Divot

McKay Coppins is an award-winning Atlantic journalist and New York Times bestselling author of Romney: A Reckoning. After gaining access to Mitt Romney’s private journals, he turned to America’s sports betting boom. The Atlantic gave him $10,000 to gamble through an NFL season. He knew the risks. Within weeks, he was hiding bets from his children and checking DraftKings at church.
In this episode of Divot with Derek Andersen, McKay reveals how investigating gambling became a test of his own self-control. He explains how betting apps keep customers coming back, why sportsbooks can limit consistent winners, and how he convinced himself he could beat the odds. They explore prediction markets, teenage gambling, and what happens when watching sports becomes a reason to keep betting.
The conversation also goes inside Washington, where McKay describes how protecting a position can replace the desire to serve. He shares what Romney’s journals revealed about private doubts, why he left Newsweek for BuzzFeed, and the advice that changed his career: choose the interesting job over the impressive one. His final measure of success comes from the people who know him best, his wife and children.
Follow McKay - X (Twitter): https://x.com/mckaycoppins
Follow Derek - X (Twitter): https://x.com/derekjandersen
Episode Partners:
Google for Startups
If you’re building AI agents, Google Cloud’s new Startup Guide to AI Agents covers choosing the right model, managing memory, and building agents you can control and scale. Download the guide: https://goo.gle/agent_guide
Timestamps:
02:50 — How Political Power Changes People’s Priorities
14:07 — Inside Mitt Romney’s Private Journals
22:59 — How Sports Betting Became a Casino in Your Pocket
26:40 — The App Features That Keep Bettors Coming Back
29:37 — Why Sports Leagues Embraced Gambling
36:30 — Prediction Markets: What Happens When Everything Is a Bet?
42:26 — What Parents May Be Missing About Teenage Gambling
44:10 — The Atlantic’s $10,000 Sports Betting Experiment
46:34 — He Knew the Risks—and Still Got Hooked
48:53 — Why Sportsbooks Can Limit Consistent Winners
54:50 — Choosing Meaningful Work Over an Impressive Job
1:04:39 — The Measure of Success His Résumé Can’t Capture
Follow Divot:
YouTube: https://www.youtube.com/@Divotpod
X (Twitter): https://twitter.com/divotdotorg
Instagram: https://www.instagram.com/divotpod
TikTok: https://www.tiktok.com/@divotpod
LinkedIn: https://www.linkedin.com/company/divotdotorg
Hashtags:
#SportsBetting #McKayCoppins #GamblingAddiction #TheAtlantic #DraftKings #FanDuel #PredictionMarkets #MittRomney #Psychology #SelfControl #Journalism #CareerAdvice #DerekAndersen #DivotPodcast

Patrick Drake is a chef, entrepreneur, and former Clifford Chance lawyer who traded corporate law for professional kitchens. He worked at Heston Blumenthal’s The Fat Duck and co-hosted a cooking series picked up by National Geographic. His career change began in secret: helping chefs in his law firm’s kitchen at lunchtime and working in London restaurants after leaving the office.
In this episode of Divot with Derek Andersen, Patrick explores why a career that looks successful can still feel wrong. He shares how he turned an unlikely ambition into practical steps, why he stopped talking about his plans, and the brief pitch that opened a door at The Fat Duck. The conversation challenges the pressure to live up to other people’s expectations—and asks what it takes to choose your own direction.
When a fire destroyed his home, Patrick faced a different challenge: accepting help. He opens up about vulnerability, fatherhood, and the morning pages practice that helps him untangle his thoughts. A man who had told himself “no one’s coming to rescue me” began to rethink what it meant to be self-reliant. His measure of success became the community around him and the people who show up when life gets hard.
Follow Patrick - X (Twitter): https://x.com/patrickdrake
Follow Derek - X (Twitter): https://twitter.com/derekjandersen
Episode Partners:
Google for Startups
If you’re building AI agents, Google Cloud’s new Startup Guide to AI Agents covers choosing the right model, managing memory, and building agents you can control and scale. Download the guide: https://goo.gle/agent_guide
Timestamps:
02:21 — Leaving Corporate Law to Become a TV Chef
05:43 — When Talking About Your Plans Replaces Taking Action
10:53 — The Cooking Show Pilot That Went Nowhere
15:05 — Building a New Career Before Quitting the Old One
19:32 — The Pitch That Opened Heston Blumenthal’s Kitchen
22:50 — How a YouTube Video Led to a Television Series
25:51 — What The Fat Duck Taught Him About Excellence
29:40 — The House Fire That Changed His Definition of Success
35:03 — Why Asking for Help Was His Hardest Lesson
39:29 — When to Keep Pushing—and When to Walk Away
42:12 — How Morning Pages Help Him Untangle His Thoughts
44:55 — Fatherhood, Friendship, and a New Measure of Success
Follow Divot:
YouTube: https://www.youtube.com/@Divotpod
X (Twitter): https://twitter.com/divotdotorg
Instagram: https://www.instagram.com/divotpod
TikTok: https://www.tiktok.com/@divotpod
LinkedIn: https://www.linkedin.com/company/divotdotorg
Hashtags:
#PatrickDrake #HelloFresh #CareerChange #HelloFreshUK #Entrepreneurship #FounderMindset #HestonBlumenthal #TheFatDuck #Reinvention #MorningPages #Fatherhood #Community #DerekAndersen #DivotPodcast

