This document provides an overview of machine intelligence and its everyday applications. It discusses artificial narrow and general intelligence, machine learning approaches including supervised and unsupervised learning, and deep learning and neural networks. It also demonstrates examples of computer vision, natural language processing, machine translation and other AI applications like cancer detection, image captioning and voice synthesis. The conclusion encourages embracing AI to improve applications.
“Old-school” machine learning via matrix factorization and loss function
Show NLU and how we build Botfred
AI levels up human capabilities
Algorithm can differentiate between male / female
Algorithm can detect the different parts although, for example, the sea and the suite are blue
Algorithm can make mistakes, as it is no cat but quite similar
From Google Pixel event this year
Left caption from 2 years ago
Deep Learning rocketed voice synthesizing - listen!
(with this new problems might occur such as social engineering by emulating voices of other persons)
Could be interesting to show what buzz this event created. In minute approx. 3 they shortly explain what Deep Learning is
http://makeagif.com/youtube-to-gif
From Google Pixel event this year
Phrase-based translation is based on statistical approaches
Neural machine translation looks like human translation
From Google Pixel event this year
Phrase-based translation is based on statistical approaches
Neural machine translation looks like human translation
Could be interesting to show what buzz this event created. In minute approx. 3 they shortly explain what Deep Learning is