Technology
10,995 cards. Each one is a 40-second story with a five-minute read behind it. Page 105 of 184.
Subtopics: 3D printing & manufacturing · AI & machine learning: practical · Aircraft · Automata · Aviation · Batteries · Binary · Biotechnology · Bridges · Cameras & imaging · Cars · Chemical engineering · Chips · Codes · Compression · Computing · Computing and information technology · Computing history · Cryptography · Cybersecurity · Electric vehicles · Electronics · Energy · Engineering · Everyday tech · Famous patents · Future Frontier · How it works · How Things Work: GPS to the Phone · How Things Work: Jets, Propellers and Ships · Industry · Infrastructure · Internet · Inventions · Inventors · Machinery and tools · Maths · Media and communication · Medical technology · Medieval clockmaking · Mercedes-Benz Engineering · Military technology · Motorbikes · Navigation and timekeeping · Networks · Nuclear power · Optical technology · Physics · Pioneers · Programming · Programming languages · Radar · Radio · Researched · Robotics: practical · Robots · Rockets · Security · Semiconductors · Ships & boats · Skyscrapers · Smartphones · Software · Sound & audio technology · Spaceflight · Super-series Wave 1: Car To Wheels · Super-series Wave 1: From Power Station To Pc · Super-series Wave 1: The Journey Of Crude Oil · Technology today · Technology: General · Textiles · Trains & locomotives · Transportation · UK & Ireland: inventions · Video game technology · Web · Wireless
Is AutoGPT the next evolution of ChatGPT?Explore the potential of AutoGPT, a tool designed to supercharge the capabilities of ChatGPT through autonomous execution.
How digital manufacturing is transforming the construction industryExplore the shift from traditional building methods to a digital manufacturing revolution, featuring footage from Laing O'Rourke's Explore Industrial Park.
Inside Vienna's ambitious $6 billion urban transformation projectExplore the massive scale of Vienna's new development, a $6 billion initiative designed to function as a city within a city.
How AI learned to simulate reality by watching 600 million videosStable Video AI represents a leap in generative technology, capable of producing simulations that closely mimic real-world physics. This video explains the scale of data required …
Why cooling compressed air is vital for modern turbocharged enginesAs engines trend toward downsized turbocharger solutions, managing heat becomes critical. This video examines the fundamental differences between air-to-air and air-to-water inter…
How a cinematic classic inspired a modern short filmExplore Rosebud, a thirteen-part short film that pays homage to the legendary Citizen Kane through a creative lens.
The guest people who shaped the history of South ChinaOften called the 'guest people,' the Hakka are a massive ethnic subgroup of the Han. Their history is defined by centuries of migration, fleeing warfare and seeking new lands, eve…
The invisible code that tells the internet how to speak to youEvery time you load a webpage in your native tongue, a precise, standardized string of characters is working behind the scenes. From HTML to PNG files, the IETF language tag ensur…
Could your consciousness survive a transfer to a digital landscape?The concept of mind uploading promises a form of digital immortality, allowing us to live forever without a body. But as we attempt to map the universe's most complex object, we f…
Can a computer write its own code from pure logic?Imagine describing exactly what you want a program to do without explaining how to do it. Program synthesis is the quest to bridge that gap, moving beyond manual coding toward sys…
Could an AI rewrite its own code to trigger an intelligence explosion?The concept of recursive self-improvement suggests a future where artificial intelligence acts as its own engineer. By iteratively refining its own algorithms and hardware, a suff…
Why giving AI a face changes how we trust and learnMost famous virtual assistants lack a visual presence, but adding a body—physical or digital—transforms the interaction. Beyond simple utility, embodied agents use gestures and ex…
The predictable math behind artificial intelligence's growing powerWhy does adding more data and compute consistently make AI smarter? It isn't just a lucky trend; it follows a specific mathematical blueprint. Neural scaling laws reveal that mode…
The clever design behind a classic children's toyExplore the engineering and flexibility of the GeoSafari Jr., a toy that combined simple play with clever design. This video examines how the game functioned and why its adaptable…
How 100,000 Conversations Trained a New Generation of AI?OpenAI has released a new version of ChatGPT trained on an unprecedented dataset: 100,000 human conversations. This massive corpus suggests a significant leap in conversational fl…
Can AI master the complex logic of Euclidean geometry?Explore how DeepMind's AlphaGeometry uses a massive dataset of 100,000,000 examples to approach human-level reasoning in mathematical proofs.
