Oct 14, 2025 · 34m · we-live-to-build
Most People Know the Hard Call They Need to Make - and Don't
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Executive coach Ariel Lechner and host Sean Weisbrot examine the intersection of artificial intelligence, automated workflows, and executive leadership, offering actionable strategies for building custom GPTs, optimizing client deliverables, and overcoming behavioral inertia to make decisive leadership calls.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Sean holds 58.2% of the talking time here. How this is scored →
Sean as informed peer 5.0 Guest teaching 2.1 Guest disagreement 1.1 Sean pushing back 1.9
speaking balance: gold is Sean, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 17:10 Ariel rejects summary-based quarterly workflows
Ariel explicitly dismisses Sean's proposed intermediate summary architecture, arguing that summaries lose essential language context and result in low quality.
Hardest push from Sean ▶ 20:07 Sean corrects 'free' vs 'included' positioningSean immediately interrupts and corrects Ariel's framing of client reports as 'free,' insisting they must be positioned as 'included' so clients perceive and respect their value.
Biggest teaching moment ▶ 33:05 Ariel on leadership avoidance and executive hard callsAriel delivers the core pedagogical takeaway of the interview, educating the host on how executive avoidance of discomfort delays vital organizational decisions.
Sean holds their own ▶ 16:04 Sean designs an automated client recording pipelineSean demonstrates practical systems expertise by mapping out an end-to-end automation from Fathom recording to Google Drive routing and two-tier LLM summarization.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Sean as informed peer | Guest teaching | Guest disagreement | Sean pushing back | Why |
|---|---|---|---|---|---|---|
| The Urgency of AI Adoption in Business | 5 | 2 | 1 | 2 | Sean opens with a strong thesis on the urgency of AI adoption, claiming businesses only have a few months to adapt. Ariel agrees with the urgency but clarifies the distinction between engineering adoption and day-to-day executive personal workflow. | |
| Incentivizing Team AI Adoption and Workplace Automation | 4 | 2 | 1 | 1 | Sean frames adoption as a five-alarm issue while Ariel shares a bottom-up monthly presentation incentive model. Sean elaborates with an anecdote about workplace automation making employees faster and more efficient. | |
| AI's Impact on Executive Coaching and Knowledge Bots | 5 | 2 | 1 | 1 | Ariel details how coaches can offer custom GPT bots loaded with transcripts to serve clients between sessions. Sean connects this to creating marketing digital twins trained on Dan Kennedy and Alex Hormozi books. | |
| Data Privacy, Intellectual Property, and AI Workflow Constraints | 6 | 3 | 2 | 4 | The dialogue shifts to intellectual property and model memory retention when feeding copyrighted material. Sean explains technical nuances of podcast pre-production prompts and the necessity of text chunking over raw multimedia uploads. | |
| Automating Client Session Notes and Quarterly Reporting Workflows | 7 | 2 | 2 | 4 | Ariel asks about handling lengthy quarterly transcripts for reporting. Sean proposes an automated folder-and-summary workflow, defends the approach when Ariel pushes back against summarization quality loss, and corrects Ariel on framing deliverables as 'included' rather than 'free.' | |
| Personalizing ChatGPT Settings Versus Building Dedicated Digital Twins | 6 | 1 | 1 | 2 | Ariel describes fine-tuning global ChatGPT account personalization prompts with life context. Sean contrasts this with deploying dedicated standalone web assistants trained exclusively on proprietary domain knowledge. | |
| Hyperfocus, Habit Consistency, and Rapid Content Creation | 5 | 1 | 1 | 1 | Sean discusses his all-or-nothing hyperfocus driven by ADHD, including writing an 8,000-word book in six hours. Ariel compares voice dictation while walking to Sean's hybrid typing and Telegram voice-to-text method. | |
| The Hard Call: Overcoming Discomfort in Executive Leadership | 2 | 4 | 0 | 0 | In the concluding segment, Sean asks for Ariel's central leadership lesson. Ariel articulates that executives usually know the hard decision they need to make but actively avoid the necessary discomfort. |
Statements from this episode (11)
Lechner: Only about 20% of coaching clients personally use AI daily
“I don't think very many. I think maybe 20%.”
Weisbrot: Businesses have three to six months to adopt AI or lose
“I think you've got three months, maybe six months to get something into place. Otherwise you're going to lose it. You're going to lose the race. I think that's, it's moving that fast.”
Lechner: Urgent daily firefighting prevents leaders from adopting strategic AI tools
“I think what a lot of people struggle, struggle with is this, like, tension between, like, the urgent and the important, and a lot of these urgent problems that come up at work we're just used to responding. And important work, like, you know, how can I, you k…”
Lechner: Monthly employee AI build presentations drive adoption better than mandates
“It's hard to be top down and tell your employees you need to be using AI. You can't really control what they're doing, and so what they proposed is at the end of each month, everyone needs to either, like, build a new workflow or a product or something using S…”
Lechner: Custom AI coach bots are inevitable and already happening
“There's definitely a fear there around like, yeah, if I just, you know, someone just creates like a custom GPT of like, you know, a couple of conversations with the coach, or they pull in like books and, you know, teachers and whatever things that have inspire…”
Weisbrot: Acquaintance closed over $500k using a custom sales GPT
“Well, the guy I was telling you about, he made a sales master GPT that he uses for himself. And I'm pretty sure he loaded some documents into it. I don't know what, but he said it helped him close over half a million and in contracts and in like a few weeks ju…”
Lechner: Public thought leadership is sufficient to build effective AI models
“If you just compile, I guess you don't need their, like, IP necessarily to create something that's maybe not perfect, but good. A lot of people who spend a lot of time sharing their thought leadership who are doing interviews, I mean, there's nothing, you can …”
Weisbrot: ChatGPT Struggles to Process Multi-Episode Video Content in Post-Production
“Post-production, the only real value you get is like in transcription and summarization. But if you try to feed it a list of episodes, it struggles to like grab things and tell you about them. At least ChattoPT is not good at it yet.”
Weisbrot: Clients appreciate deliverables framed as 'included' rather than 'free'
“Cause if it's free, then it doesn't, then people don't appreciate that work. But if it's included, they appreciate it because wow, that must've been a lot of work for you to get this for me.”
Weisbrot: Domain-specific AI assistants on websites outperform general ChatGPT twins
“So I feel like instead of having something inside of chat to PT, Having something that's specific is better because then you need me for this. Okay, fine. You need me for the cost cutting business. Okay, fine. You need me for this thing. Okay, fine. So I can s…”