Overview
Explore the journey towards AI singularity in this 48-minute conference talk from NDC Conferences. Delve into the concept of Artificial General Intelligence (AGI) and its distinction from narrow AI. Gain intuitive understanding of AI fundamentals, problem-solving approaches, and the mathematical basis of AI solutions. Compare Machine Learning and Deep Learning, focusing on image recognition through Convolutional Neural Networks (CNN). Discover state-of-the-art CNN models, transfer learning, and distributed training techniques. Examine the limitations of CNN and recent research advancements. Investigate generative AI, including Variational Auto-encoders and GANs, and their applications in image reconstruction, mind reading, and super-resolution. Learn about Natural Language Processing advancements, word embedding, and AI-generated poetry using LSTM. Analyze the challenges in creating AGI, including knowledge representation and transfer across domains. Consider the potential need for a centralized/cloud brain model to overcome current limitations in artificial intelligence development.
Syllabus
Intro
A LITTLE BIT ABOUT MYSELF
GENERATING BABY NAMES
OUTLINE
HOW WE SOLVE PROBLEM TRADITIONALLY
WHATS THE PROBLEM WITH THIS?
SOLVING PROBLEM THE AI WAY
STEPS
ARTIFICIAL INTELLIGENCE
MACHINE LEARNING
DEEP LEARNING
Convolution Neural Network
Inception V3
Transfer Leaming
TRANSFER LEARNING
DISTRIBUTED TRAINING
TRAINING CYCLOPS
IMAGE GENERATION
AUTOENCODER
BIG LUGGAGE
SMALL BAG
MEMORY GAMES
LATENT SPACE DISTRIBUTION
MORPHING
LATENT VECTORS
ADDING GLASSES
REMOVING GLASSES
GAN PRO AND CON
SINGULARITY
UNDERSTANDING IS NOT A REASONING
Taught by
NDC Conferences