Latest Interviews

Workshop: Master Your Memory and Mind

Jim Kwik

The presentation identifies digital overload, distraction, and "digital dementia" as the three primary forces eroding human performance, urging attendees to replace reactive phone habits with proactive morning routines that prioritize personal agendas. It introduces the MOM and FAST frameworks to train memory and accelerate learning by emphasizing emotional motivation, focused attention, and the transition from passive consumption to active creation. Ultimately, the speaker argues that two-thirds of memory potential is trainable, positioning disciplined self-management as the essential skill required for a workforce navigating ten to fourteen distinct careers.

Effective Programming: Adding an Effect System to OCaml

This presentation details a comprehensive framework for algebraic effects that separates effect specification from implementation by treating operations like concurrency and state as explicit, resumable interruptions. It contrasts this approach with traditional monads to eliminate boilerplate and demonstrates a full conversion of the OCaml standard library, introducing syntax for tracking purity, region-based locality, and effect polymorphism directly in the type system. The work culminates in a runtime model where effect handlers manage heap-allocated call stacks while future extensions aim to enforce region safety and support multi-handler compositions for complex scenarios.

Andy Bromberg - Startup Investor School Day 4

Andy Bromberg

Startup Investor School and Andy Bromberg's presentation analyze the evolution of venture capital from the 1940s to the modern "everyone is an angel" era, highlighting how decreasing entry costs and regulatory shifts like the 2012 JOBS Act reshaped early-stage funding. The discussion specifically contrasts traditional equity models with Initial Coin Offerings, detailing ICO mechanics such as token staking, network node ownership, and the unique liquidity risks that differ from historical venture bottlenecks. Furthermore, the session addresses current industry uncertainties, including the ambiguous legal interaction between SAFEs and tokens, the global nature of crypto capital flows, and Bromberg's caution regarding irrational pricing and untested norms in the decentralized fundraising landscape.

Startup Investor School Day 4 Live Stream

Andy Bromberg, Aaron Harris, Ron Conway, Jeff

The final day of Startup Investor School featured a comprehensive analysis of early-stage investment history by Andy Bromberg, behavioral standards for angel investors presented by Aaron Harris, and a founder-centric investment philosophy from Ron Conway. The session detailed the evolution from traditional venture capital to modern token offerings like ICOs while establishing a strict code of conduct emphasizing reputation, speed, and integrity in deal-making. Ultimately, the event equipped attendees with tools to model SAFE conversions and fostered a network committed to maintaining ecosystem transparency and supporting high-potential founders.

Michael Seibel - Startup Investor School Day 2

Michael Seibel

Outgoing Y Combinator CEO Michael Seibel shares investment lessons from his 4.5-year angel career, highlighting that 50 checks resulted in 12 post-Series A companies and a billion-dollar exit despite his dual role as a YC partner. He advises investors to write larger checks with speed and minimal friction while avoiding the hype trap of Demo Day in favor of evaluating cap tables and founder potential. The session concludes by defining top-tier investors by their willingness to act quickly and noting that luck plays a significant role in success.

Geoff Ralston's Intro - Startup Investor School Day 1

Geoff Ralston

Y Combinator launched its inaugural hybrid Startup Investor School, a four-day program in Mountain View and online that educates accredited and non-accredited investors on startup selection and deal mechanics. The curriculum, delivered by volunteer instructors from the startup ecosystem, concludes with virtual and in-person invitations to Winter 2018 Demo Days for select participants. Additionally, the initiative establishes a crowd-sourced knowledge repository and leverages historical case studies to refine investor decision-making capabilities.

The Autonomy Ecosystem: Public Infrastructure (2 of 8)

Frank Chen

Accelerating transitions to autonomous electric fleets are rendering traditional gasoline tax models obsolete, prompting urgent infrastructure investments and the exploration of new per-ride revenue strategies across the Asia-Pacific region. Simultaneously, urban planners are reimagining city layouts by repurposing vast areas previously dedicated to parking and car corridors into multi-use spaces, leveraging real-time data to optimize traffic flow and reclaiming land for pedestrian and green zones. This shift promises a three-dimensional urban evolution that eliminates vehicle idling, drastically increases lane throughput, and integrates vertical expansion through tunnels and aerial corridors.

Go to Market Boot Camp for Startups: Field Sales

Mark Cranney

This presentation outlines a comprehensive framework for designing sales organizations and orchestrating the buyer's journey through three strategic layers: top-down direct sales for complex enterprise deals or bottom-up user-driven growth for product-led models. It details how to align discovery and value articulation with specific buyer personas to navigate the psychological shifts from needs assessment to risk mitigation, while employing distinct competitive strategies to secure political champions within the client organization. Finally, the content establishes rigorous operational protocols for integrating economic validation, customized proposal planning, and cross-functional forecasting to ensure that deal stages are defined by objective deliverables rather than subjective rep sentiment.

How to Manage with Ben Horowitz (How to Start a Startup 2014: Lecture 15)

Ben Horowitz

This presentation argues that effective leadership requires analyzing high-stakes decisions through the perspectives of all stakeholders to avoid unpredictable cultural side effects. It illustrates this principle through case studies on equity management, compensation cycles, and stock option policies, while citing Toussaint Louverture's historical strategy of integrating former enemies to demonstrate how inclusive thinking drives organizational success. The discussion concludes by outlining practical execution techniques for leaders, including formal review processes and the importance of pausing to mitigate the risks of reactive decision-making.

