let’s be honest, we all saw it coming. The AI conversation has shifted from augmentation to elimination. From partnership to replacement. And nowhere is this more evident than in the cold, hard recalculations of value. If a machine can do it faster, cheaper, and without paid Time off, the human is suddenly an overhead line item too burdensome to ignore.
Take Shopify, where hiring is now contingent on proving a flesh-and-blood applicant can outperform an algorithm. Or Fiverr, whose CEO bluntly told staff: AI is coming for you. Not just coders or customer service reps. Everyone. And at companies like Moderna and ThredUp, the mantra is clear, launch more, hire less. Do more with fewer humans.
This isn’t paranoia. This is preparation.
Even the measured voices, those who remind us that past technological revolutions brought new forms of employment, can’t offer specifics. They point to “new jobs,” but when pressed, the vision gets hazy. What, exactly, will those new jobs be? And more pressingly: will humans be the ones doing them?
Let’s shift the lens now to publishing, the world I know best. Let’s talk about us, the storytellers, the truth-chasers, the layout wizards, the ones still fighting for Oxford commas and long-form relevance.
We are not exempt. In fact, publishing might be the canary in the algorithmic coal mine. AI is already attending our editorial meetings. It’s rewriting press releases in seconds, generating headline variations, even laying out entire web pages. You might not see it, but you can feel its breath on your neck every time a deadline shortens or a team size shrinks without explanation.
The pressure is real. Reduce costs. Accelerate output. Scale infinitely. And conveniently, there’s a silicon savior available, one that never sleeps, never unionizes, and never pushes back on a bad lead.
So what’s left for us? Where do we carve out meaning in a world where meaning itself is being mechanized?
First, we stop pretending this is business as usual. It’s not. The industry is undergoing a tectonic shift, and we either face it head-on or watch our value erode in slow motion. Publishers must examine every function and ask, does this require a human mind, or just a well-trained model? If it’s the latter, it’s already at risk.
Second, we build hybrid systems that play to our strengths. Let AI handle the mechanical, copyfitting, metadata, SEO tagging. But let humans own the terrain where machines still falter: tone, judgment, moral complexity. Because here’s the thing, AI can mimic syntax; it can’t replicate soul. And we must make that difference visible in every article, every op-ed, every photo caption that punches above its weight.
Transparency is not a virtue. It’s a necessity. If you’re using AI to create content, say so. Loudly. Readers aren’t stupid, and trust is leaking faster than ad revenue. If your audience suspects they’re being duped, they’ll bounce faster than you can say “pageview optimization.”
We must also, desperately and defiantly, reclaim the human story. Stop writing for clicks and start writing for consequence. The algorithm wants predictable. Feed it weird. Feed it messy. Feed it essays so full of depth and contradiction that a language model chokes on the nuance. Write things that couldn’t have been imagined in a prompt. Write what scares you. That’s where meaning lives. As I said once before, we need to be stubbornly human.
And finally, we train like hell. We educate up and down the masthead, not just in tool use, but in critical thinking. We don’t fear the bot. We learn it. We test its limits. And we decide when to walk away and write it ourselves, because we know the difference.
Yes, the tide is rising. The machine is humming. The future is already knocking, and it sounds suspiciously like your managing editor asking if ChatGPT could handle that 800-word culture recap.
But here’s the good news: we still get to decide how we respond.
And here’s what the machines still don’t get, what they can’t simulate, approximate, or harvest from the statistical echo chamber of prediction: human unpredictability, human empathy, human daring. The goosebumps from a perfectly turned phrase. The gasp when a story lands exactly where it hurts or heals. The ability to read between the lines, not just for what’s there, but for what’s missing.
We, the humans, bring contradiction, context, and conscience. We know when silence speaks louder than copy. We understand that sometimes a story’s power lies not in its virality but in its vulnerability. That’s not code. That’s consciousness.
So yes, let the machines help. Let them sort the haystack. But when it comes to finding the needle, that glint of meaning in the noise, it still takes a human hand.
Publishing’s future doesn’t hinge on outpacing AI. It hinges on out-humaning it. On making things with care, conviction, and that reckless bit of magic no machine can quite replicate: soul.
Keep the presses warm. We’re not done yet.