Robert and Richard Rizk co-founded Blackbox AI with their brother Roger, taking a coding search engine into enterprise AI infrastructure. The engineers are tackling a problem bigger than model intelligence: how to make AI affordable without giving up control of sensitive data. Their approach combines open-weight models, encrypted inference, and more efficient use of NVIDIA GPUs at scale.
In this episode with Derek Andersen, the brothers explain why the cheapest AI endpoint can carry risks beyond its price, how hosting changes the privacy equation for DeepSeek, and why open-weight models can coexist with OpenAI. They also break down the difference between zero data retention and encryption, and how generating more tokens from the same GPU could unlock wider AI adoption throughout the enterprise.
The conversation goes deeper than technology. Robert and Richard reveal how three brothers turn trust and disagreement into an advantage, why Navy SEAL discipline inspires their culture, and how they approach problems others avoid. Their lesson for founders: choose work worth doing, then tackle the hardest parts one step at a time.
Follow BlackBox.ai - X (Twitter): https://x.com/blackboxai
Follow Derek - X (Twitter): https://twitter.com/derekjandersen
Episode Partners:
Google for Startups
If you’re building a company, join Google Cloud and the DeepMind team on November 4 and 5 for a briefing on Gemini 4 Argon, exploring what it could make possible for your business and how to start building with it. Includes live Q&A. Register your interest at https://goo.gle/gemini4_briefing
Timestamps:
01:05 — AI Inference: Why More Tokens per GPU Matter
04:08 — Open vs. Closed AI: Why Businesses Want Both
09:54 — DeepSeek and Data Privacy: Where Your AI Runs Matters
18:43 — Encrypted AI vs. Zero Data Retention: What’s the Difference?
25:40 — NVIDIA GPUs: Lowering AI Costs Without Adding Hardware
31:06 — Why Smarter Models Still Need Better Infrastructure
38:01 — Three Brothers, One Company: The Blackbox Origin Story
43:14 — How Trust and Disagreement Help Founders Move Faster
47:00 — The Navy SEAL Mindset Behind Blackbox’s Culture
01:01:28 — Overcoming Hard Problems: One Decision at a Time
Follow Divot:
YouTube: https://www.youtube.com/@Divotpod
X (Twitter): https://twitter.com/divotdotorg
Instagram: https://www.instagram.com/divotpod
TikTok: https://www.tiktok.com/@divotpod
LinkedIn: https://www.linkedin.com/company/divotdotorg
Hashtags:
#BlackboxAI #AIInference #AIPrivacy #NVIDIA #DeepSeek #OpenAI #AISecurity #OpenWeightModels #AIInfrastructure #StartupFounders #DivotPodcast

