The Moral Imperative Has a Read Replica

The announcement reads as a founder taking a moral stand. Cathy Tie of Origin Genomics says that editing the genes of human embryos is a “moral imperative,” because hereditary disease is cruel and we now have tools to stop it. One notices that most families carrying a hereditary disease already have a way to avoid passing it on, and it involves no editing at all. With that detail load-bearing, the announcement reads differently.

The existing method is preimplantation genetic testing. An IVF clinic produces several embryos, tests them, and transfers one that did not inherit the faulty variant. Put in systems terms, it is a read operation. Nothing in the genome is changed; the clinic looks and then chooses. Embryo editing is a write operation, and the thing being written to is the germline, which is the one database in biology that replicates to every descendant without a rollback path. Any competent architect prefers a read when a read will do. That preference is ordinary engineering discipline, and nobody should mistake it for timidity.

So the question is when a read will not do. The honest answer is a narrow set of cases. If both parents have the same recessive condition, every embryo inherits it and there is nothing healthy to select. If one parent carries two copies of a dominant variant, the same thing happens. Those families exist, and their situation is genuinely hard. They are also rare. How rare is a figure that does not appear in anything Origin Genomics has said, and it is the most important number for the whole argument.

The strongest version of Tie’s case deserves a fair hearing. Selection throws embryos away, and for some parents that is a moral cost in itself. IVF often yields only a few viable embryos, sometimes none unaffected. Editing, if it worked cleanly, would rescue the embryo in the dish rather than choosing among them. These are serious points. None of them turns a treatment for rare cases into an imperative for everyone. Each is a reason to fix the small set of cases where selection fails, and each depends on “if it worked cleanly.” That condition is the whole open problem, since off-target edits and mosaic embryos are exactly the failure modes the field has not yet closed.

Here the incentive becomes visible. A company cannot survive on the rare family with two affected parents. A company needs volume. Once the narrow medical case has been used to establish that editing is a moral imperative, the imperative has to cover the common case, where selection already works, because that is where the customers are. The Dutch put it plainly: wie betaalt, bepaalt, meaning whoever pays decides. An imperative with a sales pipeline attached has a way of widening until it matches the pipeline.

The field already has an incident report. In November 2018, He Jiankui announced the birth of twin girls whose CCR5 gene he had edited to make them resistant to HIV. Their father was HIV-positive, and sperm washing, a routine read-level technique, already prevented transmission to the children. The write operation solved a problem the girls did not have and introduced risks nobody could measure. A Chinese court sentenced He to three years in prison. The lesson of that case is not that editing is evil. The lesson is that the first people to reach for the write operation reached for it where a read had already succeeded, because a write makes a better announcement.

So the plain question, for Origin Genomics and for anyone repeating its framing: among the families you intend to serve, how many could not be served by testing and selection? If the answer is small, then the imperative is real for those few families, and the product is something else for everyone else. If the answer is large, the company should publish the number, because every clinician who has run a preimplantation testing programme will want to see it. And if no answer is offered, the silence tells you who is expected to ask and who is hoping they won’t.

A short transmission note, for anyone reading future announcements of this kind: when a moral claim arrives attached to a product, look for the cheaper method it would replace. If that method exists, the claim is really about market share.

None of this is contempt for the people involved. Founders believe their own pitch decks, which is how the decks get written. Somewhere, in a clinic on an ordinary Tuesday, an embryologist bends over a microscope, reads a result off the screen, marks one embryo as unaffected, and moves it gently to the transfer dish. No press release goes out.