A Human Algorithm

The author is, above all things, hopeful, and that shines through the work. Coleman is careful to cover first the potential consequences of AI  developed with short sighted profit goals, with less than admirable motives in mind. (This was published in 2019, it's arguable we're currently experiencing some of those consequences listed.)
 She then looks at the variety of positive outcomes (from tech currently in use/testing to full utopic society) that could result from ensuring humanity determines how/to infuse values into AI, sets up law/policy to regulate the creation and use of AI, but primarily the need for humanity itself to change how we think and act in order to ensure that as partners or thought leaders, AI will act for the good. 
I was pleasantly surprised how often the line between animals and AI was drawn in relation to the importance of recognizing and respecting non-human intelligences. 
 The writing is bulked out by multiple examples and quotes supporting the theory or practice of all matters discussed, (70 pages of footnotes in a 300 page book). I'm not sure the book needed the history lesson prevalent in the first couple chapters. The tone vacillates between earnest and melodramatic, so occasionally I got distracted from the good point being made because I felt like it was being drilled into me. 
That being said, it's refreshing to see someone care, without reservation, someone be optimistic and carry that optimism to the best possible vision of the future. We need more of that.