“MrBeast has 440 million subscribers, and I think there will be 440 million different MrBeast versions for each person based on the data.”
~@RevanLazarus
“MrBeast has 440 million subscribers, and I think there will be 440 million different MrBeast versions for each person based on the data.”
~@RevanLazarus
Honored to have joined @InfiniteL88ps on today’s episode!
I’d encourage you to check out my episode and the show as a whole. We talked about AI creators, the future of the creator economy, podcasts, and what we’re building at @usejamieai. Thanks to @nicktawil and @jposhaughnessy for having me on the podcast!
AI started as a tool for creators. Next, it may rebuild the creator economy itself.
@RevanLazarus, founder of Jamie, joins Infinite Loops (guest-hosted by @NickTawil) to discuss AI creators, personalized media, brand deals, audience data, and the future of content.
TIMESTAMPS
0:00 Intro
3:30 The AI 18-Wheeler
7:24 Retention-Optimized Interviews
12:36 Why Real Audience Connection Matters
17:29 How Sales Teams Prove Audience Fit
22:21 440 Million Versions of MrBeast
29:03 AI Creators and Short-Form Content
34:34 Why People Are Booing AI
36:47 Will AI Create More Jobs?
40:35 AI Is Bigger Than the Internet
46:38 Why AI Companies Need Niches
@owenwb If people are not uncomfortable with your price, you’re not charging enough. 99% of the time they’ll negotiate with you rather than just saying no. Keep raising prices until multiple people tell you that’s way too much.
Today marks a tremendous MILESTONE for our AI podcast producer, Jamie. We’ve officially replaced our first human producer. The news came to me this morning from one of our clients, who was so thrilled with Jamie that they will now use it for all their associate producer roles, and subsequently had to let go of one of their full-time producers. This is a super strange feeling, honestly.
On one hand, I am over the moon that we’ve actually built the product we dreamed of creating and IT WORKS! It’s proof that what we’re building is truly disruptive to the podcasting industry.
On the other hand, it feels deeply personal. Not even a year ago, I was doing THIS JOB, and could’ve been replaced by Jamie (check out the video below). I know the grind, craft, and effort it takes to put together a show. Although I don’t know this producer, I feel for them. I’ve sat in their seat and to see that role automated by something we created is both thrilling and sobering.
For the first time, I feel the responsibility that comes with real innovation. Innovation creates opportunity and efficiency, but also completely changes people’s livelihoods.
For myself and our team, this responsibility is not taken lightly. Today is a reminder of the importance of what we’re building and the weight we carry moving forward.
Our AI is SPYING on your guests… let me explain. Jamie is learning to listen to EVERY past interview or podcast your guest has done.
Why?
Because the #1 complaint we hear from podcasters is:
“I don’t want to ask the same questions every other show has asked.”
or
“I don’t have time to sit through every past interview just to spot one intriguing detail.”
Current AI tools just scrape bios, headlines, or summaries, leading to the same boring questions and insights. But the best questions come from the CONTEXT - the subtle answers a guest gave on a random YouTube interview a year ago, and how you expand on those.
So we’re training Jamie to:
✅ Watch/listen to past interviews
✅ Transcribe and analyze them
✅ Surface fresh angles
✅ Avoid repeat questions
✅ Help you walk into every interview with a unique perspective fit for your show
It’s early, but tests are already pulling out incredible insights that we would’ve completely missed otherwise! So sick!
If you’re a podcaster, producer, or media team member - what would you want from a feature like this?
I’d love your feedback as we build!
Almost a year ago, I started building out my first podcast. Frankly, I had no idea what I was doing, and with 3 celebrity hosts counting on me, I needed to figure it out FAST.
I researched "how to be a podcast producer?", spoke to producers I knew, and aimlessly tried to prepare. I was out of my element.
A few days before our first episode, I paid a freelance producer $750 to write me a run-of-show (a game plan). Maybe I didn't give him enough time and materials, or my bar is just super high, but it wasn't very good, and it was definitely not what I imagined.
I spent the next 24 straight hours researching our guest and what our audience might want: scrolling Reddit, Twitter, TikTok, Instagram, watching every podcast and interview they've done, and asking ChatGPT everything it knew about the guest and how they related to the show we were about to start. Again, it was SH*T.
The final product: 17 pages of scrambled notes and questions on Google docs, most of which said the same thing just in a different way.
The result: The most frustration I'd ever felt and lack of executable material.
I continued this process for another 5 episodes until I knew, I either needed to pay someone to do this full-time or commit all my time to being a podcast producer. Neither was an option.
I looked at our Head of Engineering, Michael Mathieu, and said we need to fix this.
After speaking to countless others, I knew we weren't alone.
So for the past few months, we've been working non-stop on Jamie, an AI co‑producer for media teams and podcasters (link in the comments).
Jamie:
- Analyzes your past content to understand tone, style, and recurring themes.
- Surfaces fresh hooks and angles by looking at what is performing well on other podcasts and trending topics from your niche.
- Watches guests' past interviews and podcasts so you don't have to (aka you're not repeating what's already been asked).
- Prioritizes what your audience cares about most by analyzing comments, social platforms, and YouTube Studio data.
- Builds tailored research briefs on each guest, including their lesser‑known insights, so you show up ultra‑prepared.
- Creates a damn good run-of-show with all of these elements taken into account.
- Iterates with you over time, learning what resonates and sharpening its suggestions every episode.
We've signed a bunch of fantastic early customers, proven the core use case, and hit our first revenue milestones! I am super proud of our team: Peri Pascal, Michael Mathieu, and George Su, for that!
Now I'm going to share the whole process publicly as we scale, which means:
- Wins and setbacks
- Lessons learned in real-time
- Metrics as we grind to $1M in ARR
- and I promise no AI-generated posts lol
I hope you follow along and check out Jamie as we build! Shout out to Olympic Champions: Noah Lyles, Rai Benjamin, and Grant Holloway for making this happen!
The @joinoffside team and I will be attending @boardroom’s GAME PLAN tomorrow in Santa Monica, CA. I’d love to meet up with anyone attending or in town! Going to be great! #gameplan