by Arvid Narayanan & Says Kapoor

WILL AI BE GOOD FOR THE FUTURE?

Review by Gunnel Minett

There is no doubt that Artificial Intelligence is having an impact on our lives already. But will it be a change for the better or will it lead to a world where computers take over and become rulers of the world? As the authors point out, this debate has already begun. And the arguments range from very positive to a variety of nightmare scenarios. 

One of the first things to clarify, the authors point out, is that:”…AI refers to a vast array of technologies and applications, most people cannot yet fluently distinguish which types of AI are actually capable of functioning as promised and which types are simply snake oil.” (P 2) There are also clearly exaggerated believes regarding the possibilities offered by AI, by many of the people working with AI. Meanwhile other people are predicting nightmare scenarios where AI will take over in a vicious and mindless way and simply kill off human beings as no longer necessary. 

Neither prognosis is true, the authors argue. Computers are not that clever, and definitely not vicious, unless so programmed, and they add: “…and we will make the argument that predictive AI not only does not work today but will likely never work, because of the inherent difficulties in predicting human behaviour.” (P3) Instead they point to the human factor, or more specifically the owners of the companies which are developing these techniques. Unlike conventional science, where new developments are screened by peer reviews, the tech companies tend to get away with making bold statements about their products. AI does not require the same level and amount of independent evidence to substantiate their claims as is required in other areas of science. This can, and has already in some cases, led to serious problems, in particular when it comes to predictive AI. Unlike the human brain, AI can’t predict the future very well (usually equivalent to chance). These computer programs simply scan previously collected data and draws conclusions from it and presents these conclusions as predictions. But, as we all know, unforeseen events can happen. One such example is COVID, where we were all faced with a new scenario that demanded a different way of life. The predictive AI had very little useful historic data to help it and offered weird suggestions such as drinking bleach (which sadly even some politicians thought was a good idea). 

The book gives several similar examples as to how AI is being hailed by decision-makers and companies alike as the new way forward in many aspects of life. However, the book argues, there is little evidence that such claims are true. This is for financial reasons rather than any lack of computing power. There are many ways in which AI can be used as a very effective tool, but in many cases it requires careful scrutiny by human beings before reliably positive results can be achieved. And this, of course, requires more investment than the companies are willing to allocate.

An important area dealt with in the book is how AI is used in medicine. Unfortunately the authors do not address the more philosophical issues around psychological ill health. Emotions and consciousness, for instance, are not discussed. But this is an area where it is even more interesting to understand the difference between AI-aided care and the role of human carers. It is already a trend in medicine to treat psychological ill health with drugs and to label psychological problems as ‘syndromes’ requiring expert help. 

From the AI perspective there are many advantages in this approach, not least financial. But as we traditionally used to know, life is sometimes more complex: not all psychological ill health needs drugs or labels. Sometimes a friendly talk or simply human contact may help, while, of course, more serious problems will need more expert help. In particular, a growing number of children are growing up in a computerised world (which seems to have a negative impact). Consequently, the authors argue, we may need to take a  closer look at how we use AI. If not, what might be a huge help for us may turn into computers controlling us, rather than vice versa. In order to achieve this we need to understand how AI works so that we can use it to our advantage and not buy into it as a modern version of snake oil. This is the authors main point and their motive for writing this book.

Princeton University Press, Oxford, 2024, 360 pp, Hardcopy, ISBN 978-0691249131