Latest Interviews

Culture and Revolution

Ben Horowitz, Toussaint L'Ouverture, Tanya Cushman

This presentation defines organizational culture as the collective behavior exhibited when leadership is absent, arguing it must be deliberately engineered rather than left to self-generate. Through historical analysis of Toussaint Louverture's successful reprogramming of Haitian slave society and modern case studies involving leaders like Steve Jobs and Reed Hastings, the speaker demonstrates that culture is shaped by recruitment, shocking rules, and prioritization decisions. The outcome of these deliberate actions determines whether an organization survives high-stakes environments or suffers economic collapse, establishing cultural transformation as a critical, transferable skill for leadership.

A Whirlwind Tour Through Tech Trends in China

Connie Chan

China's mobile-first ecosystem leverages extreme urban density and a super-app dominance by WeChat to bypass legacy credit infrastructure, enabling seamless digital payments and unified social commerce through QR codes. This environment has spawned a unique influencer economy where "Wang Hong" figures manufacture and sell products directly to millions, while live streaming and gamified engagement models drive billions in revenue without traditional app silos. Consequently, global brands are compelled to adapt strategies that integrate communication, payments, and entertainment into singular mobile experiences rather than replicating fragmented Western models.

The End of Cloud Computing

Peter Levine

The event forecasts the obsolescence of centralized cloud computing in favor of edge intelligence, where autonomous devices like self-driving cars and drones become decentralized data centers to ensure real-time processing speed. This architectural shift redefines the cloud as a model training hub that aggregates data from trillions of sensors to update local machine learning algorithms through a continuous feedback loop. Consequently, the industry faces a massive market expansion into distributed IoT ecosystems, driving a fundamental transformation in programming languages, supply chain economics, and the operational roles of IT leaders.

Quantum Computing: A Primer

Richard Feynman, Frank Chen

Pioneered by Richard Feynman's insight into natural quantum processes, this technology utilizes superposition and linear algebra to solve intractable mathematical problems like optimization and simulation that exceed classical computing capabilities. Current hardware, constrained by superconducting cooling near absolute zero and brief coherence windows, is rapidly evolving alongside critical algorithms such as Grover's and Shor's to accelerate deep learning, revolutionize quantum chemistry for ammonia production, and threaten existing cryptographic standards. As major university labs, corporate R&D centers, and state-sponsored entities converge on hardware and software development with Bell Labs-era investment levels, the field is approaching a maturity ready to drive substantial startup formation and global economic impact.

Marketplaces

Jeff Jordan

Andreessen Horowitz is capitalizing on a "Marketplace 2.0" renaissance by investing in ventures like Airbnb and OfferUp that leverage smartphone ubiquity and two-sided network effects to disrupt traditional retailers and stagnant classified models. These platforms achieve superior unit economics by substituting community labor for physical inventory, thereby generating 80–90% gross margins compared to the ~25% typical of one-to-many retailers. By enabling rapid scaling without heavy capital expenditure for warehousing or logistics, these networked businesses create defensive moats that allow high-margin marketplace revenues to subsidize broader ecosystem growth.

Chris Re

Chris Re

Deep Dive is an open-source system that transforms unstructured data like emails and PDFs into structured SQL databases by unifying extraction and cleaning into a single statistical inference problem. Utilizing "Hogwild!" algorithms that relax consistency constraints, the architecture achieves up to 1,000 times the speed of competitors while surpassing human accuracy in specific domains. The technology has already enabled the rapid construction of a global fossil record with a single student and facilitated real-world arrests in human trafficking cases through automated pattern detection.

Building Your Board | Glenn Kelman

Glenn Kelman, Selina Tobaccowala, James Slavitt, Jeff Markowitz, Mark Singer, Emily Melton, Austin Ligon, Reed, Julie Bornstein, Bob Milad, Jason, Janie, Scott, Chris Roski

Redfin CEO Peter van de Ven transformed his company's governance by shifting board relations from adversarial to collaborative, implementing strict communication protocols and off-site retreats to foster honest dialogue and strategic alignment. To ensure operational efficiency, the firm recruited board members based on peer-scale experience and mission alignment rather than corporate prestige, resulting in employed agents who are three times more productive than industry standards. This deliberate restructuring of founder-board dynamics and executive accountability has enabled Redfin to maintain higher Net Promoter Scores while generating hundreds of millions in annual revenue.