How ancient paper folding techniques are revolutionizing modern engineeringDiscover how the centuries-old art of origami provides a blueprint for complex mechanical design, from deploying massive solar arrays in space to creating microscopic compliant me…
Is a new era of free, hyper-realistic AI simulations arriving?Explore the latest advancements in AI-driven simulations designed to mimic reality with unprecedented precision. This video examines research into models that bridge the gap betwe…
How do the world's most famous M1 Garand clones actually compare?General George S. Patton once called the M1 Garand the finest battle implement ever created. This video examines how that legendary design influenced and was replicated by other n…
Classify the sentiment of text with a Transformers pipelineUse the pipeline() function to label a sentence as positive or not, with a pretrained sentiment model.
Why GPS needs four satellitesSignals from at least four satellites give a receiver its distance to each, so it can calculate its position and the time.
How lidar maps the world in 3DA lidar times laser pulses to a surface and back, and many ranges together give three-dimensional information.
How do computers read text using a tokenizer?Transformer models cannot understand raw text directly. You will learn how a tokenizer acts as a bridge, converting human language into numerical data that a neural network can pr…
How AI Agents Use Tools and PromptsUnderstand the core architecture of an AI agent by exploring how they combine large language models with external tools and structured instructions to perform complex tasks.
How a model reaches convergenceYou can track a machine learning model's progress by watching its loss curve. Understanding this process helps you know exactly when your training has finished successfully.
Automating Model Sharing with the push_to_hub APILearn how to seamlessly sync your machine learning models to a central repository. Understanding this mechanism allows you to automate version control and metadata generation duri…
Can Large Language Models Teach Robots to Perform Complex Tasks More Efficiently?This video examines the Eureka research paper, which explores how large language models can be utilized to design rewards for robotic systems. It highlights a significant advancem…
How a Large Language Model predicts the next piece of textYou use them every day to write emails or code, but how do these massive AI models actually function? Learn the fundamental mechanism behind the text generation that powers modern…
How to read machine learning learning curvesBy analyzing the shape of loss and accuracy curves, you can determine if your model is learning effectively, overfitting to your data, or failing to capture essential patterns.
How to create a source release in ROS 2Learn the standardized workflow for releasing new features and bugfixes in ROS 2. Understanding this process ensures that repository versions remain consistent across the ecosyste…
How attention mechanisms handle long sequencesStandard AI models struggle with long texts because their computational needs grow quadratically. You will learn how specialized attention mechanisms use sparsity to process much …
Why the world's most accurate force measurement requires a million-pound stackTo ensure jets, rockets, and planes function safely, engineers must calibrate force transducers against a massive standard. This video explores the NIST facility in Washington DC,…
How neural rendering techniques can generate complex 3D shapes from scratchThis video explores the research behind Differentiable Signed Distance Function Rendering. It demonstrates how modern computational methods allow for the synthesis of intricate 3D…
Can an artificial intelligence learn to master Minecraft through massive internet data?This video examines how researchers developed an embodied agent capable of performing complex tasks in Minecraft. By leveraging vast amounts of online information, the system demo…
Can robots learn to manipulate objects just by observing human video footage?This video examines a new method that enables robotic hands to acquire dexterity by analyzing human movements from online video content. It highlights a significant shift in how m…
Can Google's new autoregressive model turn simple text prompts into high-fidelity digital art?This video examines the capabilities of Google's Parti, a text-to-image generation model. By leveraging an autoregressive approach, the system demonstrates an impressive ability t…
Achieving real-time ray tracing through advanced spatiotemporal resampling techniquesThis video examines how researchers have rearchitected spatiotemporal resampling to finally enable real-time ray tracing. It highlights the massive efficiency gains required to re…
What happens when the line between digital simulation and physical reality begins to blur?This video covers the NVIDIA GTC presentation by Károly Zsolnai-Fehér, exploring the intersection of simulation and reality. It highlights the collaborative effort behind the proj…
How a new AI model mastered the complex world of Minecraft through video observation.This video examines the research behind Video PreTraining, a method allowing an AI to learn how to play Minecraft by watching human gameplay footage. It highlights a significant s…
