Videos de Ciencia
Videos sobre avances científicos y descubrimientos.
Ciencia 255 videos
Stanford que Ensina Mais sobre LLMs do que a Maioria dos Profissionais de IA
#llm #stanford #chatgpt #inteligenciaartificial #ia Em vez de assistir a uma hora de Netflix, assista a esta palestra de 2 horas da Stanford que vai te ensinar mais sobre como LLMs como ChatGPT e Claude são construídos do que a maioria das pessoas trabalhando em empresas de IA de ponta aprende em suas carreiras inteiras.Essa é uma das aulas mais valiosas que você vai encontrar sobre o funcionamento real dos grandes modelos de linguagem. Direto, profundo e sem enrolação.Se você quer entender de verdade como essas tecnologias são feitas por dentro, essa palestra é ouro puro. Pare o que está fazendo e invista essas 2 horas. Vale muito mais do que parece.Comente: você prefere usar IA ou entender como ela realmente funciona?
IA que Falla 🤖❌ | Errores, Riesgos y Fallos de la Inteligencia Artificial
❗ La inteligencia artificial no es perfecta… y cuando falla, las consecuencias pueden ser graves. 🤖⚠️ En esta MASTERCLASS descubrirás por qué falla la IA, cuáles son los errores más comunes y qué riesgos pueden afectar a las organizaciones y a las personas. 💣 📖 Aprende: ▫️ Principales fallos de la inteligencia artificial ▫️ Sesgos en los algoritmos ▫️ Problemas de calidad de datos ▫️ Decisiones incorrectas automatizadas ▫️ Riesgos éticos y de seguridad 🎯 Ideal para quienes quieren entender los límites de la inteligencia artificial, el gobierno de datos y la analítica. 🚀 Dale play y descubre cómo prevenir errores y usar la IA de forma más segura y responsable. #iaquefalla #inteligenciaartificial #erroresia #aiethics #datagovernance #bigdata #analiticadedatos #sesgos #machinelearning #tecnologia #datos #ciberseguridad #transformaciondigital #riesgos #innovacion
Autoencoders - Explained
An autoencoder is a neural network that learns to compress its own input and rebuild it. We start with the impossible-sounding task — squeeze a 49-pixel image down to two numbers and expand it back — then build the hourglass architecture, split it into encoder and decoder, and derive the reconstruction loss that trains the whole thing without a single label. By the end you'll see why the bottleneck is the real trick: forcing the network to carry meaning through a narrow waist is what organises the latent space and turns one idea into denoising, anomaly detection, and generation. *Related Videos* ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ Variational Autoencoder - Explained: https://youtu.be/geH5HnRapRs Generative Adversarial Networks (GANs) - Explained: https://youtu.be/G-fXV-o9QV8 Convolutional Neural Networks (CNNs) - Explained: https://youtu.be/YGILT182T6w Recurrent Neural Networks (RNNs) - Explained: https://youtu.be/8G1fImBCMcQ Backpropagation is Just the Chain Rule: https://youtu.be/VCGlYxGJZ04 Activation Functions in Neural Networks - Explained: https://youtu.be/slp222E_0d4 UMAP - Explained: https://youtu.be/kwILqPNZyeo Normalization vs Standardization - Explained: https://youtu.be/87C5hkTY8RI *Contents* ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ 00:00 - The Impossible Shortcut 00:22 - The Bottleneck 00:53 - Encoder and Decoder 01:21 - Reconstruction Loss 01:48 - The Latent Space 02:08 - Why This Matters *Follow Me* ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ 🐦 X: @datamlistic https://x.com/datamlistic 📸 Instagram: @datamlistic https://www.instagram.com/datamlistic 📱 TikTok: @datamlistic https://www.tiktok.com/@datamlistic 👔 Linkedin: https://www.linkedin.com/company/datamlistic *Channel Support* ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ The best way to support the channel is to share the content. ;) If you'd like to also support the channel financially, donating the price of a coffee is always warmly welcomed! (completely optional and voluntary) ► Patreon: https://www.patreon.com/datamlistic ► Bitcoin (BTC): 3C6Pkzyb5CjAUYrJxmpCaaNPVRgRVxxyTq ► Ethereum (ETH): 0x9Ac4eB94386C3e02b96599C05B7a8C71773c9281 ► Cardano (ADA): addr1v95rfxlslfzkvd8sr3exkh7st4qmgj4ywf5zcaxgqgdyunsj5juw5 ► Tether (USDT): 0xeC261d9b2EE4B6997a6a424067af165BAA4afE1a #autoencoders #deeplearning #machinelearning
