AI and Life in 2030 (Mis)interpreted by James Fan fanj@hellovera.ai - PowerPoint PPT Presentation

AI and Life in 2030 (Mis)interpreted by James Fan fanj@hellovera.ai

Outline 1. History of AI 2. Study overview and summary 3. Defining AI 4. Trends in AI

AI will kill us all! AI is awesome!

Birth of AI: 1956 Dartmouth Workshop

1956-1973

1974-1980

Little Computing Underestimate Power d Difficulties “Build a vision system in one summer” Translate: “The spirit was willing but the flesh was Result: weak” Limited “The vodka was good, but the meat was rotten.” Intelligence

1980-1987

1987-1993

1997: IBM 2011: IBM Watson 2030: What’s Deep Blue beat defeated top next? reigning world Jeopardy! Chess champions. Champion 2005: Stanford 2016: Google won DARPA AlphaGo beat a Grand top human Go Challenge player

“... takes a much longer time coming than you think, and then it happens much faster than you would have thought.”

100 Year Study on AI Panel found no cause for concern that The overarching purpose of the One AI is an imminent threat to humankind. Hundred Year Study’s periodic expert review is to provide a collected and No machines with self-sustaining long- connected set of reflections about AI term goals and intent have been and its influences as the field developed, nor are they likely to be advances. developed in the near future. Instead, increasingly useful applications of AI, with potentially profound positive impacts on our society and economy are likely to emerge between now and 2030, the period this report considers.

What is AI and What is not AI ● GPS navigation Artificial intelligence is that ● Book recommendation activity devoted to making ● Playing game machines intelligent, and ● Driving vehicle intelligence is that quality that ● Web search enables an entity to function ● Wikipedia appropriately and with ● Internet of Things foresight in its environment. ● Operation Research -- Nils J. Nilson ● Tax filing software

Logic & Machine Search Reasonin Techniques Learning g Natural Planning Vision Speech Language Areas Processing Question AI in Machine Demos/ Robotics Answerin Games Translation g Applications

Machine Learning Supervised Unsupervised Reinforcemen Learning Learning t Learning Artificial Neural Network Deep Learning Natural Language Robotics Vision Speech Processing

A Brief History of Neural Network 1957: Frank 1974-1980: AI Winter 1987-1993: AI Winter Rosenblatt invented perceptron 1969: Minsky and Papert 1980s: Multi-layer 2000s: Deep learning proved limitations of perceptron perceptron

What Next? ● Buchanan, Bruce G. A (Very) Brief History of Artificial Intelligence AI Magazine 2005 ● https://www.coursera.org/learn/ machine-learning# ● https://www.coursera.org/learn/ neural-networks