Towards intellectual freedom in an AI Ethics Global Community.

The recent incidents involving Dr. Timnit Gebru, Dr. Margaret Mitchell, and Google have triggered an important discussion emblematic of issues arising from the practice of AI Ethics research. We offer this paper and its bibliography as a resource to the global community of AI Ethics Researchers who argue for the protection and freedom of this research community. Corporate, as well as academic research settings, involve responsibility, duties, dissent, and conflicts of interest. This article is meant to provide a reference point at the beginning of this decade regarding matters of consensus and disagreement on how to enact AI Ethics for the good of our institutions, society, and individuals. We have herein identified issues that arise at the intersection of information technology, socially encoded behaviors, and biases, and individual researchers' work and responsibilities. We revisit some of the most pressing problems with AI decision-making and examine the difficult relationships between corporate interests and the early years of AI Ethics research. We propose several possible actions we can take collectively to support researchers throughout the field of AI Ethics, especially those from marginalized groups who may experience even more barriers in speaking out and having their research amplified. We promote the global community of AI Ethics researchers and the evolution of standards accepted in our profession guiding a technological future that makes life better for all.

Maternal education and newborn thyroid-stimulating hormone level in a congenital hypothyroidism screening program.

Objective: To examine the relationship between neonatal thyroid function and the formal education of mothers.Study design: Participants came from a population-based congenital hypothyroidism (CH) screening program in Tianjin, China.Methods: Of 66,390 registered births in 2015, 60,568 mothers and newborns had complete data. Mothers were categorized into one of four categories based on their educational attainment: (a) midschool or less; (b) high school or equivalent; (c) university; or (d) post graduate. Newborn thyroid-stimulating hormone (TSH) level was measured on day 3-7. Two neonatal groups were created using cutoffs of TSH > 10 µIU/ml and TSH > 20 µIU/ml. Odds ratios (OR) for CH risk by maternal education were estimated from logistic regression models after adjusting for potential confounders.Results: For TSH > 10 µIU/ml, the screen positive incidence rate for CH was 1:201 or 4.98 per 1000 births; for TSH > 20 µIU/ml, the incidence rate was 1:2222 or 0.45 per 1000 births. Screen positive incidence rates decreased with increasing maternal education level. Compared to mothers with a postgraduate education, the ORs (95% CI) for midschool or less, high school or equivalent, and university were 2.09 (1.08, 4.04), 1.45 (0.73, 2.90), and 1.61 (0.85, 3.06) using a cutoff of TSH > 10 µIU/ml. At the higher cutoff of TSH > 20 µIU/ml ORs (95% CI) for midschool or less and high school or equivalent were 3.05 (1.20,7.74) and 3.34 (1.24, 8.97), when compared to a composite reference category of university and postgraduate level education.Conclusion: Maternal education is inversely related to neonatal thyroid function though the mechanism remains unexplained.

Non-centrosomal microtubule-organising centres in cold-treated cultured Drosophila cells.

In this paper we describe a new type of non-centrosomal microtubule-organising centre (MTOC), which is induced by cold treatment of certain cultured Drosophila cells and allows rapid reassembly of microtubule (MT) arrays. Prolonged cooling of two types of cultured Drosophila cells, muscle cells in primary culture and a wing imaginal disc cell line Cl.8+ results in disassembly of MT arrays and induces the formation of clusters of short MTs that have not been described before. Upon rewarming, the clusters are lost and the MT array is re-established within 1 h. In Cl.8+ cells, gamma-tubulin-containing centrosomes are detected, both in cell extensions and in the expected juxtanuclear position, and gamma-tubulin co-localises with the cold-induced MT clusters. The MT plus-end-binding protein, Drosophila EB1, decorates growing tips of MTs extending from clusters. We conclude that the cold-induced MT clusters represent acentrosomal MTOCs, allowing rapid reassembly of MT arrays following exposure to cold.

Positioning and capture of cell surface-associated microtubules in epithelial tendon cells that differentiate in primary embryonic Drosophila cell cultures.

Using primary embryonic Drosophila cell cultures, we have investigated the assembly of transcellular microtubule bundles in epidermal tendon cells. Muscles attach to the tendon cells of previously undescribed epidermal balls that form shortly after culture initiation. Basal capture of microtubule ends in cultured tendon cells is confined to discrete sites that occupy a relatively small proportion of the basal cell surface. These capturing sites are associated with hemiadherens junctions that link the ends of muscle cells to tendon cell bases. In vivo, muscle attachment and microtubule capture occur across the entire cell base. The cultured tendon cells reveal that the basal ends of their microtubules can be precisely targeted to small, pre-existing, structurally well-defined cortical capturing sites. However, a search and capture targeting procedure, such as that undertaken by kinetochore microtubules, cannot fully account for the precision of microtubule capture and positioning in tendon cells. We propose that cross-linkage of microtubules is also required to zip them into apicobasally oriented alignment, progressing from captured basal plus ends to apical minus ends. This involves repositioning of apical minus ends before they become anchored to an apical set of hemiadherens junctions. The proposal is consistent with our finding that hemiadherens junctions assemble at tendon cell bases before they do so at cell apices in both cultures and embryos. It is argued that control of microtubule positioning in the challenging spatial situations found in vitro involves the same procedures as those that operate in vivo.