3. Deep Learning Explained | Applications, Challenges & Ethics
Dive deep into the world of Deep Learning Models in this comprehensive video covering Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), Diffusion Models, and Transformers. Whether you're a student, AI enthusiast, or professional, this video breaks down complex concepts into easy-to-understand explanations. We begin with an introduction to deep learning, the driving force behind modern AI systems like ChatGPT and Google Gemini. Then, we explore how VAEs learn compressed probabilistic representations, how GANs generate realistic data through adversarial training, and how Diffusion Models create high-quality images by reversing noise. Next, we uncover the power of Transformers, the revolutionary architecture behind today’s most advanced natural language processing and multimodal systems. From text generation and translation to image synthesis, these models are transforming industries worldwide. 🚀 What you'll learn in this video: What deep learning models are and why they matter How VAEs, GANs, Diffusion Models, and Transformers work Real-world applications in healthcare, finance, entertainment, and robotics Key challenges like computational cost, bias, and interpretability Ethical concerns including misinformation, privacy, and responsible AI 🌍 Real-World Impact: Discover how deep learning powers innovations such as disease detection, fraud prevention, recommendation systems, autonomous vehicles, and more. ⚠️ Challenges & Ethics: We also discuss the limitations and ethical implications of AI, including bias in datasets, deepfakes, and the importance of transparent AI systems. 📌 Disclaimer: This video is created for educational and knowledge purposes only. The content is AI-generated, and while efforts have been made to ensure accuracy, some information may be incorrect or outdated. Viewers are encouraged to verify facts independently. #DeepLearning #ArtificialIntelligence #MachineLearning #Transformers #GANs #VAEs #DiffusionModels #AI #DataScience #NeuralNetworks #TechEducation #FutureOfAI #AIEthics #GenerativeAI deep learning models, VAE explained, GAN tutorial, diffusion models AI, transformers neural networks, generative AI explained, artificial intelligence basics, machine learning models, GPT and BERT explained, deep learning applications, AI in healthcare, AI ethics issues, neural networks tutorial, AI technologies 2026, generative models comparison
Robotique & IA
🤖🇯🇵 **IREX 2025 à Tokyo dévoile le futur des robots !** Des humanoïdes sauveteurs aux robots compagnons, le Japon et la Chine présentent une nouvelle génération de machines capables d’aider les humains dans les tâches dangereuses, médicales et quotidiennes. 🚀 🔥 Au programme : • **Caleido 9**, le robot de sauvetage géant de Kawasaki • **Friends**, l’assistant domestique pour personnes âgées • **Nurabot**, déjà utilisé dans les hôpitaux • **Corleão**, le cyber-cheval futuriste à hydrogène 🐎⚡ • Des humanoïdes chinois ultra avancés qui accélèrent la course à l’IA robotique 🤯 La frontière entre science-fiction et réalité disparaît peu à peu… 👀 #Robot #IA #Humanoide #Tokyo #IREX2025 #Tech #Innovation #FutureTech #ArtificialIntelligence #Robotique #Kawasaki #AI #Futur #Technologie
iHairium in 60 seconds: AI hair & scalp diagnostics, explained by the founder 💇♀️
Meet Ilia Krav, founder of iHairium, the AI platform for hair & scalp diagnostics in beauty. In 60 seconds, Ilia walks through what iHairium does for a haircare brand. Customers point their phone at their hair and scalp, get a diagnostic from a proprietary computer-vision model (trained on 150,000+ expert-labeled images), and walk away with a personalized routine pulled from your catalog. Trial-and-error haircare shopping becomes a guided purchase, products that work and customers who keep coming back. The platform today: ✔️ 475,000+ diagnostics completed ✔️ 35+ integrations across brands, salons, retailers, and telehealth ✔️ 4+ years in production ✔️ SOC 2 Type II, GDPR and CCPA compliant Works on Shopify, custom ecommerce, native apps, in-store kiosks, salon chairs - any touchpoint where you sell or recommend haircare. Learn more about our solutions & book a call ➡️ ihairium.com/for-health-beauty-brands #beauty #haircare #ai #ecommerce #saas #beautytech #aibeauty #DTC #beautybrand #beautyindustry #beautybusiness #aitools #founderinterview #beautypersonalization #beautyecommerce #aistartup #aihaircare #personalization #foundertalk #leadgeneration #CRO #beautyentrepreneur #shopify #aiforbeauty #b2bsaas
