Latest Interviews
Showing 1381–1395 of 1,630 interview transcripts.
Clear all filters- Lex Fridman1h 11m
Rajat Monga: TensorFlow | Lex Fridman Podcast #22
Launched by Google Brain in 2011, the TensorFlow ecosystem has evolved from a proprietary deep learning library into a globally adopted platform with over 41 million downloads and 1,800 contributors. The framework is currently transitioning to version 2.0, which defaults to eager execution and unifies its API around Keras to resolve developer confusion while preserving backward compatibility for enterprise systems. Driven by a distributed governance model and competition from alternatives like PyTorch, the project aims to democratize machine learning across diverse hardware, from mobile devices to cloud TPUs, by simplifying model development and addressing enterprise data organization challenges.
- Milken Institute57 min
Right or Left: Who's Best for Freedom?
Josh Barro, Kenneth Hersh, Rich Lowry, Gene Sperling, Felicia Wong
In a debate contrasting FDR's positive liberties with classical negative freedoms, Felicia Wong and Gene Sperling argue that true liberty requires protection from want and fear, while Rich Lowry and Kenneth Hirsch contend that freedom is strictly defined by the absence of government coercion. The panel further explores how economic policies like the Affordable Care Act and anti-discrimination laws either expand individual agency or threaten consumer choice, while noting that shifting political narratives now force conservatives to defend liberty against a backdrop of authoritarianism and rising public demand for government intervention. Ultimately, the discussion highlights a deep ideological divide over whether freedom entails securing material security and dignity through state action or preserving it through limited government and market competition.
- Milken Institute52 min
Inside a Blockbuster Franchise: "Avengers: Endgame"
Julia Boorstin, Anthony Russo, Joe Russo, David Levinson, John McCutchan
Directors Joe and Anthony Russo leveraged a unique back-to-back production strategy and a global day-and-date release to propel *Avengers: Endgame* to a record-breaking $1.2 billion opening weekend. Their collaborative "mastermind" approach, which integrated scientific consultation and disciplined ensemble management, transformed the film into a universal cultural phenomenon while reaffirming the distinct value of the theatrical experience. Building on this success, the Russo brothers are now expanding their AGBO production company to champion artist ownership and emerging talent with projects like *Cherry*, all while maintaining a format-agnostic commitment to both streaming and cinema.
- Y Combinator56 min
Analyzing Billions of Transactions to Understand Consumer Behavior - Michael Babineau and Kevin Hale
Michael Babineau, Kevin Hale, Craig Cannon, Mike
Co-founded by Mike and Lillian, the company delivers a financial analytics platform that empowers investors and corporations to analyze over 50 billion credit card transactions to solve complex business intelligence problems. With a diverse team of 60 data scientists and engineers, the firm processes unstructured data for entities ranging from startups to major retailers like Stitch Fix and Amazon, transforming raw transactional noise into actionable strategic insights. Supported by strategic funding from Goldman Sachs and Citi, the platform achieved 150 clients through organic growth by offering a customizable toolset rather than static reports, allowing users to generate their own unique information edges.
- Milken Institute1h 1m
Investing with Vision: A Conversation With SoftBank's Rajeev Misra
Rajeev Misra, Mike Milken, Michael Milken
SoftBank Vision Fund has deployed $80 billion across 80 portfolio companies to aggregate fragmented markets through AI-driven platforms and preferential debt financing. Led by Rajiv Mehta, the fund targets 30% long-term returns by applying a minority stake strategy to "fearless" founders while leveraging SoftBank's ecosystem for global expansion. With a second fund planned for immediate launch, the organization aims to capitalize on the transition from internet growth to an AI revolution that restructures industries like transportation and hospitality.
- Y Combinator50 min
Vidit Aatrey on Building Meesho, India's Top Reselling Platform, with Adora Cheung
Founded in 2015, Meesho is a social commerce platform founded by CEO Adit Atrai that empowers nearly 500,000 users, predominantly women from Tier 2 and 3 cities, to generate supplemental income by reselling goods via WhatsApp and Instagram without requiring inventory or working capital. The company operates by leveraging existing social trust to overcome India's market barriers, a strategy supported by Atrai's direct engagement with top users and a focus on indigenous product development rather than replicating foreign models. By shifting from a supply-focused tool to a peer-to-peer marketplace in 2016, Meesho has established a competitive moat that enables over 700 employees to facilitate transactions for non-branded goods while addressing unmet needs in the non-metro demographic.
- a16z49 min
10+1 Lessons from Serial Entrepreneur Justin Kan
Fourteen-year veteran entrepreneur Justin Kahn applies his extensive experience from founding and exiting ventures like Twitch and Justin.tv to lead Atrium, a technology-enabled law firm designed to offer speed and transparency for startups. Transitioning from high-market-risk consumer apps to execution-focused B2B models, Kahn leverages a massive Series A round to signal trustworthiness while advocating for a "conscious company" culture built on empathy and radical responsibility. His approach emphasizes that sustainable success requires treating startup building as a multi-decade marathon rather than a sprint, integrating rigorous personal development with strategic business growth.
