Episode Overview

Podcast: Proven Podcast
Host: Charles Schwartz
Guest: Richard White (Founder & CEO, Fathom)
Episode: The Surprising Future of AI
Date: October 15, 2025

This episode explores the realities and future trajectory of artificial intelligence with Richard White, CEO of Fathom. White shares candid insights drawn from his experience building one of the leading AI companies in the productivity space, providing a grounded perspective on AI's promise—and pitfalls—for business, society, and individual workers. The conversation cuts through hype to examine both the existential hopes and practical headaches facing AI practitioners and end-users, offering strategic advice for entrepreneurs, employees, and leaders navigating rapid technological upheaval.

Key Discussion Points & Insights

1. The Reality Behind AI Success Rates (00:29–06:14)

• Richard's background: Founder/CEO of Fathom, market-leading AI note-taking tool.

  • "We've basically built an AI that...so you can just focus on your conversations and not doing a bunch of kind of data entry work." (00:34)

• Harsh realities: Building with AI is fundamentally less predictable than traditional software.

  • “AI has completely upended how we think about building software... It's more like R&D now, whereas before it was more of a manufacturing process.” (01:21)
  • Fathom experiences a 60% failure rate for AI initiatives; an MIT study cites 95% as industry average. Many AI experiments don’t meet the intended outcome.
  • "It's easy to get the AI to spit out something. Our part is getting it to spit out the right thing." (02:49)

• Judging AI output ("Always Incorrect"): It’s often binary with old software—now, subjective quality is central, and that’s hard to measure.

• Vivid analogy: Comparing a 60% AI failure rate to boarding a plane with such odds: "I want to get on a plane that had a 60% failure...I mean, I would get married because that's a 62% failure ratio." (02:34)

2. The Workflow Overhaul and "Jenga" QA Method (04:43–07:54)

• AI is easier for individuals than teams: Personal users can edit AI output; at scale, lack of oversight creates risks.
• Fathom's approach: “All they do all day is play what I call kind of like an AI version of Jenga...pushing on the blocks, finding models that work for specific use cases.” (05:07)
• AI products and models behave inconsistently and have unpredictable shelf lives.
• When evaluating vendors: “We make every vendor give us a 90 day pilot for AI and have a whole team that QAs it. When we don't do that, it almost never works.” (05:46)
• Onto the "LLM treadmill": Upgrades happen so often that maintenance eats into development, as older models often lose support. (06:38)

3. The Pace of Change and Startups vs. Incumbents (09:26–12:12)

• Big companies lack the muscle: AI forces a new paradigm, and large incumbents are often unprepared.
• Opportunity for startups: “The incumbents are completely out of their depth in how to build software in this new era.” (10:02)
• Change cycles have shortened—what took 10 years is now possible in two; volatility is the new norm.
  • “It's like there's so much psychological change that just undoes, you know, if you valuations for SaaS businesses.” (23:09)

4. AI’s Macro Trajectory: Fire or Fire Hazard? (12:12–15:50)

• AI's impact: "This is the greatest technological shift of my lifetime. Bigger than mobile, bigger than social." (12:12)

• The path to AGI: Split among experts—will we plateau, or see another leap? Latest models show diminishing returns.

• Market volatility: Some companies skyrocket to $1B, then vanish—opportunities are immense, but ephemeral.

• Wealth and organizational transformation: The prospect of a billion-dollar company run by a single person is now plausible.

  "Sam Altman talks about the first billion dollar company with a single person. I think that's highly possible." (15:50)

5. Adaptation Strategies for Entrepreneurs and Employees (17:11–22:36)

For Entrepreneurs:

• Best time for niche, application-layer businesses: Faster, cheaper, and focused solutions win.
  • “You don’t need a CS degree or a team of six engineers...you can really tailor the stuff to specific use cases.” (17:11)
• Build in-house, but expect tools to have a short shelf life—six to nine months—before needing to buy or rebuild.

For Employees:

• The old model is gone: Degrees may not buffer you against disruption.
  • “Your fear is well founded, first of all...What you’re seeing when you look at college enrollment is down, trade school enrollment is up.” (21:20)
• Future belongs to those who can augment their own jobs—or help companies improve AI success rates.
• Tangible advice: Trade skills will surge, and expertise in integrating or managing AI is increasingly valuable.

6. Boardroom Fears & Society-Wide Impacts (23:09–26:40)

• Rapid disruption: Boardrooms are bracing for existential questions—AGI, regulatory risks, global competition.

• Societal adaptation: Reference to "Manna" by Marshall Brain—a parable on possible AI-driven futures.

  • “Everyone’s busy fighting, trying to put the genie back in the bottle. The genie’s not going back.” (24:11)

• National and geopolitical competition: The AI arms race between the US and China is a factor in regulatory politics.

• Societal fear: Paths are open—could go dystopian or utopian; vigilance and education are key.

  "I think we would be foolish...This feels like the critical moment in human civilization." (25:54)

7. The Hopeful Side: Positive AI Outcomes (26:40–34:09)

• Medical breakthroughs: AI makes preventive healthcare affordable, enables earlier detection, and accelerates research.

  • “You don’t have to be rich to get life-extending care well ahead of some acute medical crisis.” (27:18)

• Automation in public safety: Self-driving cars could reshape cities, reduce fatalities, and free urban space.

• The new application of tools: Success now depends on adaptability; expectations and operational models must change.

  “Having that adaptability was vitally important.” (32:16)

• Workflow examples: Tools like Gamma can generate hundreds of slides in a week, changing the nature of presentations.

