The work was done using a gene editing technique called CRISPR (pronounced "crisper").
The idea of gene editing is to make specific changes in a particular gene, just as you might correct a spelling mistake. Gene editing has been around for decades, but in organisms other than mice it used to be difficult, expensive and time-consuming.
The CRISPR method - the name refers to characteristic sets of repeating chunks of DNA known as "clustered regularly interspaced short palindromic repeats" - developed in just the past few years, has changed all that, allowing biologists to achieve in weeks what used to take years.
The ease, speed and cheapness of CRISPR has made it possible for more people to experiment with gene editing. Last month, it was reported that a handful of teams are trying to modify human embryos using the method. Now one of those teams, led by Junjiu Huang at the Sun Yat-sen University in Guangzhou, China, has published its results.
"Because ethical concerns preclude studies of gene editing in normal embryos," the team writes, the researchers used human eggs that had been fertilised by two sperm rather than one.
These "polyspermic" eggs may develop for a few days but never develop normally and are discarded by fertility clinics.
Of the 86 eggs injected, just four were successfully modified - an efficiency rate far lower than required to make human germline gene editing a practical prospect. The others either did not survive, or were not successfully modified.
There were also changes to genes other than the globin gene. Such "off-target" alterations are a big concern, because they could cause serious illnesses.
However, Huang and his colleagues found what could be a serious problem: the embryos were a mixture of modified and unmodified cells - so-called genetic mosaics. That means the results of preimplantation genetic testing could be misleading.
But should this kind of research be done at all? That depends on whether you think modifying the inheritable DNA of the human germline is acceptable. Some have called for a moratorium on this kind of work, and according to Huang, the paper was rejected by the journals Science and Nature in part because of ethical concerns.
Polls in various countries, however, indicate that there is actually substantial public support - sometimes over 50 per cent - for using germline modification.