Understanding Artificial Intelligence

by Scientific American (Creator), Rodney Brooks (Foreword), Sandy Fritz (Compilation)

Drawn from the pages of Scientific American and collected here for the first time, this work contains updated and condensed information, made accessible to a general popular science audience, on the subject of artificial intelligence.

Genres: Science

90 pages, ebook

First published March 1, 2002

About the author

Scientific American , as an institutional author, is a popular science magazine founded by Rufus M. Porter and controlled by Nature Publishing Group since autumn, 2008. Mariette DiChristina has been editor-in-chief since December, 2009.

Reviews

[Brian · Rating 4 out of 5]

Understanding Artificial Intelligence is a collection of ten essays first published in Scientific American. The essays are written by different authors, including Marvin Minsky, “the mastermind of artificial intelligence” to whom the final chapter in the book is dedicated.

The collection was published in 2002, and is dated in terms of the technology and ideas, but it makes for a fascinating read, allowing us to look back at how experts thought artificial intelligence would develop over the coming decades.

Some of the predictions turned out to be accurate, but many are wildly out in terms of the expected timeframe, and I found it especially interesting to recognise exactly how far we haven’t come in the last twenty years.

An area in which we have progressed is facial recognition. One of the earlier chapters claimed it was all but impossible to automate recognition of a friend’s face, as the rules for recognising a face could not be written down. Nowadays, of course, many of us take this capability for granted as part of the every day authentication mechanism we use to access our mobile phones.

I would recommend reading Understanding Artificial Intelligence. Some of the chapters are technical and difficult to follow, but the collection provides an enlightening glimpse back in time into the field of AI and what we thought it would become.

[to'c · Rating 3 out of 5]

This is an old book so don't read it to get up-to-date on the latest AI. But it's an informative look back on the hopes and dreams. It also contains many articles that are amazingly relevant even today. You may be surprised to find that the current fascination with neural networks is approximately 70 years old! Yeah. It's just taken that long for the hardware to catch up to the computational needs of mimicking the human brain.

The 3 star rating is jut because it is an older book and mostly out-of-date. But I found it a good fast and fun read anyway.

[Zegsta · Rating 3 out of 5]

Una serie de editores de Scientific American comparten una colección de sus artículos de la revista en un libro, y vale la pena leerlo. La escritura no es muy compleja y no entran en muchos tecnicismos, pero no por ello dejan fuera los temas que ahora nos suenan tan conocidos (especialmente a aquellos que trabajan en el ámbito) de la lógica difusa, Machine Learning, temas de súper conductores o nanotecnología.

Fue retador el libro, pero está rápido (tomé un día para leerlo). Creo que una de las partes más rescatables de él son su inicio, donde hacen una crítica a la prueba de Turing para evaluar cuando una máquina puede pensar, y los últimos capítulos donde hablan de los límites entre la mente humana y la mente biónica del futuro - habrá muchos cuestionamientos tecnológicos y éticos al respecto.