Book Review : Prediction Machines: The Simple Economics of Artificial Intelligence By Ajay Agrawal ; Joshua Gans and Avi Goldfarb

The relevance of this book has just sky-rocketed considering our current uncertain state of the world. Even though I read this book a while ago, when I revisited my bookshelf under lockdown, it once again caught my eye. So if you are looking for some interesting and relevant reading in the next few weeks, you can download the book from various sites, go to www.predictionmachines.ai  and click the ‘Buy the book’ button which will lead you to a few links of where you can get the book (no need to go out with the Kindle and Audiobook versions available).

Technology has already impacted our world dramatically in the past few years, but the rate of change is accelerating exponentially, impacting social behaviours. There are many articles and discussions hypothesising how and what behaviours will change for the long-term due to this pandemic, but now, more than ever, business leaders need to consider the imminent rise of AI and how their business will approach and navigate this new normal.

Artificial Intelligence (AI) is often portrayed in a Sci-Fi-movie-kind-of-way and some believe that machines will take over the world, but in this book three economists look at AI through their lens and demystify it by viewing it as a drop in the cost of prediction. They provide a framework for all business leaders (CEOs, managers, policy makers, investors and entrepreneurs) to approach the unknown.

5 Key take-outs:

• Prediction is at the heart of making decisions within uncertainty.
• Through applying simple economics and viewing prediction as a commodity, Predication Machines offers clear insights regarding the business implications of cheaper prediction.
• The impact of prediction tools will increase productivity as machines and humans have distinct strengths and weaknesses, but combining the strengths of each means ‘Predicting Better Together’; and being even marginally better, can lead to a big difference in market share and revenue.
• Prediction is not a decision and as prediction becomes better, the value of human judgement will increase.
• Businesses will start by adopting AI tools to support their strategy, but then these tools may become powerful enough to change the strategy. 

AI is already being used in a number of industries such as Google, Uber and commercial airline flights, but the full potential of AI is yet to be seen. Considering the current pandemic, there have also been numerous articles written about how AI can democratise the healthcare sector and the education system, improve food-security and Korea has even used it to flatten the curve. Pre-pandemic there are also examples of how it can accelerate computational chemistry and climate science. It may be a good time to get ahead of this curve.

A really insightful and valuable read and they offer a free tool on their website called the ‘AI canvas’ which can guide one through the process. (www.predictionmachines.ai)