“Science is under attack and scientists are losing credibility,” Gary Smith warns at the outset of ‘Distrust: Big Data, Data-Torturing, and the Assault on Science’ (Oxford University Press, £25, ISBN 9780192868459). “There are three prongs to this assault on science: disinformation, data torturing, and data mining. Ironically, science’s hard-won reputation is being undermined by tools invented by scientists. Disinformation is spread by the internet that scientists created. Data torturing is driven by scientists’ insistence on empirical evidence. Data mining is fuelled by the big data and powerful computers that scientists created.”

The diagnosis is convincing. Throughout ‘Distrust’, Smith explains clearly the phenomena that are eroding trust in science: fake news, confirmation bias, p-hacking, the myriad problems with the world of scientific publishing.

Smith has crammed his book with case studies. There are some well-worn examples and some more original ones, such as an examination of the failure of Zillow’s algorithmically driven house-flipping venture, which beautifully demonstrates the danger of trusting ignorant and narrow algorithms to replace human expertise. One memorable section on data mining demonstrates that the stock price of Urban Tea, a Chinese tea distributor, can be predicted to fall four days after Donald Trump posts tweets containing the word ‘with’. Smith is at his best when taking these concepts familiar in rationalist circles and applying them to the Data Age.

‘Distrust’ has been compared with Carl Sagan’s ‘The Demon-Haunted World’ and there is a similarity there, although it lacks Sagan’s style and originality. It is the latest instalment in that slightly tired trend of ‘skeptical’ [sic] books by the likes of Richard Dawkins and Ben Goldacre. The trend gave rise to many great books – witty, provocative, game-changing – but, looking back, they feel very much of their time. An awful lot has changed since then.

There remains a need for books that tackle irrational thinking, and Smith is fighting the good fight. After all, as he demonstrates very thoroughly, there are still astonishing numbers of people in the UK and US who believe in dangerous pseudoscience. However, there is a growing acknowledgement that people who believe irrational things will not be won over by rational arguments like those presented in ‘Distrust’. Scientists will need to look to other means of persuasion. This could include addressing in good faith the ugly parts of the history of science – the eugenics, the deeply unethical experiments – and condescending to engage non-scientists with art and storytelling.

Additionally, ‘Distrust’ might have benefitted from a stronger editorial hand to rein-in the digressions and keep it focused on the argument at hand (towards the end, the book moves beyond the ‘disinformation, data torturing, data mining’ argument to tackle AI, data privacy and other big fraught issues of today).

It is not entirely clear for whom the book is intended. Perhaps for those who already trust science, would like to learn more, and perhaps persuade others? That group will find ‘Distrust’ an affirming read filled with clear explanations and interesting case studies. They are unlikely, however, to finish it equipped to appeal to the irrational mind of an anti-vaxxer.

“Science is under attack and scientists are losing credibility,” Gary Smith warns at the outset of ‘Distrust: Big Data, Data-Torturing, and the Assault on Science’ (Oxford University Press, £25, ISBN 9780192868459). “There are three prongs to this assault on science: disinformation, data torturing, and data mining. Ironically, science’s hard-won reputation is being undermined by tools invented by scientists. Disinformation is spread by the internet that scientists created. Data torturing is driven by scientists’ insistence on empirical evidence. Data mining is fuelled by the big data and powerful computers that scientists created.”

The diagnosis is convincing. Throughout ‘Distrust’, Smith explains clearly the phenomena that are eroding trust in science: fake news, confirmation bias, p-hacking, the myriad problems with the world of scientific publishing.

Smith has crammed his book with case studies. There are some well-worn examples and some more original ones, such as an examination of the failure of Zillow’s algorithmically driven house-flipping venture, which beautifully demonstrates the danger of trusting ignorant and narrow algorithms to replace human expertise. One memorable section on data mining demonstrates that the stock price of Urban Tea, a Chinese tea distributor, can be predicted to fall four days after Donald Trump posts tweets containing the word ‘with’. Smith is at his best when taking these concepts familiar in rationalist circles and applying them to the Data Age.

‘Distrust’ has been compared with Carl Sagan’s ‘The Demon-Haunted World’ and there is a similarity there, although it lacks Sagan’s style and originality. It is the latest instalment in that slightly tired trend of ‘skeptical’ [sic] books by the likes of Richard Dawkins and Ben Goldacre. The trend gave rise to many great books – witty, provocative, game-changing – but, looking back, they feel very much of their time. An awful lot has changed since then.

There remains a need for books that tackle irrational thinking, and Smith is fighting the good fight. After all, as he demonstrates very thoroughly, there are still astonishing numbers of people in the UK and US who believe in dangerous pseudoscience. However, there is a growing acknowledgement that people who believe irrational things will not be won over by rational arguments like those presented in ‘Distrust’. Scientists will need to look to other means of persuasion. This could include addressing in good faith the ugly parts of the history of science – the eugenics, the deeply unethical experiments – and condescending to engage non-scientists with art and storytelling.

Additionally, ‘Distrust’ might have benefitted from a stronger editorial hand to rein-in the digressions and keep it focused on the argument at hand (towards the end, the book moves beyond the ‘disinformation, data torturing, data mining’ argument to tackle AI, data privacy and other big fraught issues of today).

It is not entirely clear for whom the book is intended. Perhaps for those who already trust science, would like to learn more, and perhaps persuade others? That group will find ‘Distrust’ an affirming read filled with clear explanations and interesting case studies. They are unlikely, however, to finish it equipped to appeal to the irrational mind of an anti-vaxxer.