London Futurists

AI overview: 2. The Big Bang and the years that followed

September 07, 2022 London Futurists Season 1 Episode 3
London Futurists
AI overview: 2. The Big Bang and the years that followed
Show Notes

In this episode, co-hosts Calum Chace and David Wood continue their review of progress in AI, taking up the story at the 2012 "Big Bang".

00.05: Introduction: exponential impact, big bangs, jolts, and jerks
00.45: What enabled the Big Bang
01.25: Moore's Law
02.05: Moore's Law has always evolved since its inception in 1965
03.08: Intel's tick tock becomes tic tac toe
03.49: GPUs - Graphic Processing Units
04.29: TPUs - Tensor Processing Units
04.46: Moore's Law is not dead or dying
05.10: 3D chips
05.32: Memristors
05.54: Neuromorphic chips
06.48: Quantum computing
08.18: The astonishing effect of exponential growth
09.08: We have seen this effect in computing already. The cost of an iPhone in the 1950s.
09.42: Exponential growth can't continue forever, but Moore's Law hasn't reached any theoretical limits
10.33: Reasons why Moore's Law might end: too small, too expensive, not worthwhile
11.20: Counter-arguments
12.01: "Plenty more room at the bottom"
12.56: Software and algorithms can help keep Moore's Law going
14.15: Using AI to improve chip design
14.40: Data is critical
15.00: ImageNet, Fei Fei Lee, Amazon Turk
16.10: AIs labelling data
16.35: The Big Bang
17.00: Jürgen Schmidhuber challenges the narrative
17.41: The Big Bang enabled AI to make money
18.24: 2015 and the Great Robot Freak-Out
18.43: Progress in many domains, especially natural language processing
19.44: Machine Learning and Deep Learning
20.25: Boiling the ocean vs the scientific method's hypothesis-driven approach
21.15: Deep Learning: levels
21.57: How Deep Learning systems recognise faces
22.48: Supervised, Unsupervised, and Reinforcement Learning
24.00: Variants, including Deep Reinforcement Learning and Self-Supervised Learning
24.30: Yann LeCun's camera metaphor for Deep Learning
26.05: Lack of transparency is a concern
27.45: Explainable AI. Is it achievable?
29.00: Other AI problems
29.17: Has another Big Bang taken place? Large Language Models like GPT-3
30.08: Few-shot learning and transfer learning
30.40: Escaping Uncanny Valley
31.50: Gato and partially general AI

Music: Spike Protein, by Koi Discovery, available under CC0 1.0 Public Domain Declaration

For more about the podcast hosts, see https://calumchace.com/ and https://dw2blog.com/