Artificial General Intelligence by Julian Togelius.
Julian holds a BA in Philosophy, Computer Science, and Psychology from Lund University, MSc from the University of Sussex, and a PhD in Computer Science from the University of Essex. He was an associate professor at the Center for Computer Games Research, IT University of Copenhagen. Today he is an associate professor at the Department of Computer Science and Engineering at the New York University.
This book is a wide overview of the concept of Artificial General Intelligence (AGI) which predicts AI will hold the ability to understand, learn, and apply knowledge like humans.
I certainly found this book very enjoyable to read about where the general direction of a robust AI will impact society. There are of course examples today including Google’s Go which not only defeated the world’s greatest Go player, but generated moves that caught all humans off guard. Yet that same AI system cannot beat you in Tetris.
Dzejla and Emin are certainly providing a complex topic in a friendly manner to make complex concepts easy to understand. Readers will learn about deploying the correct type of database solution for your organization’s applications. This was a very enjoyable book to read.
Will we see AGI in our lifetime?
Julian is illustrating multiple approaches and definitions of general human intelligence from psychology, ethology, and computer science perspectives. He is revealing two technical views to a AGI. First, Foundation Models which are self-supervised that enable models to learn from vast amounts of data. Secondly, Open-Ended Learning which focuses on learning via dynamic virtual environments. This permits continuous adaptation and growth.
I would recommend How Smart Machines Think by Sean Gerrish. as he writes about how today’s super AI computer can beat you at Go, but will be defeated by any user playing Tetris.
In the closing chapters address broader questions about artificial general intelligence and if these systems would be conscious. Actually, Julian writes about the risks AGI may indeed pose to humanity. Are we aware how AGI could in fact alter our society in unplanned ways.
In conclusion, Julian is providing an understanding of the future AI world that we will (one day in the future) confront. Do we understand the technical and ethical challenges involved?