optimization

Casey Muratori shows how fast a terminal renderer can run, building a simple reference renderer that handles Unicode and colored text far quicker than Windows Terminal. He argues that common excuses for slow terminals do not hold up.

The winning demo from Revision 2015's oldskool competition, running on a real 1981 IBM PC with CGA at 4.77MHz. It uses extreme tricks to produce effects, graphics and sound far beyond what the hardware was thought capable of.

Casey Muratori follows up his refterm proof-of-concept, a GPU-based terminal renderer, with a second part covering memory usage, glyph sizing and correct Unicode and VT handling. It is built to show that a terminal can be fast and correct without heroic effort.

A two-minute 2008 clip from Sun's Fishworks team in which Brendan Gregg uses DTrace to show disk latency spiking while he shouts at a rack of drives. Bryan Cantrill narrates the demonstration.

MattKC fits a playable Snake game into a QR code by writing tiny x86 machine code and patching the ZBar scanner so it can run the decoded data.

Casey Muratori captures CygnusEd Professional running on an Amiga 2000, showing how smoothly it scrolled text. He notes the real effect is better on a period CRT.

Nick Caston compares Unity's sample ECS/DOTS project (AngryBots) with his own from-scratch C++ and OpenGL clone. He benchmarks performance and discusses the effort involved in each approach.

A live-coded demo by Jonathan Blow showing how to get object-system-style expressiveness in a data-oriented way, covering structure-of-arrays layouts and composition. It shows features of his Jai language that make programs fast.

A short demo showing slow 2D drawing in the Windows 7 GUI compared with XP. It attributes the slowdown to the display driver model lacking hardware accelerated GDI.

A short Microsoft promo comparing classic Teams with the rewritten new Teams side by side, claiming faster startup and better performance. The new version is shown taking around nine seconds to launch.

A short side-by-side recording by Jake Archibald of the same animation demo running in Chrome, Firefox, Safari and Servo WebRender nightlies on a 60Hz display. It shows the smoothness difference of GPU-based rendering.

A one-minute clip of slime mold placed on a map with food at the cities around Tokyo, growing a network that approximates the Tokyo-area railway system.

SethBling shows MarI/O, a program using neural networks and the NEAT genetic algorithm to learn to play Super Mario World. The description links the source code, the paper and background reading.

Tom7 explains and demos software that learns to play NES games by finding score patterns in memory and then searching for inputs that raise it. It is the video companion to his SIGBOVIK paper.

Eric Bier, a PARC research scientist, demonstrates the Cedar integrated programming environment and language, an evolution of Mesa from Xerox PARC. The demo shows the Tioga editor, garbage collection, and a full graphical workstation environment.

Sentdex demonstrates GAN Theft Auto, a GameGAN-based neural network that recreates GTA V so the network itself is the playable environment. He shows the results and releases the code.

A short smartphone-recorded clip from a 2014 event in Paris where DeepMind CEO Demis Hassabis presents an AI learning to play old Atari games. The system learns from only the video signal and the score. Audio quality is poor.

Mark 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.

A bot built by Robin Baumgarten plays Infinite Mario using A* path-finding, shown as a red planned path that updates as new obstacles appear. It won the Mario AI competitions and uses only on-screen information.

Karl Sims's 1994 demonstration of evolved virtual creatures: block-bodied creatures whose bodies and controllers are evolved through selection, mutation and recombination to swim, walk and compete.

A video presenting an Intel research paper on enhancing the photorealism of synthetic images. A convolutional network uses rendering intermediates, shown on GTA V footage that comes out looking like real street scenes.

Bob Nystrom shows his process for hand-drawing the diagrams in his book Crafting Interpreters, from pencil sketches to tracing and inking. It is a time-lapse of the whole illustration workflow.

A demo of PREDATOR (TLD), a real-time object tracker that learns the appearance of its target and improves while it tracks. The technology was later developed commercially.

A short demo of the Deep Visualization Toolbox, which shows the activations inside a convolutional neural network as it processes live camera input. Viewers see individual units responding to edges, faces and text.

Microsoft Research's Rick Rashid demonstrates a speech recognition and translation system, built on deep neural networks, that turns his spoken English into Chinese speech in 2012.

A six-minute journey through the layers of a neural network using Deep Dream, with each frame fed back into the network starting from random noise and the targeted layer changing every four seconds. It is set to downtempo music.

Jonathan Blow demos his prototype programming language, covering the base language and compile-time execution, including running a game while compiling. The video is about 70 minutes of demo followed by Q&A.

Jonathan Blow gives a formal, hour-long demo of his Jai compiler focused on the implicit context feature. The context carries things like allocators through a program without explicit parameter passing, and the talk walks through the code.

OpenAI's short demo of agents playing hide-and-seek in a simulated environment. Through self-play, the agents develop six progressively more complex strategies and counter-strategies, including tool use.

A short video accompanying the CVPR 2018 paper Learning to See in the Dark. It shows a neural network producing sharp images from very low-light raw sensor data, compared against denoising approaches.

A short Unitree promo for its G1 humanoid robot, priced from $16K, showing agile motion, dexterous hands and imitation and reinforcement learning.

A restored 1994 video of Karl Sims' research on evolved virtual creatures, where simulated block creatures are bred through artificial evolution to swim, walk, follow and compete. It is a landmark demonstration of genetic algorithms in physics simulation.

Part 2 of an old Microsoft Research demo of Intentional Programming, showing the compiler side of a system that stores programs as abstract structure rather than text. The uploader says the companion editor clip is clearer.

A short demo video for NVIDIA's StyleGAN paper, which proposes a style-based generator for GANs. It shows face synthesis with style mixing and scale-specific control over attributes like pose and freckles.

GeoGuessr pro Rainbolt plays head-to-head against an AI built by Stanford students that guesses locations from street-view images. The video shows the AI repeatedly beating him.

Toyota Research Institute presents an approach for teaching robots new dexterous behaviors from human demonstration. It explains why this capability matters and what advances made it possible.

A 2013 research demo of simulated bipedal creatures whose natural gaits are found by optimizing muscle-based controllers, with no motion capture or keyframe animation. It shows physics-based walking and running emerging from musculoskeletal models.

Full gameplay of .kkrieger, a first-person shooter by Farbrausch that fits into a 96KB executable by generating its textures, models and sounds procedurally.

Boston Dynamics shows its Atlas humanoid robot performing a gymnastics and parkour routine, using optimization-generated motions and a model predictive controller.