VIDEO
GPT-3 has 175 billion parameters/synapses. Human brain has 100 trillion synapses. How much will it cost to train a language model the size of the human brain?
REFERENCES:
[1] GPT-3 paper: Language Models are Few-Shot Learners
https://arxiv.org/abs/2005.14165
[2] OpenAI's GPT-3 Language Model: A Technical Overview
https://lambdalabs.com/blog/demystifying-gpt-3/
[3] Measuring the Algorithmic Efficiency of Neural Networks
https://arxiv.org/abs/2005.04305
Other Videos By Lex Fridman 2020-08-24 Math Meme Review with Grant Sanderson (3Blue1Brown) 2020-08-23 Grant Sanderson: Math, Manim, Neural Networks & Teaching with 3Blue1Brown | Lex Fridman Podcast #118 2020-08-21 Best hidden feature of Python | Chaining comparison operators 2020-08-20 Sheldon Solomon: Death and Meaning | Lex Fridman Podcast #117 2020-08-19 Lex Fridman Podcast | Now in ASCII + Name Change 2020-08-16 Sara Seager: Search for Planets and Life Outside Our Solar System | Lex Fridman Podcast #116 2020-08-14 Dileep George: Brain-Inspired AI | Lex Fridman Podcast #115 2020-08-11 2+2=5 in Java 2020-08-09 Russ Tedrake: Underactuated Robotics, Control, Dynamics and Touch | Lex Fridman Podcast #114 2020-08-07 Best feature of Python | List Comprehensions 2020-08-01 GPT-3 vs Human Brain 2020-07-31 Manolis Kellis: Human Genome and Evolutionary Dynamics | Lex Fridman Podcast #113 2020-07-29 Ian Hutchinson: Nuclear Fusion, Plasma Physics, and Religion | Lex Fridman Podcast #112 2020-07-27 The most controversial Python feature | Walrus operator 2020-07-26 Richard Karp: Algorithms and Computational Complexity | Lex Fridman Podcast #111 2020-07-24 When will the 200 billionth person be born? 2020-07-21 Jitendra Malik: Computer Vision | Lex Fridman Podcast #110 2020-07-20 Will Javascript Take Over the World? | Brian Kernighan and Lex Fridman 2020-07-19 Neural networks learning spirals 2020-07-19 C Programming Language | Brian Kernighan and Lex Fridman 2020-07-18 Brian Kernighan: UNIX, C, AWK, AMPL, and Go Programming | Lex Fridman Podcast #109
Tags: openai
neural network
deep learning
machine learning
language models
nlp
gpt 3 demo
gpt lex