A European fertility organization is calling for international limits on how many children can be conceived from a single sperm donor, according to MIT Technology Review’s Jessica Hamzelou. The proposal targets a problem that has become harder to ignore as DNA testing and online registries expose sprawling groups of donor-conceived half-siblings.
Hamzelou reported the case of Ties van der Meer, a 47-year-old conceived at a private fertility clinic with sperm from an anonymous donor. Van der Meer has found one sibling, but he does not know how many more may exist. Other donor-conceived people have discovered dozens, and in some cases hundreds, of half-siblings. One person who found 25 half-siblings described the experience to MIT Technology Review as feeling mass-produced.
The fertility group’s argument is straightforward: a donor’s genetic contribution should not be allowed to scale without a cross-border ceiling. National rules can be weak medicine when fertility treatment, donor sperm and family searches do not stop at national borders. The proposal, as described by MIT Technology Review, is meant to reduce the chance that a single donor produces very large numbers of children, though the report also notes that such a policy has limits.
AI researchers look beyond text
MIT Technology Review is also turning to a separate AI question: how machines can build useful models of the physical world. Large language models have made software much better at handling text, but robotics and other embodied systems need more than fluent language. They need ways to represent objects, motion, cause and effect, and the messy geometry of real spaces.
Researchers describe one answer as world models, a class of AI systems intended to help machines understand and operate in physical environments. MIT Technology Review said it will discuss the technology in a LinkedIn Live event on Tuesday, July 14, with Will Douglas Heaven, its senior editor for AI, and Sam Sinha, founding AI researcher and head of world models at 1X Technologies. The session is scheduled for 9:30 a.m. PDT, 12:30 p.m. EDT and 5:30 p.m. BST.
Other technology developments
- Apple has sued OpenAI, alleging the company stole trade secrets to build consumer hardware, according to CNBC. The BBC reported that the suit claims OpenAI hired Apple staff to obtain information, while Reuters reported that Apple also sued former employees Chang Liu and Tang Tan.
- Omar Yaghi, a Nobel Prize-winning chemist known for creating “molecular sponges,” is leaving the United States to lead a China-based AI institute focused on discovering new materials, according to the Los Angeles Times and The New York Times.
- The European Union is moving toward tighter social media restrictions for children, including a proposal to bar under-13s without adult supervision, according to The New York Times. Politico reported that EU officials also told Meta to disable autoplay and infinite scroll.
- Meta removed an Instagram AI image feature after criticism, TechCrunch reported. The New York Times said the tool could generate images based on public accounts and automatically included users with public profiles.
- Bloomberg reported that Phia, a shopping app co-founded by Phoebe Gates, claimed affiliate sales through fake clicks that it had not actually driven.
Wired also reported that San Francisco police drone footage was accidentally released, exposing hours of surveillance video. NPR reported that AI platforms are training political bots to imitate candidates in campaign texts, which sounds about as pleasant as campaign texting already is.
This story draws on original reporting from MIT Technology Review.
