Email has only had two truly big moments in its history.
The AOL CD that showed up in your mailbox in 1998. And the BlackBerry that untethered your inbox from your desk.
The third one is happening right now.
This week on Audience Bridge Insights, I sat down with Dan Oshinsky, owner of Inbox Collective, to talk about what AI is about to do to the inbox, and what newsletter operators should actually do about it.
Dan's résumé speaks for itself.
He ran newsletters at BuzzFeed and The New Yorker before launching Inbox Collective, where he now helps newsrooms, nonprofits, and independent newsletters get more readers and make money.
I caught his talk on AI and newsletters at the Newsletter Conference back in May.
It was the conversation everyone was still having in the hallways afterward.
So I brought him on the show to keep it going.
You can watch the entire podcast now or scroll down to get the full breakdown.
The Third Big Wave
Dan frames email history in three moments.
Wave one: email becomes a platform.
Everyone gets an AOL screen name and an embarrassing email address.
Wave two: mobile.
Suddenly people check their inbox 50 times a day instead of twice.
Both waves were great for newsletters. More people, more time in the inbox, more chances to reach your reader.
Wave three is AI.
And this one is different, because nobody, not even the Gmail team, knows what the inbox looks like in three, five, or ten years.
Here's the uncomfortable truth Dan put on the table: nobody loves checking email. There's too much of it, and most of it is stuff you never wanted.
AI is very good at learning what you want, filtering what you don't, and handling the repetitive stuff behind the scenes.
It's coming for the inbox whether we like it or not.
Write for the Human, Not the Robot
The first question every operator is asking: who am I writing for now?
People… Deliverability… AI?
Dan's answer hasn't changed.
Write for the person.
His logic is simple.
Email is too deeply integrated into our lives, especially our work lives, to disappear.
What changes is what survives the filter.
The stuff that I'm gonna make time for is gonna be really good stuff that's voicey, it comes from a person, it has expertise, it has perspective.
His framework for what wins: depth.
Depth of knowledge. Depth of insight. Depth of access. Depth of analysis.
Dan's a die-hard Washington Nationals fan.
His point: he would never outsource his Nats fandom to an AI agent spitting out weekly stat summaries.
If a great weekly newsletter broke down the farm system and what's happening behind the scenes, he'd make time for it every single week.
Nobody outsources the things they actually care about.
If your newsletter is something people care about, AI filtering is your friend. If it's something people tolerate, you're already dead.
The Mailbox at the End of the Driveway
Dan's best analogy of the episode: your physical mailbox.
Over the last two decades, our reliance on physical mail collapsed.
Yet Dan still walks to the end of his street for the handful of things he loves, the New Yorker, the Atlantic, and (his confession, not mine) every AARP publication.
The junk still shows up. It goes straight in the trash.
That's the inbox's future.
AI gets better at identifying the marketing messages you never engage with and simply stops showing them to you.
What's left is the stuff you look forward to.
For good operators, that's not a threat. That's a cleaner room to stand out in.
Macro, Not Micro
This was Dan's most practical advice for anyone panicking about every AI press release.
Think about the macro, not the micro.
The macro:
AI is coming. Write deep content. Nail your email authentication. Get your verified mark certificate for that checkmark in the inbox. Build your brand. Build automations that trigger relevant messages at the right time.
The micro:
That screenshot going around of Gmail rewriting pre-header text with its own AI summary.
These inboxes roll out lots of different tests and features. And some of them stick and many of them do not.
Remember the Gmail tabbed inbox era?
Dozens of features launched. Most quietly died because users didn't care.
Teams that panic-optimize for every new feature burn time on things that may not exist in a year.
Teams that do the fundamentals win in every version of the future.
Where AI Actually Earns Its Keep
We compared notes on where AI belongs in a newsletter operation right now.
Dan's list: process, brainstorming, and analytics.
Not voice. Not strategy.
His favorite example: a newsroom client spending five to six hours a month manually sorting CSVs into spreadsheets. He taught Claude the process once. Now they upload the file, and seconds later they get the reports, plus dashboards they never had before.
My biggest use: is subscriber data analysis.
We export the raw activity data - signup date, first open, first click, last click, all of it - and analyze exactly where subscribers fall off, by acquisition source.
That's how we find the opener bloat: the "engaged" subscribers who open constantly but never click. Most of them are bots, security scanners, and prefetched images, not people.
And it's why I live by the 2-Minute Rule: any activity within two minutes of delivery isn't a human.
Cut that noise and the real picture appears: when to sunset subscribers, when a win-back actually works, and when it's not worth sending at all.
Gmail already knows which of your opens are real. The question is whether you do.
The Win-Back Nobody Else Is Running
Here's where I think AI gets genuinely exciting for monetization and reactivation.
Most win-back campaigns are evergreen. Generic. Set once, forgotten for months, barely moving the needle.
Now imagine a win-back that knows what's happening in your niche this week, knows what that subscriber actually read before they went dark, and writes each message specific to that person.
That's not a canned "we miss you" email. That's a reason to come back.
Dan's excited about the same shift on the front end: AI-optimized welcome flows where survey answers drive personalized offers, upsells, and welcome series that would be impossibly complicated to build by hand in an ESP.
The one-size-fits-all automation era is ending on both ends of the subscriber lifecycle.
The Real Fear: The Unknown
I asked Dan what newsletter publishers should actually be worried about.
His answer: the unknown itself.
The entire newsletter economy is built on one assumption - that email doesn't change much.
No algorithm. No single company setting the rules. A stable place to build.
AI challenges that stability for the first time.
Ten years ago, if you asked Dan what email would look like in a decade, the answer was "pretty much the same."
Ask him what June 30th, 2036 looks like, and his honest answer is: he truly doesn't know.
Let's not cling too tightly to the past. What we build next may be more interesting and more exciting than what's come before.
My Take After This Conversation
Three things stuck with me:
1. Depth is the new moat.
Knowledge, insight, access, analysis.
An AI-filtered inbox kills commodity content first. If your newsletter is an RSS feed in reverse chronological order, "that's how we've always done it" won't save you.
2. The fundamentals got more important, not less.
Authentication, brand certificates, engagement hygiene, smart automations.
Every future version of the inbox rewards senders who prove they're reputable and relevant. Chase the macro, ignore the micro.
3. AI belongs behind the scenes, not in your voice.
Process, analysis, segmentation, timing.
That's where the leverage is. Your perspective, your choices, your angles - that's the part no model replaces, and the part your readers are actually subscribed to.
If AI cleans up the inbox the way Dan and I both expect, the operators with real engaged audience density will be the ones left standing.
See you next week,
Chris Miquel
P.S. Everything Dan and I talked about comes down to one thing: in an AI-filtered inbox, only real engagement survives. The bots, the prefetched opens, the security scanners - none of it will save you.
That's why we built Smart Delivery, domain-level deliverability and engagement monitoring so you know what Gmail is actually doing with your email before your numbers fall off a cliff. And Smart Reactivation, which finds the dormant subscribers still worth winning back using real click signals, not opener bloat.
If you want to know how much of your "engaged" list is actually human, book a call and let's find out.
