Get the latest ideas from World Economic Forum.
Plus the best new takeaways about artificial intelligence from other top podcasts — read in minutes, not hours.
or
By continuing, you agree to Podbrain's Terms and Privacy Policy.
Nicholas Thompson is the CEO of The Atlantic and best-selling author of The Running Ground. He has been driving digital transformation at the 169-year-old publication while also setting American records in ultramarathons. Thompson's memoir explores how his running performance dramatically improved in his 40s, from a 2:43 marathon to 2:29, after understanding the mental nature of physical limits.
The conversation covers Thompson's leadership philosophy drawn from running principles, including his 'non-goals' framework for prioritization and how to manage organizational energy like training load. He discusses his complex relationship with his late father, who came out as gay and lived a chaotic life, yet remained Thompson's biggest supporter in running.
Thompson addresses The Atlantic's AI strategy, including their partnership with OpenAI for content licensing and traffic preservation. He explores the broader implications of AI democratization versus concentration of power in a few companies, and shares his daily content creation practice of posting tech videos across social platforms.
From Mental Barriers to Marathon Breakthroughs
The Running Ground chronicles Thompson's counterintuitive improvement from 2:43 marathons in his 30s to 2:29 and American ultramarathon records in his 40s, driven by understanding mental limits rather than just physical training.
A high school track experience revealed how mental barriers work: running what he thought was 5:45 pace for an 11:30 two-mile, he actually ran 5:20 pace because he didn't know how fast he was going - 'my brain would have gotten too scared' - Thompson
Breaking the 2:40 marathon barrier unlocked rapid progression: 'I couldn't get under 2:40, I couldn't get under 2:40 and then I got under 2:40 and within a year it was under 2:30' - Thompson
Pain in running is often mental rather than physical - the body being afraid and thinking 'you're going too fast, you can't handle this, you better slow down' rather than actual physical limitation.
Learning from an Imperfect Father
Thompson's father Scott was a promising politician whom John Kennedy said would be president, but came out as gay and lived a chaotic life that Thompson described as 'my mom was an art historian and my dad runs a brothel in Bali.'
Despite the chaos, Scott Thompson provided unconditional support: 'it's rare to have people in your life who truly want the best for you at every moment' and 'my dad always wanted the absolute best for me' - Thompson
Thompson's grandfather believed success was zero-sum within the family - 'as my father did better, it would make him do worse. Like the universe had only a set amount of success that could be given to Thompson's.'
This experience taught Thompson to work with difficult people: at a job interview, saying 'my father's an alcoholic, but we get along great' helped him get hired by a boss who also drank too much.
The Non-Goals Framework for Leadership
Thompson's 'non-goals' principle: 'what is a thing that is valuable that I do not have time for... then everything I do has to be more valuable than that' as a way to structure priorities.
Developing a video strategy for The Atlantic serves as Thompson's long-standing non-goal - valuable but not economically viable given organizational constraints and other priorities.
Leadership requires understanding organizational capacity: 'how much pressure can they take' and finding the right challenge level, as his mentor David Remnick said about not asking someone to jump 7 feet when they can do 5 feet.
Thompson uses Trello boards to track five daily priorities, next three tasks, and general projects, ensuring focus on 'the really important stuff, building the relationships that matter the most.'
Running as Leadership Training
Running teaches the fundamental principle that 'you do not ever get better without consistent effort. And if you put in consistent effort, you always get better' - applicable to impossible workplace projects.
Training involves strategic rest: running hard 2-3 times per week with recovery days, which Thompson applies to work scheduling - 'you put in you do some really hard stuff but then you have to maybe take it a little easier the next day.'
Learning to distinguish injury from fatigue is crucial: 'if the right leg feels a lot worse than the left leg, there's probably something wrong' versus equal soreness indicating normal training adaptation.
Running without music develops body awareness and coordination: 'you're trying to make your mind more aware of how your body works... you're trying to improve your neurological system' rather than just getting motivated.
The Atlantic's AI Strategy and Digital Transformation
AI companies 'violated our terms of service' and 'probably violated copyright law' by taking content to build competitive models, with most offering nothing in return - some The Atlantic has sued, others they're negotiating with.
OpenAI's partnership provides compensation for data usage and collaboration on AI search that 'respects journalists and drives traffic to you' - a model Thompson wishes other tech companies would follow.
Thompson describes AI as 'a giant magnet' for journalism's needle-in-haystack work, while The Atlantic must adapt subscription paywalls and advertiser reach for an agent-driven world with less search traffic.
As CEO, Thompson focuses on revenue generation through advertising and subscriptions rather than editorial oversight: 'I don't oversee the editorial. I don't even know what we're publishing today.'
AI Democratization vs Corporate Concentration
Current AI power concentration creates geopolitical instability: 'all of the closed source AI power concentrated in the United States' contributes to why 'the US stock market is soaring' and US geopolitical dominance.
Scaling laws require 'billions of dollars to train models' because 'what makes models better is more compute, more data, more everything' - concentrating power in few companies with massive resources.
New AI architectures like 'world models, neural symbolic AI, continuous learning models' could shift the paradigm from needing 'more money' to needing 'a better idea' for democratization.
Audrey Tang's framework distinguishes between the 'vertical race' of major AI companies toward infinity and the 'horizontal race' where civil society builds tools that may eventually change the vertical competition.
Book Recommendations and Daily Practices
For management, Thompson recommends Scaling People by Claire Hughes Johnson as 'a great guide to how to think about how to be a boss and how to have hard conversations.'
Endure by Alex Hutchinson was Thompson's top running book from the past two years, focusing on athletic performance and limits.
Talk to Me by Rich Benjamin, Thompson's best book of the previous year, tells the story of a young man exploring his Haitian heritage and 'how families process and understand pain.'
Thompson creates daily tech videos across platforms reaching nearly 2 million people, filming 2-minute segments about whatever interests him in technology that day, often asking friends to hold his phone in locations like Davos.
From World Economic Forum. Get a note like this from every new episode.