Page 6 of 24
In briefGeorge Hotz of Comma AI gives a 2017 conference talk on lessons from building a self-driving system. He argues about how driving should be defined and why the common approach to autonomous vehicles is flawed.
In briefMario Zechner explains why he built pi, a minimal coding agent, out of frustration with existing agent harnesses. He also discusses how AI agents are straining open source projects and argues for slowing down.
In briefYannic Kilcher explains how he trained a language model on three years of 4chan's /pol/ board and let it post on the site. He evaluates how good the model was and discusses the ethics and fallout.
In briefCoffee Break analyzes about 40,000 videos from seven months of YouTube's trending tab. It examines whether trending is biased toward mainstream media and late-night shows.
In briefAndrej Karpathy, then Tesla's head of AI, gives a CVPR 2021 workshop keynote on how Autopilot's neural networks work. A large part covers the move to a vision-only stack, including why radar was dropped and how vision produces depth estimates.
In briefPeter Norvig's 2007 Google talk on theorizing from data, covering what huge datasets can do for language understanding, translation, information extraction and inference. It argues that more data often beats cleverer models.
In brief3Blue1Brown's second neural network chapter explains cost functions and gradient descent, showing how a network learns by minimizing error during training.
In briefMark Rober and Jabril build an app that uses machine learning to decode baseball sign sequences. The video also explains neural networks in a simple, accessible way.
In briefMark Raasveldt of CWI presents DuckDB, an embeddable analytical database described as the SQLite for analytics, in a CMU Database Group talk. He explains why it was built, what it is for, and how it works.
In briefLex Fridman's long conversation with Andrej Karpathy on Tesla's self-driving approach, Optimus, neural networks, transformers, AGI and more. It also touches on his OpenAI work and working with Elon Musk.
In briefNovelist Daniel Suarez warns in a TED talk about autonomous robotic weapons and drones. He argues that taking the decision to kill out of human hands threatens democratic accountability and calls for a ban.
In briefPeter Norvig's UBC distinguished lecture on how massive amounts of simple data, such as for language modeling, can yield understanding that hand-built models could not.
In briefChristopher Domas's REcon 2013 talk on dynamic binary visualization for reverse engineering. He shows how rendering binary data as interactive visual maps and trigraphs reveals structure that hex dumps hide, leading to his Cantor.dust tool.
In briefCorridor Digital makes a roughly seven-minute anime short using Stable Diffusion and related AI tools, based on a technique they teach in a longer tutorial. It was an early showcase of AI-assisted animation.
In briefAlexander Nix of Cambridge Analytica presents at the 2016 Concordia Summit on using big data, psychographic profiling and behavioral microtargeting in elections.
In briefLex Fridman interviews chip architect Jim Keller about the future of computing, AI, life and consciousness. Topics include CPU design, branch prediction and why imperfect designs often win.
In briefTom Scott describes trying ChatGPT to fix his Gmail labels and being unsettled by how capable it was. He argues that we cannot know where we are on the technology curve, comparing it to Napster.
In briefNeil Halloran's animated, data-driven documentary visualizes the human cost of World War II and compares it with other wars and the trend toward fewer war deaths in recent decades.
In briefMark Rober shows an automatic bullseye dartboard that tracks a thrown dart with motion capture and moves in real time so it always hits the center. He spent over three years on the project.
In briefSimon Knowles, co-founder and CTO of Graphcore, talks about "intelligence processors", chips designed specifically for machine learning. The talk was given at the RAAIS 2017 summit and covers Graphcore's architecture.
In briefVeritasium explains why many real-world systems follow power laws rather than normal distributions. It covers self-organized criticality, with simulations of sandpiles, magnetism and forest fires.
In briefA university lecture from CS480/680 by Pascal Poupart on attention mechanisms and transformer networks, worked through on a chalkboard.
In briefAndrej Karpathy builds a bigram character-level language model from scratch, introducing torch.Tensor and the framework of training, sampling and loss evaluation. It is the first step toward a GPT-style Transformer in his makemore series.
In briefGeoff Cumming uses a simulation to show how wildly the p value varies from one replication to the next because of sampling variability. He argues for estimation and confidence intervals over null hypothesis significance testing.