Tesla Optimus vs Boston Dynamics Atlas — Who Wins in 2025? #robot #technology #robotics #future
Tesla has more money. More factories. More engineers. And yet Boston Dynamics still builds the better robot. So what's actually going on — and why might that gap not last? In this video we break down the real difference between Tesla Optimus and Boston Dynamics Atlas. Not the viral clips. Not the hype. Just what actually matters — physical capability, artificial intelligence, and the one factor most people completely overlook. 🤖 What We Cover: ⚡ Why Boston Dynamics still leads on pure physical capability 🧠 How Optimus learns differently from every other robot on Earth 🏭 Why Tesla's manufacturing machine changes everything 💰 The $30,000 question that decides who actually wins 🌍 Why the better robot might not be the one that changes the world Today Atlas wears the crown. But the real race isn't about who builds the most impressive machine. It's about who builds the machine the world actually uses. 💬 Who do you think wins — Tesla Optimus or Boston Dynamics Atlas? Drop your verdict in the comments 👇 🔔 Subscribe to Robo Report — The AI Robot Revolution, Decoded. New videos every week tracking the biggest developments in AI robotics without the hype.
LLM Fundamentals Explained | How ChatGPT & Large Language Models Work”
🚀 Want to understand how ChatGPT and AI actually work? In this video, we break down Large Language Models (LLMs) in a simple and beginner-friendly way! Learn the fundamentals of LLMs, Tokens, Embeddings, Hidden States, AI Reasoning, and Generative AI with easy explanations and real-world examples. This video is perfect for students, beginners, M.Tech learners, AI enthusiasts, engineering students, interview preparation, and anyone curious about Artificial Intelligence. 📌 In this video, you will learn: ✅ What is a Large Language Model (LLM)? ✅ How LLMs work step-by-step ✅ Evolution of AI language understanding ✅ Tokens, Embeddings & Hidden States explained ✅ Difference between traditional NLP and Generative AI ✅ How ChatGPT understands language ✅ Real-world applications of LLMs ✅ AI concepts explained in simple language 🎯 Who should watch this? ✔ Engineering Students ✔ M.Tech / B.Tech Students ✔ AI & Machine Learning Beginners ✔ Cybersecurity & IT Students ✔ Interview Preparation Candidates ✔ Tech Enthusiasts 🔥 If you found this video useful, Like, Share & Subscribe for more AI, Cybersecurity, Python, Software Testing, and Technology videos! #LLM #ArtificialIntelligence #ChatGPT #MachineLearning #AIForBeginners #DeepLearning #NLP #LLMExplained #AITutorial #GenerativeAI
Diagnóstico de SOP: IA vs medicina tradicional (los datos reales)
7 de cada 10 mujeres con síndrome de ovario poliquístico reciben un diagnóstico erróneo o tardío, pero una nueva red neuronal promete cambiar este panorama clínico radicalmente. La medicina está atravesando un punto de inflexión gracias a la inteligencia artificial aplicada al diagnóstico hormonal. Investigadores han desarrollado un algoritmo capaz de detectar patrones en el SOP que los métodos convencionales pasan por alto. Analizamos cómo el procesamiento de datos a gran escala mejora la precisión clínica, reduciendo años de incertidumbre para las pacientes. Entender esta tecnología no solo es vital para la salud femenina, sino que marca el estándar de la medicina personalizada moderna. 