I am a big fan of Stephen and is beautiful visualisations. I think this book serves as a good counterpoint to the amount of marketing disguised as "big data". It raises several critical and important questions organisations should ask themselves before spending millions on the new initiative.
I do not however agree that all of Big Data is pure marketing hype. When you strip away the layers there is some significant changes that have happened since the introduction of Hadoop and cloud based storage and processing which Stephen fails to acknowledge. For example: - Cost of storage has plummeted and it has brought analytical horsepower (if not capability) within the reaches of much smaller organisations. Not only those who could afford the "big iron" of the past. - Availability of curated public data-sets has dramatically increased allowing forward thinking organisations to use these in combination with internal data for better "data sense-making" (Stephen's term) - There has been an explosion in interest, tutorials and people willing to learn more - so the marketing hype has helped, though as usual it has over-promised. - One more nuanced area where the "domain expertise" is being challenged in in the deep learning space - to be fair this is still on the cutting edge of science with few organisations having the skill-sets to fully utilise this. Here learning algorithms have surpassed human capabilities in several areas (Go as well as retro games are the better known examples). And these have been driven more by computing power and the "proto-science" of deep learning than necessarily experts in the game.
I would recommend it for people who like to have the fully perspective of what topics mean and are able to hold both the promise and cynicism necessary to navigate the world of big data.