Innovation produces net-negative utility. Prefer comprehensive and robust evidence.

For writers such as Steven Pinker (for example, in his book 'Enlightenment Now'), the human condition is steadily improving, and this improvement can be picked up across a range of indicators: fewer homicides, fewer infant deaths, longer life expectancies, improving human rights, increasing literacy, increasing average education, increases in mean (and also median) wealth, fewer people in extreme poverty (less than 1 USD/ day), etc.

However, while not denying significant progress in some areas, particularly the reduction in violence, cherry-picking data to support a comforting Whiggish conclusion is unlikely to generate robust evidence.

To show that the picture is more nuanced, it is worth compiling an alternative, complementary list of dimensions where public health and environmental wellbeing are deteriorating, and highlighting that many of these are probably due to innovation: for example, the appearance and rise of lung cancer (it was relatively unheard of, even among doctors, by 1900: Proctor 2012); the problematic increase of allergies such as asthma (which are somehow conspicuously absent in Amish communities, Holbreich et aL 2012); the rapid increase of anaphylactic shock (Lee et aL 2017); the prolonged decrease in human sperm counts (Levine et al. 2017); the positive relation between new cancer cases and per capita income (which cannot be explained by improved detection potential or longer lives: Luzzati, Parenti, and Rughi 2018); the rise of child-hood obesity (Ebbeling, Pawlak, and Ludwig 2002); the global burdens posed by air pollution (Akamoto 2003), the collapse of insect populations (Kwun 2019), pervasiveness of fluorinated chemicals in the environment (Lon 2019). plastic waste accumulating in the oceans (Camay 2016); the 'annihilation' of vertebrate populations in the contra of the current human-induced mass extinction event (CehaEns. Ehrhch, and Than 2017): the increasing share of Earth's surface made desolate and uninhabitable by human activity (e.g. Chernobyl. Fultushisna. polluted battlefields such as Verdun in France (Bausinger. holmium and Preuff 2007; Coupe: 20181, the Union Carbide disaster area in Bhopal with continuing groundwater contamination, perhaps also including landfill sites, nuclear waste disposal areas, former mining Mel, landmine fields); and so on.

Rather than cherry picking data to support a preconceived view that innovation is either always good or always bad, we should seek to better understand the nature of innovation (in ceder it> better design innovation policy), and hence there is a need for a more nuanced discussion that rearticulates an old point Innovation hat a &meson as well as a rate of change (Nelson 1962). Innovation can have good and bad effects, and those positive and negative outcomes are typically unevenly distributed. Choices about innovation are therefore complex and often contested, and the selection environment that weeds Mal the 'bad' innovations is not something that can he taken for granted. Nor can it be assumed In a pluralistic society. let *lune world, that definitions of what cultists as good or bad are universally or even widely aimed, It would be concerning if people who occupy important positions in society overlook the fact that the divisions of good and bad are reared in the selection environment for innovation, and think that innovation produces good outcomes naturally.