🤖 EN ESTE VÍDEO: ✅ Cómo los algoritmos detectan biomarcadores que la analítica estándar ignora ✅ El papel del machine learning en la reducción de falsos negativos ✅ Comparativa entre el diagnóstico humano y el análisis predictivo por IA ✅ Implicaciones éticas y de privacidad en el manejo de datos médicos sensibles Estamos ante un cambio de paradigma que transformará la atención médica en tiempo récord. Suscríbete para no perderte los avances más disruptivos de la IA y dinos, ¿crees que los algoritmos deberían tener la última palabra en diagnósticos médicos? ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ 🔔 SUSCRÍBETE para no perderte ninguna noticia: https://youtube.com/@fjqg?sub_confirmation=1 📡 SÍGUEME EN TODAS LAS REDES: 🌐 Web & blog → https://mybestia.com ▶️ YouTube → https://youtube.com/@fjqg?sub_confirmation=1 📸 Instagram → https://www.instagram.com/quintinogiaia/ 🎵 TikTok → https://www.tiktok.com/@mybestia 💬 Telegram → https://t.me/franciscoquintinogarcia_bot 🦋 Bluesky → https://bsky.app/profile/mybestia.bsky.social 🐘 Mastodon → https://mastodon.social/@mybestia 💼 LinkedIn → https://www.linkedin.com/in/franciscoquintino/ ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ 📂 MÁS VÍDEOS DEL CANAL: 🤖 IA y Automatización → https://youtube.com/@fjqg/videos 🔐 Ciberseguridad & Hack → https://youtube.com/@fjqg/videos 📱 Tech y Gadgets → https://youtube.com/@fjqg/videos ⚙️ Desarrollo & Código → https://youtube.com/@fjqg/videos ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ 📌 SOBRE MYBESTIA — Francis Quintino Soy Francis Quintino, creador de contenido tech desde Palma de Mallorca 🌴 Especializado en inteligencia artificial, automatización con IA, ciberseguridad, pentesting autónomo, creación de páginas web con IA, e integración de IA en empresas. Todo el contenido de mybestia.com se genera con pipelines de IA propios — sin relleno, sin clickbait, datos reales. ▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬ 🏷️ MYBESTIA, inteligencia artificial, medicina, SOP, síndrome de ovario poliquístico, salud femenina, tecnología, innovación, ciencia, diagnóstico médico, algoritmo, machine learning, salud tech, medicina personalizada, salud 2024, Big Data salud, IA aplicada, ciberseguridad datos, bienestar, avances científicos #IA #SaludTech #SOP #MedicinaPersonalizada #Ciencia #MyBestia
Como o "cérebro" da IA funciona? #inteligenciaartificial #chatgpt #openai
Como a IA moderna funciona? Uma rede neural artificial é um modelo matemático inspirado no funcionamento do cérebro humano. Essa é uma das bases tecnológicas de sistemas modernos de IA, incluindo LLMs como o ChatGPT. #largelanguagemodels #ia #neuralnetworks
The AI Factory: Engineering Modern LLM Inference Pipelines | Uplatz
Modern AI systems are no longer simple models running isolated predictions—they operate like massive digital factories processing billions of requests, orchestrating GPUs, managing memory, and delivering intelligent responses at global scale. In this video, we explore “The AI Factory” and break down how modern LLM inference pipelines are engineered for performance, scalability, and efficiency. This video is by Uplatz. You’ll learn how large language model inference works behind the scenes, from token generation and request routing to distributed GPU execution and response optimization. We explain why inference engineering has become one of the most critical challenges in the generative AI era. The video dives into core components of modern inference pipelines including model serving, batching strategies, KV cache management, GPU scheduling, vector databases, retrieval-augmented generation (RAG), and low-latency orchestration systems. You’ll understand how organizations optimize infrastructure to reduce inference costs while maintaining performance and responsiveness. We also explore technologies and frameworks used in production AI systems such as Kubernetes, Ray, and vLLM for scalable AI deployment and inference acceleration. Additionally, we discuss concepts like model quantization, mixture-of-experts (MoE) architectures, inference parallelism, autoscaling, observability, and AI infrastructure optimization. Learn how companies engineer AI platforms capable of serving millions of users across enterprise applications, copilots, AI agents, and multimodal systems. Whether you're an AI engineer, platform architect, DevOps