Explore insights from Lukas Biewald, founder of Figure Eight and Weights & Biases, in this 44-minute conference talk from the Full Stack Deep Learning March 2019 bootcamp. Dive into Biewald's journey in the machine learning industry, from founding the largest data labeling company to his current role in developing tools for ML practitioners. Gain valuable perspectives on AI's progress, industry predictions, and real-world applications in speech recognition, face recognition, and personalized medicine. Examine the challenges faced in data labeling and deep learning projects, and learn about the importance of training data, data cleaning, and human-in-the-loop systems. Discover insights on breakthrough technologies, free datasets, and the evolving landscape of AI tools and platforms. Understand the practical aspects of shipping deep learning projects and the future of AI applications in various industries.
Lukas Biewald on Founding Weights & Biases and FigureEight - Full Stack Deep Learning - March 2019
Overview
Syllabus
Introduction
Lukass career
Hacker cred
AI is working
People making ludicrous predictions
Speech
Brands
CocaCola
Face Recognition
Blue River
Oxy
Personalized Medicine
Siobhans OS
Problems with FigureEight
Deep Learning vs ML
Space Odyssey
Netflix Prize
Selfdriving cars
Machine learning can be unpredictable
Puppy vs Bagel
Alphago
Tesla
Criminal Risk Scores
How to Inspect Systems
Adversarial Systems
Machine Learning Requires Training Data
Peter Norvig
Not hiring annotators
Cleaning your data
Data science joke
Andre Karpati
Lukass last compiler error
Why do we do it
Looking at more data
Realworld breakthroughs
Free datasets
Humans vs Computers
Design Patterns
Training
Human Loop System
Real Case Study
Better Tools Platforms
SageMakers
YouTube Channel
Medical Applications
FigureEight
ImageNet
Cancer Images
Microwave Oven
Selfdriving suitcase
How to ship deep learning projects
Contact Lukas Biewald
Lukas Biewald on Labeling
Taught by
The Full Stack
