I attended the Deep Learning & AI Assistant Summit organized by ReWork in San Francisco. The speakers in the conference were top AI experts from companies like Google, Facebook, Amazon, Uber etc.
There were also presentations by startups and a demo stage where some of the selected startups showed their work. Deep Learning is a hot and rapidly evolving area. It is a powerful subfield of machine learning, based on algorithms inspired by the brain called artificial neural networks.
The algorithms automate the extraction of representations from data, and have helped solve challenges in: computer vision, natural language processing, speech recognition, video processing and image recognition.
Personally I found a presentation by Disney Research’ Michael Carr very interesting. It was about modelling team strategies using Deep Imitation Learning. They had developed an automatic “ghosting” system for a basketball match which illustrated where defensive players should have been (instead of where they actually were) based on the locations of the opposition players and ball.
An AI assistant is a software agent that can perform tasks or services for an individual. An automatic chatbot, Apple’s Siri and Amazon’s Alexa are examples of AI assistants. Natural Language Understanding is the next big challenge to conquer to create great AI assistants. This will enable companies to build realistic conversational interfaces and smooth dialog between AI and user. There were some interesting applications for AI assistants presented in the conference. For example a company called Woebot was developed to address the growing crisis of mental health. Woebot is an automated conversational agent (chatbot) who helps you monitor mood and learn about yourself by chatting with an AI.
This was an excellent conference for networking and getting to know the current state of Deep Learning and Assistant AIs.
Written by Pekka Alatossava