exact instructions

A family comedy video in which kids write instructions for making a peanut butter and jelly sandwich, and a parent follows them with exact literalness, producing chaos. This is the classroom-friendly version.

A family video in which kids write instructions for making a peanut butter and jelly sandwich and a parent follows them exactly as written, with comic results. It demonstrates how precise instructions must be.

Christopher Domas presents at Black Hat 2017 how page fault analysis and processor fuzzing can exhaustively search the x86 instruction set. The talk uncovers undocumented instructions and hardware bugs and shows limits of disassemblers.

A 2018 vlog showing how to use the windshield wipers on a Tesla Model 3, including auto mode, manual settings and voice commands. It is labeled outdated.

Christopher Domas presents research at DEF CON 26 showing a hidden alternate instruction set and undocumented hardware backdoor in some x86 processors, found on a third-party CPU. He uses it to ask how feasible wider hardware backdoors are.

A short comedic complaint by ProZD about YouTube search failing to surface his own video even when searching its exact title. It illustrates how poor search results have become.

Christopher Domas's Derbycon 2015 talk on the M/o/Vfuscator, a compiler that emits only x86 mov instructions, showing mov is Turing complete. It explains how a full machine is built from mov and why the output is a nightmare to reverse engineer.

CGP Grey explains that postal codes are routing instructions for mail rather than geographic regions. He looks at the quirks and problems this causes.

The first lecture of Stanford's CS107 Programming Paradigms, taught by Jerry Cain, giving an overview of the course. The course goes from C and assembly through C++, concurrency, Scheme and Python.

Matt Godbolt, creator of Compiler Explorer, tours surprising optimizations modern compilers perform, such as turning loops into closed-form sums and using popcount instructions. He explores what the compiler does behind the scenes.

8088 MPH is a demoscene production for the original 1981 IBM PC with CGA graphics and a 4.77MHz 8088, shown with real composite output. It competed at Revision 2015 and pushes the hardware well past what it was thought capable of, breaking emulators.

MIT 22.01 lecture by Michael Short explaining the Chernobyl accident technically, including RBMK reactor positive feedback, xenon poisoning and the cascade of human errors during the test.

Matt Ranney of Uber talks at GOTO Chicago 2016 about what Uber learned from growing past 1,000 production microservices. He covers regrets, sprawl and organizational effects.

Bill Dally of NVIDIA gives a University of Washington lecture on trends in deep learning hardware. It covers numeric formats, specialized instructions, energy per operation and co-design of hardware and software.

Vox's Joe Posner, with help from 99% Invisible, explains why bad doors are everywhere. It covers Don Norman's long-running complaint about doors and the idea that good design should not need instructions.

Stand-up Maths explains why the perimeter of an ellipse has no simple closed-form equation, showing the infinite series and presenting approximation formulas.

12tone tells the story of pitch inflation and how concert A came to be standardized at 440 Hz, including attempts at legal standards. It is a short music history explainer.

John McCutchan of Google explains how Dart and the Dart VM expose SIMD instructions to speed up web software. He covers what SIMD is and how it applies to graphics and games.

A three-minute student film from Campus MovieFest showing drivers in every lane holding exactly the speed limit on an Atlanta-area interstate and the traffic jam that results.

MSNBC's clip of the April 2020 White House coronavirus briefing in which President Trump responds to a DHS presentation on disinfectants by musing about injecting disinfectant into the body. The segment is three minutes long.

Rob Miles discusses GPT-3 on Computerphile, covering how scaling parameters kept improving results over GPT-2 and how the model nearly learned arithmetic without anyone knowing exactly how.

A short clip from HBO's Silicon Valley in which Gavin Belson fires the Nucleus team.