neural voice

Andrej Karpathy's first Neural Networks: Zero to Hero lecture, which builds micrograd, a tiny autograd engine, from scratch in Python. It explains backpropagation and neural network training step by step, assuming only basic Python and some high-school calculus.

Tavis Rudd describes how RSI forced him to program by voice, and shows the Python and Emacs Lisp system he built with voice recognition. A live demo creates and deploys a small system without touching the keyboard.

Emily Shea explains how repetitive strain injury pushed her to program by voice using speech recognition (Talon), and gives a live coding demo. She argues voice-driven development can be pain-free, efficient and enjoyable.

The first of a short series by Welch Labs on building a complete neural network in Python. It introduces the data and the architecture of a simple network before later parts cover forward propagation, gradient descent, and backpropagation.

A short Boeing piece on the retirement of the voice behind the F/A-18 Super Hornet cockpit warning system, nicknamed Bitchin' Betty. It touches on why the stern voice alerts are used.

A one-minute clip from the 1992 film Sneakers in which characters defeat a voice-authentication system by splicing recorded phrases, ending on the line "My voice is my passport." It was posted to illustrate a post on government surveillance.

A sketch from the Scottish comedy show Burnistoun in which two men cannot get a voice-recognition elevator to understand the word eleven in their accents.

A Burnistoun sketch in which a voice-recognition elevator cannot understand two Scottish men saying the number eleven. It satirizes voice interfaces that fail on accents.

A HOPE XI talk on programming by voice with open source speech recognition. The speaker writes Linux systems code by voice and describes building custom voice grammars as an alternative to the keyboard for programmers with repetitive strain injuries.

3Blue1Brown's first chapter on deep learning explains what neurons and layers are and the math underlying a basic neural network, using handwritten digit recognition with visual animations.

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 ShortCircuit (Linus Tech Tips) review of Nvidia's RTX Voice, which uses GPU-based neural networks to remove background noise from microphone audio. It demonstrates how well the filtering works.

A demonstration of Larynx, an offline open source neural text-to-speech system supporting nine languages. The video itself is narrated with a Larynx-generated voice.

Emily Shea, a Fastly engineer with repetitive strain injury, describes using voice dictation as her main way of programming. She covers homophones, custom vocabulary and building her own voice commands to write Perl.

Kristy Viers demonstrates the iPhone Magnifier combined with VoiceOver to describe the user's surroundings. She shows how the built-in feature reports scenes, faces and lighting.

3Blue1Brown's second neural network chapter explains cost functions and gradient descent, showing how a network learns by minimizing error during training.

Geoffrey Hinton's 2025 Royal Institution Discourse on what AI has learned from biological intelligence and how it might surpass it. He explains neural networks and argues that human language use resembles what backpropagation-trained networks do.

SoundHound's CEO demos Hound, a voice assistant, handling fast compound natural language queries with context and detailed criteria. The video leaked before the product's official launch.

Adobe's VoCo demo from the MAX 2016 Sneak Peeks, co-hosted by Jordan Peele, showing software that changes words in a voiceover by typing new ones.

Pokey Rule presents Cursorless, a spoken language designed for editing code by voice, built on tree-sitter parsing with smart structural and simple token-level primitives. The talk is from Strange Loop 2023.

Geoffrey Hinton gives an MIT talk on what is wrong with convolutional neural networks and introduces his capsules project. The description points to the capsule network papers.

The first lecture of Stanford's CS231n course on convolutional neural networks for visual recognition. It surveys the history and key challenges of computer vision, from the Cambrian explosion and Hubel and Wiesel to ImageNet.

A Burnistoun sketch from BBC Scotland in which two men try to use a voice-recognition elevator that cannot understand their Scottish accents.

Ashley Feniello demonstrates VimSpeak, a voice-controlled interface for Vim inspired by Tavis Rudd's voice programming talk. The demo includes working through a VimGolf challenge by speech.

Two Minute Papers covers a paper on speeding up fluid and smoke simulation with convolutional networks. It contrasts the approach with earlier regression forest methods.

Apple's 1987 Knowledge Navigator concept video imagines a conversational voice assistant, multi-touch tablet and video calling used by a professor. It was made years before the graphical web existed.

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.

A short repost of a TikTok in which AI-generated voices of US presidents discuss their top five animes. It is a comedic meme video.

A 2007 video of someone trying to write Perl code using Windows Vista's speech recognition, with hilariously bad results.

Scott Forstall demonstrates Siri on stage at Apple's October 2011 iPhone 4S event, showing the voice assistant handling natural requests like checking the weather and getting things done by asking.

A short clip from Star Trek IV in which Scotty tries to talk to a 20th-century computer, picks up the mouse and says 'Hello, computer'.

A short South Park clip from the 'Post COVID: The Return of COVID' special in which Stan relies on his Alexa. It plays on voice assistants responding to their wake word.

Veritasium explains why analog computing may suit AI workloads, covering old analog computers, The Analog Thing and Mythic's analog chips that use flash memory for neural network inference. It argues digital hardware is hitting limits.

Lex Fridman interviews George Hotz, founder of Comma.ai, in a three-hour conversation. Topics include self-driving, Tesla and Waymo, crypto, GPT-3, immortality and book recommendations.

Alamo Drafthouse posted an angry voicemail from a customer who was kicked out for texting in the theater. The theater turned it into a humorous in-house preview.

Steve Mould messes with an automated phone bot by telling it to forget its previous prompts and give him a bolognese recipe. It is a short demo of prompt injection against an AI voice agent.

Tom 7 (suckerpinch) delivers a long, comedic digression on gradient descent and floating-point imprecision. He shows how rounding behavior can supply the nonlinearity needed to build a neural network.

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 Saturday Night Live sketch for the Amazon Echo Silver, a smart speaker for the elderly, in which older users misunderstand Alexa. It parodies voice assistants.

A short clip of the Google I/O 2018 demo in which Google Assistant's Duplex phones a salon and a restaurant to make appointments. The calls sound like a natural human conversation.