Skip to main content

Artificial Intelligence - Melanie Mitchell *****

As Melanie Mitchell makes plain, humans have limitations in their visual abilities, typified by optical illusions, but artificial intelligence (AI) struggles at a much deeper level with recognising what's going on in images. Similarly in some ways, the visual appearance of this book misleads. It's worryingly fat and bears the ascetic light blue cover of the Pelican series, which since my childhood have been markers of books that were worthy but have rarely been readable. This, however, is an excellent book, giving a clear picture of how many AI systems go about their business and the huge problems designers of such systems face.

Not only does Mitchell explain the main approaches clearly, her account is readable and engaging. I read a lot of popular science books, and it's rare that I keep wanting to go back to one when I'm not scheduled to be reading it - this is one of those rare examples.

We discover how AI researchers have achieved the apparently remarkable abilities of, for example, the Go champion AlphaGo, or the Jeopardy! playing Watson. In each case these systems are tightly designed for a particular purpose and arguably have no intelligence in the broad sense.  As for what's probably the most impressively broad AI application of modern times, self-driving cars, Mitchell emphasises how limited they truly are. Like so many AI applications, the hype far exceeds the reality - when companies and individuals talk of self-driving cars being commonplace in a few years' time, it's quite clear that this could only be the case in a tightly controlled environment.

One example, that Mitchell explores in considerable detail are so-called adversarial attacks, a particularly AI form of hacking where, for example, those in the know can make changes to images that are invisible to the human eye but that force an AI system to interpret what they are seeing as something totally different. It's a sobering thought that, for example, by simply applying a small sticker to a stop sign on the road - unnoticeable to a human driver - an adversarial attacker can turn the sign into a speed limit sign as far as an AI system is concerned, with potentially fatal consequences.

Don't get me wrong, Mitchell, a professor of computer science who has specialised in AI, is no AI luddite. But unlike many of the enthusiasts in the field (or, for that matter, those who are terrified AI is about to take over the world), she is able to give us a realistic, balanced view, showing us just how far AI has to go to come close to the more general abilities humans make use of all the time even in simple tasks. AI does a great job, for example, in something like Siri or Google Translate or unlocking a phone with a face - but AI systems still have no concept of, for example, understanding (as opposed to recognising) what is in an image. Mitchell makes it clear that where systems learn from large amounts of data, it is usually impossible to uncover how they are making decisions (which makes the EU's law requiring transparent AI decisions pretty much impossible to implement), so we really shouldn't trust them with important outcomes as they could easily be basing their outcomes on totally irrelevant inputs.

Apart from occasionally finding the explanations of the workings of types of neural network a little hard to follow, the only thing that made me raise an eyebrow was being told that Marvin Minsky 'coined the phrase "suitcase word"' - I would have thought 'derived* the phrase from Lewis Carroll's term "portmanteau word"' would have been closer to reality.

There have been good books on the basics of AI already, and excellent ones on the problems that 'deep learning' and big data systems throw up. But without a doubt, Mitchell's book sets a new standard in giving an understanding of what's possible and how difficult it is to go further. It should be read by every journalist, PR person and politician before they pump out yet more hype on the AI future. Recommended.

* Polite term
Hardback:   

Kindle 
Using these links earns us commission at no cost to you

Comments

Popular posts from this blog

Magdalena Zernicka-Goetz - Four Way Interview

Magdalena Zernicka-Goetz is Professor of Mammalian Development and Stem Cell Biology at the University of Cambridge, and Bren Professor in Biology and Bioengineering at Caltech. She has published over 150 papers and book chapters in top scientific journals and her work on embryos won the people’s vote for scientific breakthrough of the year in Science magazine.Her new book, co-authored with Roger Highfield, is The Dance of Life: symmetry, cells and how we become human.

Why science?

I fell in love with biology when I was a child because I loved doing experiments and seeing what happened. It was fascinating and enormous fun. I also fell in love with art at the same time. Art and science are both based on experiments and uncovering new paths to understand the world and ourselves. Why do we think the way we think? Where do our feelings come from? Is the 'right' answer always right? Where do we come from? How do parts of our body communicate with each other?  What is the nature of ti…

The Dance of Life - Magdalena Zernicka-Goetz and Roger Highfield ****

There is without doubt a fascination for all of us - even those who can find biology a touch tedious - with the way that a tiny cellular blob develops into the hugely complex thing that is a living organism, especially a human. In this unusual book which I can only describe as a memoir of science, Magalena Zernicka-Goetz, assisted by the Science Museum's Roger Highfield, tells the story of her own career and discoveries.

At the heart of the book, and Zernicka-Goetz's work, is symmetry breaking, a topic very familiar to readers of popular physics titles, but perhaps less so in popular biology. The first real breakthrough from her lab was the discovery of the way that a mouse egg's first division was already asymmetrical - the two new cells were not identical, not equally likely to become embryo and support structure as had always been thought.  As the book progresses, throughout the process of development we see how different symmetries are broken, with a particular focus on…

Meera Senthilingam - Four Way Interview

Meera Senthilingam is currently Content Lead at health start-up Your,MD and was formerly International Health Editor at CNN. She is a journalist, author and public health researcher and has worked with multiple media outlets, such as the BBC, as well as academic institutions, including the LSHTM and Wellcome Trust. She has Masters Degrees in Science Communication and the Control of Infectious Diseases and her interests lie in communicating global health issues to the general public through journalism and working with global health programmes. Her academic research to date has focused on tuberculosis, particularly the burden of drug-resistant tuberculosis and insights into the attitudes and behaviours of the people most affected. Her new book is Outbreaks and Epidemics: battling infection from measles to coronavirus.

Why science?

I have always found science fascinating and have always had a strong passion for it. My friends in high school used to find it amusing to introduce me to people…