Can a single photograph now generate a high-resolution, animated digital human head?This video explores MegaPortraits, a new neural network capable of creating high-quality, megapixel-resolution head avatars from just one source image. It demonstrates how modern …
How Google's new artificial intelligence lets you travel through a single photographThis video explores a new AI technique that transforms a static image into a continuous, navigable landscape. By analyzing a lone picture, the system generates new perspectives, a…
How artificial intelligence reconstructs clear scenes from extremely dim and grainy photographsThis video examines a machine learning technique that synthesizes high-quality, bright images from low-light, noisy raw data. It explains how researchers enable computers to perce…
Simulating One Billion Particles: A New Milestone in Fluid DynamicsThis video examines a breakthrough method for modeling fluid motion at an unprecedented scale. By utilizing a highly efficient mathematical approach, researchers can now simulate …
Can artificial intelligence reconstruct realistic human faces from simple two-dimensional images?This video explores a new artificial intelligence technique capable of creating three-dimensional human faces from flat imagery. By utilizing dense landmarks, this method advances…
How Stable Diffusion democratized high-resolution image generation for everyoneThis video explores the research behind Stable Diffusion, a powerful model for generating high-resolution images. It highlights how this technology became accessible to the public…
Can Google's Minerva AI Solve Complex Quantitative Reasoning Problems Like a Human?This video examines Minerva, a language model developed by Google designed to tackle quantitative reasoning. It explores how the model approaches complex mathematical and scientif…
How Google's DreamBooth AI Learns to Place Your Specific Subjects Into New ScenesThis video examines DreamBooth, a method that allows AI models to learn unique subjects from just a few images. By teaching the system to associate a specific entity with a unique…
Can artificial intelligence turn a single static photograph into an endless cinematic journey?This video explores a new machine learning technique that allows users to travel through a two-dimensional image as if it were a three-dimensional environment. It demonstrates how…
How robots can now navigate using simple natural language commandsThis video examines a new robotic navigation system that integrates language, vision, and action models. By leveraging large pre-trained artificial intelligence, these machines ca…
Can neural networks make complex 3D simulations faster and more efficient?This video examines a method for representing high-detail volumetric data using hierarchical neural networks. By optimizing how 3D spaces are stored and processed, this technology…
How Google's new AI technique allows precise control over image generation promptsThis video explores Prompt-to-Prompt, a method for editing images generated by AI. By manipulating cross-attention layers, users can modify specific elements of an image while mai…
Can artificial intelligence generate complex 3D models from simple text descriptions?This video explores DreamFusion, a new AI technique that creates 3D assets from text prompts. By leveraging 2D diffusion models, this technology bridges the gap between static ima…
How NVIDIA achieved real-time ray tracing, a feat once considered computationally impossible.This video explores the technical breakthroughs that enabled real-time ray tracing in computer graphics. It examines the methods used to simulate light transport, transforming a p…
Bridging the gap between digital simulations and the physical worldThis video explores the research paper titled The flow from simulation to reality. It examines how modern computational simulations are achieving unprecedented levels of visual fi…
Can AI help video game characters move more realistically?Explore how a new research paper introduces a method for creating game animations that look nearly indistinguishable from reality by mastering the underlying patterns of motion.
How OpenAI trained a speech recognition model on 680,000 hours of audio data.This video examines the technical approach behind OpenAI's Whisper, a system designed for high-performance speech recognition. By leveraging a massive, diverse dataset, the model …
How Latent Diffusion Models Are Revolutionizing High-Resolution Image SynthesisThis video explores the technical advancements behind Stable Diffusion. It examines how latent diffusion models enable the creation of high-quality images, offering a glimpse into…
Can a robot master table tennis against a human opponent?Google researchers have developed an artificial intelligence capable of playing table tennis. This video examines how a robotic system learns to compete in fast-paced matches agai…
How modern AI learns to complete text by filling in the gapsThis video examines a technical approach that improves how language models learn to generate text by training them to predict missing segments within a sequence. It explains how t…
How do researchers simulate the physics of one million crushing particles?This video explores a new method for simulating complex physical interactions, such as a hydraulic press crushing a million particles. It highlights how researchers achieve stable…
Intellect Plus