Discover
Unscripted with Jeff Pedowitz
Unscripted with Jeff Pedowitz
Author: The Pedowitz Group
Subscribed: 1Played: 13Subscribe
Share
© Copyright 2026 All rights reserved.
Description
Most podcasts let their guests give the answers they came prepared to give. This one doesn’t.
Unscripted is where Jeff Pedowitz — CEO of The Pedowitz Group and one of the most recognized voices in AI-driven revenue marketing — sits down with the builders, thinkers, and leaders who are actually shaping what comes next. Not the keynote version. The real version.
Every conversation runs through the same throughline: AI is no longer the story, it’s the substrate. What does that mean for the companies being built, the policies being written, the institutions trying to hold, and the humans navigating all of it?
Guests come from everywhere — drug discovery, politics, investment, education, creative work — because the AI question doesn’t belong to any one industry. It belongs to everyone trying to do something that matters in a world that’s changing faster than the frameworks we use to understand it.
No scripts. No softballs. No prepared answers left standing.
Jeff asks what no one else thought to ask. That’s the whole show.
Unscripted is where Jeff Pedowitz — CEO of The Pedowitz Group and one of the most recognized voices in AI-driven revenue marketing — sits down with the builders, thinkers, and leaders who are actually shaping what comes next. Not the keynote version. The real version.
Every conversation runs through the same throughline: AI is no longer the story, it’s the substrate. What does that mean for the companies being built, the policies being written, the institutions trying to hold, and the humans navigating all of it?
Guests come from everywhere — drug discovery, politics, investment, education, creative work — because the AI question doesn’t belong to any one industry. It belongs to everyone trying to do something that matters in a world that’s changing faster than the frameworks we use to understand it.
No scripts. No softballs. No prepared answers left standing.
Jeff asks what no one else thought to ask. That’s the whole show.
27 Episodes
Reverse
What does it take to make AI safe? According to my latest Unscripted guest, you start by breaking it.
Sahil Agarwal has a PhD in applied math from Yale. His path ran from nonlinear dynamics at Oxford to climate modeling to hunting for planets outside our solar system, and it ended up somewhere few people expected. Today he leads Enkrypt AI, where his team red teams the frontier models businesses are building on. They push those models until they fail, then build the guardrails that stop them.
As a kid in India, while other kids were outside playing, Sahil was taking things apart with a screwdriver. His parents scolded him for breaking more than he fixed. Turns out that was training.
A few things he told me that I can’t stop thinking about:
“No model is safe.” It wasn’t said for shock value. It’s just what his team keeps finding.
A model that refuses a harmful request once can often be talked into it slowly, one harmless-looking question at a time.
An attack doesn’t have to be text. It can hide inside an ordinary-looking image or audio file.
If your agent goes rogue, you can’t fire it. The liability lands on the business that deployed it.
On a scale of 1 to 10, how worried should we be? His answer: a 7 or 8.
But this isn’t a doom conversation. Sahil uses AI every day and believes in its potential. His point is that safety isn’t a yes-or-no question. It’s about how much risk you’re willing to take, and putting the right guardrails in place now instead of bolting them on later. As he puts it, we’re still the ones deciding whether this ends up looking like WALL-E or The Terminator.
If you’re deploying AI agents in your business, this is a conversation you’ll want to hear.
Halfway through today’s interview, Kathleen Kennedy Townsend paused, reached off camera, and held up a letter.
Her father, Robert F. Kennedy, wrote it to her the day President Kennedy was buried. She was 12 years old.
“Dear Kathleen, you seem to understand that Jack died and was buried today. As the oldest of the Kennedy grandchildren, you have a particular responsibility now. A special responsibility to John and to Joe. Be kind to others and work for your country. Love, Daddy.”
On the worst day of his life, he didn’t write about anger or revenge, and he didn’t ask who was responsible. He wrote about kindness, responsibility and service.
As Kathleen put it: “When terrible things happen to you, the reaction doesn’t have to be, I’m going to be mean back. That’s a waste of your energy.”
She has spent her life answering that letter.
She became the first woman to serve as Lieutenant Governor of Maryland. She made Maryland the first state to require student community service, after showing up to the hearing with 800 supporters behind her.
She co-founded the Georgetown Center for Retirement Initiatives when zero states had auto-IRA programs. Today 18 do, and billions of dollars are saved by workers who never had a retirement plan before. Most of them will never know her name, and that’s fine with her.
When I asked whether her father’s and uncle’s work was ever finished, she corrected me right away: “There is no finish. The work goes on.”
Thank you, Kathleen, for trusting me with something so personal, and for a conversation I’ll carry with me for a long time.
Your best employee might be the reason you can’t scale.
I’ve hired them, and I’ve defended them in meetings. There was the account manager who crushed his number every year and was universally disliked by everyone around him. There was the consultant who was brilliant with clients and quietly toxic to the team.