How to Build an Exchange

Claudio, Ron, Pete

A US equities matching engine utilizes a deterministic, single-instance architecture to process three million messages per second while ensuring fair, latency-critical order execution across thousands of symbols. The system employs C++ and OCaml on a dedicated machine to replace complex distributed consensus with high-speed, single-threaded logic, allowing state machines to recover from failures within 30 to 60 seconds via message log replay. This design guarantees regulatory reproducibility and financial reliability by eliminating consensus overhead, enabling clients to achieve precise atomicity and global risk enforcement through simplified, high-throughput message sequencing.

Why Functional Programming Doesn't Matter

John Hughes, Tony Hoare

Jane Street, a high-frequency proprietary trading firm executing millions of daily trades, attributes its engineering success to OCaml's expressive static types rather than traditional functional features like laziness or strict purity. This type system enforces correctness by eliminating null pointer exceptions, forcing logic updates at compile-time to prevent bugs, and encoding business invariants to exclude impossible states. Consequently, the firm prioritizes predictable performance and verifiable clarity to protect its capital against errors, viewing advanced type systems as a more critical asset than higher-order functions or declarative purity.

MIT 6.S094: Deep Learning for Human-Centered Semi-Autonomous Vehicles

Researchers are collecting billions of high-speed video frames from semi-autonomous Teslas to train deep learning models that detect critical driver metrics such as body pose, gaze direction, and cognitive load. By shifting from fully supervised to semi-supervised annotation strategies, the team achieves an 84-fold reduction in human effort while accurately identifying micro-saccades and emotional cues to overcome current privacy and trust barriers. This data-driven approach aims to replace static crash test assumptions with dynamic occupant monitoring, ultimately enabling vehicles to adapt passive safety systems based on real-time human behavior.

Effective ML 2011 Harvard CS51 Part 2

Jane Street's engineering philosophy prioritizes explicit error handling, strict boilerplate avoidance, and managed complexity to ensure robust high-frequency trading systems. By leveraging OCaml's predictable performance and enforcing rigorous code review protocols alongside small, autonomous teams, the firm balances mathematical precision with human oversight in a hybrid trading model. This approach treats software development as a discipline of clarity and maintainability, where static typing and disciplined mutation control enable reliable operation in volatile markets.

Atul Butte | Driving Public Health: A Trillion Points of Data

Atul Butte

Addressing the exponential growth of biomedical data, the presentation details UCSF's initiative to consolidate electronic records from 14.1 million patients into a unified warehouse that shifts disease tracking from code-based to molecular-based systems. This infrastructure enables low-cost computational predictions, such as the rapid repurposing of imipramine for lung cancer, which accelerated a clinical trial from concept to authorization within fifteen months. By creating dynamic maps of patient disease progression, the project aims to redefine accountable care through continuous, population-level monitoring rather than isolated clinical encounters.

Dr. Eric Kandel, Nobel Prize-winning Neuroscientist: Talks at GS Session Highlights

Dr. Eric Kandel

A comprehensive analysis of memory systems distinguishes between localized brain functions, such as Broca's and Wernicke's areas, and the dual nature of memory identified in patient H.M. as either explicit or implicit. Research into storage mechanisms reveals that long-term memory relies on the CREB transcription factor and RBAB48 protein to physically grow synaptic connections, while age-related decline stems from a specific deficit in the dentate gyrus distinct from Alzheimer's pathology. This decline is reversible through interventions like injecting osteocalcin or engaging in physical exercise, which boost necessary protein levels to restore cognitive function in aged models.

Jessica Livingston Speaks at Female Founders Conference 2015

Jessica Livingston

The event details Y Combinator's evaluation criteria, which prioritize founder determination and character over demographics, resulting in an acceptance rate for female founders that mirrors the applicant pool. It outlines the organization's origin as a non-profit experiment that evolved into a profitable model supported by a familial alumni network designed to mitigate critical issues like co-founder disputes. Key figures emphasize their unique "social radar" interview process and the ongoing commitment to open-source investment tools and founder education to sustain the community's integrity.

Danae Ringelmann at Startup School SV 2014

Danae Ringelmann

Indiegogo founder Denae Robinson launched the world's largest global crowdfunding platform in 2008 to democratize access to capital after witnessing systemic barriers prevent entrepreneurs and artists from securing funding. Facing a prolonged fundraising drought that forced the team to forgo salaries for four years, Robinson pivoted the company to an open, internet-based model featuring flexible funding options that enabled breakthroughs like uBiome's market validation and the Hour of Code initiative's massive educator mobilization. Guided by a culture of fearlessness and diversity, the organization has established itself as a meaningful tool for solving fundamental problems, earning recognition from figures like Bill Gates and the White House for its societal impact over immediate financial returns.