MIT 6.S094: Convolutional Neural Networks for End-to-End Learning of the Driving Task

This lecture explores the application of Convolutional Neural Networks to computer vision challenges, specifically demonstrating how deep learning models surpass human performance on benchmarks like the CIFAR-10 dataset to enable autonomous driving systems. It details the architectural differences between convolutional, pooling, and fully connected layers while contrasting browser-based training tools like ConvNet.js with robust offline implementations in TensorFlow. The session concludes by addressing critical industry hurdles such as data scarcity for rare edge cases and the necessity for near-perfect accuracy to ensure safety in real-world deployment scenarios.

MIT 6.S094: Deep Reinforcement Learning for Motion Planning

Lex

Participants in the Deep Traffic competition develop deep reinforcement learning agents to navigate a seven-lane highway simulation, aiming to achieve an average speed of 65 mph or higher through autonomous decision-making. Utilizing client-side JavaScript and Andrej Karpathy's ConvNet.js library, competitors train neural networks via experience replay and Q-learning algorithms without relying on explicit ground truth data for vehicle actions. The initiative serves as a practical framework for exploring the complexities of autonomous driving, highlighting both the potential of simulation-based learning and the critical challenges of aligning reward functions with real-world safety constraints.

Nuts and Bolts of Applying Deep Learning (Andrew Ng)

Andrew Ng, Hugo Larochelle, Andrej Karpathy, Richard Socher, Sherry Moore, Ruslan Salakhutdinov, John Schulman, Pascal Lamblin, Adam Coates, Alex Wiltschko, Quoc Le, Yoshua Bengio, Shubho Sengupta, lexfridman, Peter, Andre, Shubo, Sammy

Baidu structures its 1,000-person AI organization around unified data warehouses and integrated ML-HPC teams to drive deep learning performance that scales linearly with data volume rather than traditional algorithms. The presentation outlines critical diagnostic frameworks for bias and variance, emphasizing human-level error as a benchmark for defining theoretical limits and guiding the shift toward end-to-end learning in data-rich perception tasks. Finally, the discussion establishes practical heuristics for product automation and career development, advocating for synthetic data engineering and the rigorous "dirty work" of replicating research papers to master the field.

TensorFlow Tutorial (Sherry Moore, Google Brain)

Sherry Moore, Hugo Larochelle, Andrej Karpathy, Richard Socher, Ruslan Salakhutdinov, Andrew Ng, John Schulman, Pascal Lamblin, Adam Coates, Alex Wiltschko, Quoc Le, Yoshua Bengio, Shubho Sengupta, lexfridman, Zach, Pichin Lo

Google Brain's Sherry Moore presented a tutorial on transitioning from research to production using the TensorFlow framework, highlighting its open-source architecture that supports diverse applications like image recognition, voice processing, and deep learning. The session detailed core concepts such as data flow graphs, placeholders, and session execution while guiding attendees through hands-on labs for linear regression and MNIST digit classification. Moore also outlined the platform's extensive portability across mobile and cloud devices and invited community contributions to further develop the library's modular design.

Theano Tutorial (Pascal Lamblin, MILA)

Pascal Lamblin, Hugo Larochelle, Andrej Karpathy, Richard Socher, Sherry Moore, Ruslan Salakhutdinov, Andrew Ng, John Schulman, Adam Coates, Alex Wiltschko, Quoc Le, Yoshua Bengio, Shubho Sengupta

This presentation details Theano, an eight-year-old symbolic expression compiler that enables high-performance deep learning by automatically differentiating mathematical graphs and compiling them into optimized C++ or CUDA code. The session demonstrates the framework's core capabilities, including graph manipulation for neural network backpropagation, GPU acceleration via shared variables, and sequence modeling through the `scan` operator, while showcasing practical implementations of logistic regression, convolutional networks, and LSTMs on datasets like MNIST. Addressing deployment challenges inherent in its tight Python integration, the discussion concludes by highlighting Docker containers as the standard solution for distributing models and outlines a roadmap for enhanced 3D convolution and cuDNN support.

Deep Learning for Speech Recognition (Adam Coates, Baidu)

Adam Coates, Hugo Larochelle, Andrej Karpathy, Richard Socher, Sherry Moore, Ruslan Salakhutdinov, Andrew Ng, John Schulman, Pascal Lamblin, Alex Wiltschko, Quoc Le, Yoshua Bengio, Shubho Sengupta

Deep learning has revolutionized speech recognition by replacing traditional, error-prone pipeline architectures with end-to-end neural networks that map raw audio directly to text, achieving character error rates below 6% in Mandarin. This shift utilizes techniques such as Connectionist Temporal Classification and advanced data augmentation to overcome historical limitations in accuracy and scalability, enabling systems to match human transcriber performance while significantly increasing user productivity. As researchers address computational bottlenecks through optimized training strategies like dynamic batching, these models are transitioning from experimental benchmarks to production-ready tools for consumer applications ranging from real-time captioning to hands-free vehicle control.