  "Are they gonna be exactly what I had before? No, no. But...you have to be flexible." (33:05)

8. Toolkits, Secret Sauce, and Fun (30:53–34:09)

• No secret Silicon Valley tools: Most use ChatGPT, Magic Patterns, Gamma, etc.—what matters is how you compose and string them together.
  • "At the high end…you actually end up using multiple models…from different providers in that pipeline.” (31:04)
• Fun matters: Team delight in AI-generated imagery and building playful branding—“Not to be fun is not to be discounted in the workplace.” (34:09)

9. Planning and Leading in the Age of AI (38:14–41:36)

• 90-day planning cycles: Anything beyond 90 days is guesswork due to rapid industry shifts.

  • “In a lot of companies, planning is this art of self-deception...it’s important to have hypotheses about the future…but then react locally.” (38:15)

• Leading remote and high-trust organizations: Optimize not for size, but for trust and autonomy (Dunbar’s Number—~150).

  “You should trust by default. You didn’t want to trust them by default, you shouldn’t have hired them.” (42:14)

• AI’s role in knowledge management: AI tools can query years of meeting data, surfacing insights no human could process alone.

10. Organizational Culture and Hiring (42:14–47:43)

• High-trust over low-trust in remote teams: Success comes when teams trust by default and give autonomy.
• Quick to cut misfits: Misalignment should be clear within 60–90 days, though a good hiring pipeline should minimize failures.
• Hiring litmus test: “Would you trust this person to feed your children? If you can’t say yes to that, then you have failed in the hiring process.” (44:22)

11. Personal Mission, Politics, and the Next Chapter (47:43–49:55)

• Personal blinds and passion: White’s entrepreneurial superpower is finding new obsessions—he’d get excited about plumbing if needed.
• Leadership in public life: Encourages people of good judgment to enter politics, despite its negative reputation.
  • "Politics not enough. I think people of high character, good judgment go into politics because they judge it to be ev negative. And it is ev negative. But that's not why you do it, right?" (48:44)

Notable Quotes

• On AI’s unpredictability:

  "It's like, it's not as predictable as what we had before this in SaaS." (08:45, Richard White)

• On market volatility:

  "Companies go from zero to a hundred million, a billion in revenue in two years...and then go back down to zero two years later." (12:12, Richard White)

• On societal impact:

  "You do not have the luxury to sit on the sidelines. Either you're going to be panhandling or you're going to embrace AI because this is just, this is what it is. This is electricity." (16:08, Charles Schwartz)

• On job transformations:

  "If you can be an expert in replacing your own job with AI, that gives you a job over the next couple years." (22:36, Richard White)

• On boardroom fears:

  "It used to be you build a software company and usually at least 10 years before someone really disrupted you, and then now it's like five years. Pretty soon it'll be two years." (23:09, Richard White)

• On regulatory and geopolitical competition:

  "There's an arms race between us and China around AI." (24:11, Richard White)

• On adaptability and workflow:

  "When you pick a new tool, you have to understand what you used to do, how you used to operate, will also have to change." (32:16, Charles Schwartz)

• On optimism despite uncertainty:

  "I still wake up every day feeling pretty optimistic. Yeah, I look inside my window, feel less optimistic, but like, I feel like we'll get there." (41:36, Richard White)

• On remote company culture:

  “You should trust by default. You didn’t want to trust them by default, you shouldn’t have hired them. But you should direct them by default. You should give them room to run.” (42:14, Richard White)

Timestamps for Key Segments

• AI's failure and unpredictability: 01:21–04:06
• "Jenga" QA and Model Upgrades: 05:07–07:54
• Challenges for big companies: 09:26–10:02
• Five-year industry outlook / AGI: 12:12–15:00
• Rise of single-person, multi-million dollar companies: 15:50–16:08
• Advice for entrepreneurs: 17:11–19:06
• Advice for employees & societal impact: 19:06–22:12
• Boardroom anxieties & AI regulation: 23:09–25:28
• Positive impacts & medical breakthroughs: 27:18–29:14
• Tools and workflows in the AI age: 30:53–34:09
• Leadership, planning, and remote orgs: 38:14–41:36
• Hiring, trust, and culture: 42:14–47:43
• Personal mission, passion, and service: 47:43–49:55

Memorable Moments

• Comparing AI failure rates to marriage and airplane crashes. (02:34)
• "AI Jenga": The daily process of testing, abandoning, and retesting models to find what works. (05:07)
• "The genie’s not going back in the bottle." (24:11)
• The EM-dash and emoji saga: The chat’s recurring frustration with GPT outputs. (36:28, 37:19)
• Imagining a future where a single person runs a $1B company: (15:50)
• Passion for building and willingness to get excited about plumbing as a metaphor for adaptability. (47:43)

Tools & Practical Takeaways

• Plan in 90-day increments, not years.
• Demand 90-day pilots from AI vendors and rigorously QA them.
• Favor small, high-trust teams—trust by default, move quickly to cut misfits.
• Don’t expect tools or workflows to stay static—adapt rapidly.
• AI is a career and business accelerant for those at the application layer.
• Fun and creativity with AI tools boosts morale and engagement.

Closing Note

Richard White’s frank, nuanced outlook underlines that AI is both existentially disruptive and full of opportunity—for those able and willing to adapt at every level, from individuals to boardrooms. The conversation leaves listeners with a sense of urgency and agency: the future isn’t guaranteed, but isn’t doomed—if we act, adapt, and keep questioning the path forward.