It took me too long to see it. You can put your values on every wall in the office, but if you keep employing people who break them, your team hears a very different message.
This week on Unscripted, I sat down with Russ Reeder, who has put a name on this problem.
Russ has led nine PE-backed companies across six continents. He started Rightsline, which now runs the rights and licensing back office behind Netflix, Hulu, Sony and Universal. He helped sell Media Temple to GoDaddy. Today he runs KeyDelta, helping leadership teams fix the operating model before they bolt AI onto it.
His new book is called Heroes Don’t Scale.
A “hero” is the engineer who is the only one who knows how the code works. It’s the rep who hoards every client relationship, or the manager who treats information like currency. Everyone celebrates them, and the company slows down around them.
Russ has replaced five CEOs. Three of them were heroes.
A few more moments from the conversation stuck with me:
Back in 2006, his team had encrypted movies streaming from hotel basements into people’s homes. They had the studio deals and the set-top boxes, and Warner Brothers was ecstatic. One controlling shareholder killed it. You know who built that business instead.
AI is about to create a new generation of heroes. Picture the one employee who vibe-coded your sales automation and is the only person with the login.
People who are scared of AI taking their job should be scared. Russ argues the cure is curiosity, because you can’t be curious and afraid at the same time.
Companies are bought, not sold. Cutting your way to $5M more in EBITDA doesn’t fool the buyer.
Russ asks one test question in his book: if your key person quit tomorrow, would the business still work?
Most leaders say no. Russ says nearly every company would survive. The fear of losing that person is usually the real problem.
So tell me in the comments: who is the hero in your company, and what is it costing you?
#Leadership #AI #PrivateEquity #Scaling #Unscripted
Most economic forecasts come from people with something to sell. Erik Britton is a rare exception.
He began his career inside the Bank of England, went on to help lead Oxford Economics, and now runs Fathom Consulting in London, one of the few genuinely independent macro shops in the world, where governments, global asset managers, and pension funds go for analysis that sits where economics, geopolitics, and markets meet.
In this conversation, Erik makes a call most people in the AI conversation won’t make out loud: either this is the most transformational technology in human history, bigger than electrification, the computer, or the jet engine, or it’s a bubble. He takes the bubble side, and walks through the arithmetic behind it.
We get into:
Why current AI capex implies returns that would require a 10 to 15 percent increase in US GDP
The difference between a dot-com bubble and a 2008 bubble, and why hidden debt in shadow banking, private equity, and venture capital makes this one closer to the second
Why AI displacement is hitting the top of the skills distribution instead of the bottom, and what that does to the apprenticeship path into senior roles
How a central bank should classify AI: positive supply shock, negative demand shock, and financial stability risk, all at once
How Erik actually uses AI in his own work, and why he hasn’t embedded it in Fathom’s analytical workflow
What China’s dual circulation strategy is really optimizing for, and why it isn’t making its own people rich
Why every forecast critic is already running a forecast, whether they admit it or not
His direct advice for leaders building a 2027 plan: expect to get punched in the guts, and start anyway
A conversation for CEOs, CFOs, and revenue leaders who have to commit to a number in an environment nobody can model cleanly.
Read the companion article: https://www.pedowitzgroup.com/blog/its-a-bubble-or-its-the-biggest-thing-that-ever-happened-erik-britton-on-planning-for-2027
Learn more about Fathom Consulting: https://www.fathom-consulting.com
#Unscripted #AIBubble #MacroEconomics #B2B #RevenueMarketing #AIStrategy #ThePedowitzGroup
"I've been to space. I've seen Earth from the outside.
And I'm more worried about what's happening down here than anything I saw up there."
Colonel Terry Virts has flown 40 combat missions, piloted the Space Shuttle, commanded the International Space Station, directed a documentary, and earlier this year ran for Congress in Texas.
He came on Unscripted to talk about AI. We ended up talking about all of it.
The thing that stayed with me: he teaches a case study at Harvard Business School about the Challenger and Columbia disasters. Engineers were warning NASA. Managers had schedule pressure. Warning signs were normalized. Seven people died.
His TED Talk argument: the AI moment looks exactly the same.
Except this time, the carrot isn't a launch date. It's a trillion-dollar IPO. And the consequences of getting it wrong aren't seven lives.
A few things from this conversation I can't stop thinking about:
"Don't normalize deviance. If something bad's happening and it's not killing you, don't assume it's fine. Maybe you're just getting lucky."
The night he and a Russian cosmonaut watched bombs explode over Ukraine from the ISS window — and didn't say a word
He sold his Russian space suit and donated the money to Ukraine
His $5-a-day government per diem after 200 days in space
Why bioweapons, not chatbots, are what actually keep him up at night
One of the most wide-ranging conversations I've had on this show. Space, AI, democracy, corruption, civic duty — and why the founding fathers anticipated Trump but never anticipated Mike Johnson.
Full episode is live now. Link in the comments.
#AI #Leadership #Space #Democracy #Unscripted