- Y Combinator49 min
Monetizing Podcasts and Newsletters - Chris Best of Substack and Jonathan Gill of Backtracks
Chris Best, Jonathan Gill, Craig Cannon, Jack Ryder, Debdut Mukherjee
Substack leadership and industry experts analyze a shifting podcasting landscape where direct subscription models are gaining traction against traditional advertising, leveraging strategies like micro-pricing and referral incentives to monetize niche audiences. The discussion highlights that success relies less on mass download metrics and more on consistent publishing, high-quality audio storytelling, and the ability to measure engagement through direct response data rather than obsolete download counts. Ultimately, the panel predicts a future ecosystem where creators maintain direct audience relationships through independent platforms, blending ad-supported mass content with paid subscription niches to counter the consolidation risks posed by major tech giants.
- Lex Fridman1h 13m
Chris Lattner: Compilers, LLVM, Swift, TPU, and ML Accelerators | Lex Fridman Podcast #21
Chris Lattner, the creator of LLVM and Swift and former lead of Tesla's Autopilot software, currently directs compiler infrastructure initiatives at Google including TensorFlow, TPU accelerators, and the emerging MLIR project. He details how the open-source LLVM community unites competing giants like Apple, NVIDIA, and Intel by sharing expensive optimization layers while pioneering techniques that apply machine learning to solve complex register allocation challenges. His career narrative highlights a strategic shift in the industry toward integrated automatic differentiation and dynamic compilation, fundamentally reshaping how diverse languages interact with modern hardware from mobile devices to neural network accelerators.
- Y Combinator46 min
HubSpot CEO and Cofounder Brian Halligan with Wufoo Cofounder Kevin Hale
Brian Halligan, Kevin Hale, Craig Cannon, Dharmesh
HubSpot co-founder Brian Halligan details the company's evolution from a service-based consultancy to a software giant, highlighting their pivotal pivot to a freemium model and the strategic decision to maintain a $250 monthly price point for six years. He explains how the organization overcame high early churn by shifting executive compensation toward Net Promoter Score and reorienting its go-to-market strategy toward content assets rather than traditional sales tactics. Halligan further emphasizes the necessity of consumer-grade user experiences in B2B markets and advises modern founders to build their own content platforms while avoiding the "groupthink" of hiring co-founders with identical skill sets.
- a16z50 min
Seven Trends in Blockchain Computing (Spring 2019)
Olaf, the first employee of Coinbase, details the third wave of blockchain evolution driven by WebAssembly integration and Proof-of-Stake consensus to resolve current developer bottlenecks and security flaws inherent in older protocols. This upcoming phase emphasizes superior technical infrastructure, including built-in on-chain governance and sub-second finality, while shifting industry metrics from user counts to composability and open-source development speeds. As the sector transitions from a nascent launch phase to mature application development, these upgrades aim to facilitate complex use cases like decentralized finance and AI data markets without relying on ad-revenue models or arbitrary platform control.
- Y Combinator1h 4m
Software Engineer Turned Youtuber - Jarvis Johnson
Jarvis Johnson, Craig Cannon, Nathan Allebach, Taylor
Engineering manager and content creator Jarvis Johnson transitioned from a full-time career at Patreon to independent YouTube production after a viral subscriber surge validated his strategy of balancing technical deep-dives with broader industry critique. Leveraging lessons from Silicon Valley and his time managing teams, Johnson now advocates for "clickbait with positivity" while warning against the rise of soulless content farms that threaten platform authenticity. His journey highlights a shift from grade-driven multitasking to focused, interest-driven learning, aiming to provide accessible role models for aspiring technologists navigating the software industry's evolving landscape.
- Milken Institute1h 6m
Part 1: A Conversation With John Hickenlooper | Part 2: A Conversation with Seth Moulton
John Hickenlooper, Seth Moulton, Jerry Baker, Christina Bellantoni
The Milken Institute Global Conference featured two presidential conversations with Democratic candidates Governor John Hickenlooper and Representative Seth Moulton, who focused on unifying the party to defeat President Trump. Hickenlooper emphasized restoring middle-class economic stability through antitrust enforcement, tax reforms on wealth, and state-led social policies, while Moulton advocated for a generational shift in leadership driven by veteran healthcare reform, green infrastructure investment, and constitutional impeachment debates. Both candidates addressed critical national challenges including climate change, gun control, and economic restructuring, aiming to present a cohesive platform that addresses deep national divisions.
- a16z1h 15m
Clippy: The Unauthorized Biography
Chloe Condon, Steven Sinofsky, Frank Chen, BillG
Led by engineer Sam Hobson, a Microsoft team developed the Office Assistant, or Clippy, to bridge the gap between complex 1990s software and home users by applying social actor theories and natural language processing. Despite consulting Disney animators and facing cultural localization hurdles, the feature generated significant user frustration through intrusive interruptions, leading to its official retirement in Office XP by 2001. Although the specific implementation failed, the project established a legacy of intent-based interaction models and remains a prominent cultural meme that highlights the ongoing challenge of designing usable intelligent agents.
- a16z53 min
Towards a Decentralized Storage Network for the Blockchain Computer
Ali Yahya, Ben Fisch, Bruce Schneier
Decentralized storage networks solve the critical challenge of verifying file integrity across permissionless nodes by replacing institutional trust with cryptographic protocols like Proof of Retrieval and Proof of Replication. These mechanisms transform underutilized personal device capacity into a censorship-resistant infrastructure where economic incentives compel nodes to dedicate actual disk space rather than simulating storage. By enforcing high tightness constraints that disproportionately burden large centralized providers, the system ensures a scalable, fault-tolerant foundation for future decentralized applications.