professional, cloud engineer, researcher, or technology enthusiast, this video provides a practical and structured understanding of how modern LLM inference factories operate at scale. For full course browse https://uplatz.com/online-courses #LLM #GenerativeAI #AIInfrastructure #MLOps #LLMOps #ArtificialIntelligence #vLLM #GPUComputing #AIEngineering #Uplatz ---------------------------------------------- 🌐 Welcome to Uplatz – Your Gateway to Career Transformation! To access full courses or training bundles: 🌐 https://uplatz.com 📧 support@uplatz.com 🎓 About Uplatz Uplatz is a global leader in online IT and professional training, offering comprehensive courses in AI, machine learning, data science, cloud computing, cybersecurity, and enterprise technologies such as SAP, Oracle, Salesforce, and ServiceNow. With expert-led programs and real-world learning paths, Uplatz empowers learners and organizations across 190+ countries to build future-ready skills and thrive in the digital era. 📘 Explore Uplatz Course Portfolio Learn the most in-demand and emerging technologies with Uplatz: ✅ AI & Machine Learning – Agentic AI, LLMs, LangChain, Deep Learning, MLOps, LLMOps ✅ Cloud & DevOps – AWS, Azure, GCP, Docker, Kubernetes, Terraform, CI/CD ✅ Data & Analytics – Data Science, Data Engineering, Power BI, Tableau, Big Data (Spark, Kafka) ✅ Programming & Frameworks – Python, FastAPI, Django, Java, JavaScript, SQL ✅ Cybersecurity & Blockchain – Ethical Hacking, Cloud Security, Zero Trust, Blockchain & Web3 ✅ IoT & Embedded Systems – IoT Platforms, Edge Computing, Embedded C, Microcontrollers ✅ ERP & CRM – SAP (all modules), Salesforce, Oracle ERP, Microsoft Dynamics ✅ Web & App Development – Full-Stack Development, React, Angular, Node.js, Flutter 🎓 Master cutting-edge skills. Build your tech career with Uplatz. 🌐 Learn more: https://uplatz.com 🎯 Why Choose Uplatz ✔️ Job-focused, project-based learning ✔️ Globally recognized certifications ✔️ Lifetime access & affordable pricing ✔️ Career guidance and mentorship 🔔 Subscribe for weekly tech tutorials, demos, and success stories. 📲 Follow us on LinkedIn, Instagram, Twitter, and Facebook. #Uplatz #Tech #Technology #MachineLearning #CloudComputing #Learning
Dual Vision - Computer Vision. Inteligencia que ve. Tecnología que Transforma.
Giovanni Pais Lucero es Ingeniero Civil Eléctrico y Magíster en Ciencias de la Universidad de Chile, especializado en visión por computadora, robótica e inteligencia artificial aplicada a entornos industriales. Durante su formación fue reconocido entre los mejores alumnos de su generación y obtuvo la Beca Nacional de Magíster ANID en 2020. Es cofundador y CEO de Dual Vision, startup tecnológica chilena enfocada en desarrollar soluciones de inteligencia artificial y video analítica para mejorar operaciones industriales, reduciendo costos y aumentando la eficiencia mediante monitoreo y control automático en tiempo real. La empresa desarrolla tecnologías aplicadas a manufactura, logística, energía y minería, integrando visión computacional, Edge AI y sistemas autónomos de monitoreo. Dual Vision ha sido reconocida en distintos programas y concursos de innovación, adjudicándose fondos públicos y privados para el desarrollo y escalamiento de sus soluciones tecnológicas. Su experiencia combina investigación aplicada, desarrollo tecnológico y emprendimiento, liderando la implementación de sistemas de visión computacional en entornos reales de operación. Ha participado en programas de innovación, aceleración e innovación abierta, incluyendo Startup Chile Growth 10 (2025), trabajando junto a empresas e instituciones del ecosistema tecnológico e industrial. Además, representó a Chile en competencias internacionales de robótica RoboCup en 2017 y 2019, participando en el desarrollo de sistemas autónomos e inteligencia artificial. Actualmente lidera el crecimiento y escalamiento de Dual Vision, impulsando soluciones que integran inteligencia artificial, procesamiento en borde (Edge AI) y monitoreo autónomo para transformar procesos físicos en información accionable y automatizar la supervisión operacional en tiempo real.