read this as followup on the walter isaacson book on crispr and jennifer doudna, hoping for deeper discussion of unique ethical issues raised by heritable germline genome editing that he only briefly touches on in that book. there are only two interesting chapters (4 and 5), but they mainly contain summary judgements rather than argument. the others are background science explanation and cautious committee-speak about the need to find consensus.

some thoughts in or occasioned by this book:

* the treatment versus enhancement distinction is not helpful, fails to illuminate even simple examples like making people less susceptible to alcoholism

* we can speak of "species-typical capabilities" like hearing and vision, and that phrasing sidesteps some of the important but separable controversies around how we build a world hostile to those with species-atypical endowments

* author mentions that editing out some species-atypical traits (eg deafness) will dwindle base of support for the already inadequate accommodations society makes for those traits. am thinking of this like ideal versus non-ideal theories of justice, where you must be mindful of a policy feedback loop.

* another example: we do not want to edit next generation to accommodate a racist society. the focus should be on making society not racist. it seems to me one way to fill in the argument here is to point out that "treatment" here depends on diagnosis: eg, on one leftwing view racialization is symptomatic of racism, like fever is of the flu.

* overly gene-centric view is generating a lot of the bad ideas about gene-editing. genes are not "causing" income inequality, political institutions are. good news: we don't need cutting edge science to do "policy-editing."

* parents are already wanting to bring all their negative-sum status competitions to the genetic arena: making next generation 6" taller, etc. this is both the most likely and dumbest problem in the medium term.

* if we cannot get universal access in place, cooperation/cohesion likely to break down, both within and between societies

* author doesn't discuss but what is "desirable" may itself be a function of genome, so we get a very unpredictable path dependence: first we edit this gene, then next generation consequently has new desire to edit another gene, and so on perhaps so gradually as to be imperceptible. ( )