Ivan Burazin is the co-founder and CEO of Daytona, the AI infrastructure company building computers for AI agents. His 16-person engineering team rebuilt its product with a radical rule: no manually written code. With engineers directing up to 15 agents at once, he offers a firsthand look at the future of software development, where human judgment matters more than typing every line.
In this episode with Derek Andersen, Ivan makes a bold case: the next software revolution will be built for agents, not humans. He explains why SaaS could become infrastructure, how AI agents explore multiple paths in parallel, and why GPU access can become a financing problem. His challenge to startup founders: look beyond adding a chatbot and rethink who your software is built to serve.
The conversation goes deeper than technology. Ivan explores why AI could widen inequality, why founders delay hard decisions, and why human relationships still matter. From his repeated failures to his gratitude for family, he shares a different measure of success. It is not about feeling happy every day, but about ending it content that you did something difficult, meaningful, and worth the effort.
Follow Ivan - X (Twitter): https://x.com/ivanburazin
Follow Derek - X (Twitter): https://twitter.com/derekjandersen
Episode Partners:
Google for Startups
If you’re building a company, join Google Cloud and the DeepMind team on November 4 and 5 for a briefing on Gemini 4 Argon, exploring what it could make possible for your business and how to start building with it. Includes live Q&A. Register your interest at https://goo.gle/gemini4_briefing
Timestamps:
01:49 — AI Agents vs. Chatbots: What’s the Difference?
06:09 — Why Daytona’s Engineers Stopped Writing Code
11:49 — How Many AI Agents Can One Person Manage?
18:22 — Inside Daytona: Giving AI Agents a Computer
21:07 — The AI “Multiverse”: Testing Multiple Paths at Once
23:45 — Why AI Infrastructure Needs Split-Second Speed
26:25 — Build for AI Agents, Not Just Humans
28:43 — Salesforce and the Future of SaaS
34:59 — Inside the GPU Wars: The Cost of Scaling AI
45:04 — AI’s Biggest Risk? The Widening Divide
49:10 — Why Ivan Moved to San Francisco for AI
56:49 — Hard Decisions, Resilience, and Defining Success
Follow Divot:
YouTube: https://www.youtube.com/@Divotpod
X (Twitter): https://twitter.com/divotdotorg
Instagram: https://www.instagram.com/divotpod
TikTok: https://www.tiktok.com/@divotpod
LinkedIn: https://www.linkedin.com/company/divotdotorg
Hashtags:
#AI #Superintelligence #IvanBurazin #Daytona #ArtificialIntelligence #SoftwareEngineering #SoftwareDevelopment #AIAgents #AIInfrastructure #Entrepreneurship #FutureOfSoftware #FutureOfWork #DerekAndersen #DivotPodcast

Aditya Agarwal helped build early Facebook and scaled Dropbox’s engineering team from 25 to 1,000. The former Dropbox CTO is now a General Partner at South Park Commons, a venture firm with $2 billion in assets under management whose portfolio includes Baseten, valued at $13 billion in June 2026. His career spans the rise of social media, cloud software, and now artificial intelligence.
In this episode with Derek Andersen, Aditya makes a bold case: the superintelligence era is already here. He explains why AI is reshaping software engineering, recalls a 15-second AI diagnosis he estimates would have taken him hours, and reveals why coding alone is no longer enough. As Big Tech builds the infrastructure, he sees an extraordinary opening for small teams with ambitious ideas.
The conversation goes deeper than technology. Aditya challenges founders to stop following predictable paths and give their biggest ideas room to grow. From early Facebook to the lessons of building startups, he explores curiosity, patience, and the courage to begin without every answer. His message is a challenge to every builder: don’t let what you’ve already achieved limit what you dare to attempt next.
Follow Aditya - X (Twitter): https://x.com/adityaag
Follow Derek - X (Twitter): https://twitter.com/derekjandersen
Episode Partners:
Google for Startups
If you’re building applications with generative media or considering it, Google just released a brand-new technical guide. You can download it at https://goo.gle/genmediaguide
Timestamps:
06:27 — From Oracle to Facebook: Betting on the Unknown
17:23 — Why Contrarian Startup Ideas Face Rejection
18:58 — South Park Commons: Finding What’s Worth Building
29:41 — AI Is Changing the Value of Coding
33:15 — The 15-Second AI Diagnosis That Surprised a CTO
37:15 — Why the AI Revolution Is Still Early
39:43 — Why He Believes Superintelligence Is Already Here
40:39 — Can Small AI Startups Compete With Big Tech?
42:08 — AI Safety: Can We Really Slow Down?
45:26 — Imposter Syndrome: Learning Beats Knowing
46:37 — Beyond Success: What Actually Matters
Follow Divot:
YouTube: https://www.youtube.com/@Divotpod
X (Twitter): https://twitter.com/divotdotorg
Instagram: https://www.instagram.com/divotpod
TikTok: https://www.tiktok.com/@divotpod
LinkedIn: https://www.linkedin.com/company/divotdotorg
Hashtags:
#AI #Superintelligence #AdityaAgarwal #Facebook #Dropbox #ArtificialIntelligence #SoftwareEngineering #AIStartups #SouthParkCommons #SiliconValley #Entrepreneurship #StartupFounders #FutureOfWork #DerekAndersen #DivotPodcast