The AI Economy by Roger Bootle is a very good read. Roger is Chairman of Capital Economics, Europe's largest macroeconomic consultancy and he also writes a column for The Daily Telegraph.
Throughout the book, Bootle gives a clear introduction to the relevant topics surrounding AI, from the different types of universal basic income to the supporting arguments and critiques of Piketty's work on inequality. His style involves clearly laying out theories and their critiques, sometimes even in bullet point lists, before giving his own judgement.
The book ranges from discussions over the possibility of deficient demand impacting on bond yields to what school subjects will be desired in the future. This makes it an informative book for everyone from traders to students.
He remains optimistic throughout the book whilst still recognizing the challenges that the new AI economy introduces, steering a middle course between the predictions of Utopia and Dystopia often associated with the subject to offer a realistic and grounded vision of the future. For example, while confident that advances in technology will deliver higher living standards, as they have done throughout history, the immediate impact will likely make people worse off due to structural unemployment. Nonetheless, he is adamant that “AI will not spell Armageddon for employment.”
He is helped by his decision to leave discussion of 'The Singularity' (when AI becomes fully conscious) until the end. If we do reach this point, which Bootle suggests that we never will, then it would be almost impossible to predict what would happen next as human Labour quickly becomes extinct and humanity is either destroyed by technology or assimilated into it.
On the role of the state in the development of AI, Bootle gives a fairly free market approach. Although he believes the government should neither tax nor subsidise progress in this field, he gives a strong case for drawing up an appropriate legal framework as well as restrictions on development of AI in areas that could be used by criminals or for weapons.
When it comes to education, he identifies an increasing importance in the humanities and how they develop critical thinking skills. He believes that technology will dramatically improve the education sector from freeing up teachers from more mundane tasks to how online classes can bring education to a wider audience.
He moves on to cover UBI which is probably the most discussed policy area when thinking about the AI economy. Bootle argues that any implementation of UBI would be unacceptably low or unacceptably expensive and thus unfeasible. However, although he covers negative income taxes earlier in the specific chapter it might have been good if he could then have come back to them to either offer them as an alternative to the unfeasible UBI or to give an argument on why it would also be unfeasible. His only criticism of negative income taxes, which he credits to Friedman himself, was simply that they incentivise people to lie about how much they earn in order to get more money. This incentive however already exists in any system that has direct taxation and it also fairly hard to do with the PAYE system. He may have been more comfortable in not covering this area because he predicts that technological change will not necessarily either decrease inequality nor remove work - at least until the singularity. We all hope that this to be true but as he himself says, this may be an area that could do with some more academic research.
He makes the amusing observation of what we view as AI shifts all the time and much of what we consider normal computer processes now would have been considered 'AI' a few years ago. Although it is difficult to predict the future, Bootle does give a very grounded and undoubtedly realistic view of the future. This, however, is overlooking the real value of the book, which is to enable the reader to enter the debate themselves by introducing the key concepts and issues. As an introduction to the debate as a whole, this book is invaluable.