Enhancing trust in AI through industry self-governance.

Artificial intelligence (AI) is critical to harnessing value from exponentially growing health and healthcare data. Expectations are high for AI solutions to effectively address current health challenges. However, there have been prior periods of enthusiasm for AI followed by periods of disillusionment, reduced investments, and progress, known as "AI Winters." We are now at risk of another AI Winter in health/healthcare due to increasing publicity of AI solutions that are not representing touted breakthroughs, and thereby decreasing trust of users in AI. In this article, we first highlight recently published literature on AI risks and mitigation strategies that would be relevant for groups considering designing, implementing, and promoting self-governance. We then describe a process for how a diverse group of stakeholders could develop and define standards for promoting trust, as well as AI risk-mitigating practices through greater industry self-governance. We also describe how adherence to such standards could be verified, specifically through certification/accreditation. Self-governance could be encouraged by governments to complement existing regulatory schema or legislative efforts to mitigate AI risks. Greater adoption of industry self-governance could fill a critical gap to construct a more comprehensive approach to the governance of AI solutions than US legislation/regulations currently encompass. In this more comprehensive approach, AI developers, AI users, and government/legislators all have critical roles to play to advance practices that maintain trust in AI and prevent another AI Winter.

Sensory determinants of behavioral dynamics in Drosophila thermotaxis.

Complex animal behaviors are built from dynamical relationships between sensory inputs, neuronal activity, and motor outputs in patterns with strategic value. Connecting these patterns illuminates how nervous systems compute behavior. Here, we study Drosophila larva navigation up temperature gradients toward preferred temperatures (positive thermotaxis). By tracking the movements of animals responding to fixed spatial temperature gradients or random temperature fluctuations, we calculate the sensitivity and dynamics of the conversion of thermosensory inputs into motor responses. We discover three thermosensory neurons in each dorsal organ ganglion (DOG) that are required for positive thermotaxis. Random optogenetic stimulation of the DOG thermosensory neurons evokes behavioral patterns that mimic the response to temperature variations. In vivo calcium and voltage imaging reveals that the DOG thermosensory neurons exhibit activity patterns with sensitivity and dynamics matched to the behavioral response. Temporal processing of temperature variations carried out by the DOG thermosensory neurons emerges in distinct motor responses during thermotaxis.

Sensorimotor structure of Drosophila larva phototaxis.

The avoidance of light by fly larvae is a classic paradigm for sensorimotor behavior. Here, we use behavioral assays and video microscopy to quantify the sensorimotor structure of phototaxis using the Drosophila larva. Larval locomotion is composed of sequences of runs (periods of forward movement) that are interrupted by abrupt turns, during which the larva pauses and sweeps its head back and forth, probing local light information to determine the direction of the successive run. All phototactic responses are mediated by the same set of sensorimotor transformations that require temporal processing of sensory inputs. Through functional imaging and genetic inactivation of specific neurons downstream of the sensory periphery, we have begun to map these sensorimotor circuits into the larval central brain. We find that specific sensorimotor pathways that govern distinct light-evoked responses begin to segregate at the first relay after the photosensory neurons.

Balance of activity between LN(v)s and glutamatergic dorsal clock neurons promotes robust circadian rhythms in Drosophila.

Circadian rhythms offer an excellent opportunity to dissect the neural circuits underlying innate behavior because the genes and neurons involved are relatively well understood. We first sought to understand how Drosophila clock neurons interact in the simple circuit that generates circadian rhythms in larval light avoidance. We used genetics to manipulate two groups of clock neurons, increasing or reducing excitability, stopping their molecular clocks, and blocking neurotransmitter release and reception. Our results revealed that lateral neurons (LN(v)s) promote and dorsal clock neurons (DN(1)s) inhibit light avoidance, these neurons probably signal at different times of day, and both signals are required for rhythmic behavior. We found that similar principles apply in the more complex adult circadian circuit that generates locomotor rhythms. Thus, the changing balance in activity between clock neurons with opposing behavioral effects generates robust circadian behavior and probably helps organisms transition between discrete behavioral states, such as sleep and wakefulness.

Controlling airborne cues to study small animal navigation.

Small animals such as nematodes and insects analyze airborne chemical cues to infer the direction of favorable and noxious locations. In these animals, the study of navigational behavior evoked by airborne cues has been limited by the difficulty of precisely controlling stimuli. We present a system that can be used to deliver gaseous stimuli in defined spatial and temporal patterns to freely moving small animals. We used this apparatus, in combination with machine-vision algorithms, to assess and quantify navigational decision making of Drosophila melanogaster larvae in response to ethyl acetate (a volatile attractant) and carbon dioxide (a gaseous repellant).

Impairment of ubiquitylation by mutation in Drosophila E1 promotes both cell-autonomous and non-cell-autonomous Ras-ERK activation in vivo.

Ras signaling can promote proliferation, cell survival and differentiation. Mutations in components of the Ras pathway are found in many solid tumors and are associated with developmental disorders. We demonstrate here that Drosophila tissues containing hypomorphic mutations in E1, the most upstream enzyme in the ubiquitin pathway, display cell-autonomous upregulation of Ras-ERK activity and Ras-dependent ectopic proliferation. Ubiquitylation is widely accepted to regulate receptor tyrosine kinase (RTK) endocytosis upstream of Ras. However, although the ectopic proliferation of E1 hypomorphs is dramatically suppressed by removing one copy of Ras, removal of the more upstream components Egfr, Grb2 or sos shows no suppression. Thus, decreased ubiquitylation may lead to growth-relevant Ras-ERK activation by failing to regulate a step downstream of RTK endocytosis. We further demonstrate that Drosophila Ras is ubiquitylated. Our findings suggest that Ras ubiquitylation restricts growth and proliferation in vivo. We also report our intriguing observation that complete inactivation of E1 causes non-autonomous activation of Ras-ERK in adjacent tissue, mimicking oncogenic Ras overexpression. We demonstrate that maintaining sufficient E1 function is required both cell autonomously and non-cell autonomously to prevent inappropriate Ras-ERK-dependent growth and proliferation in vivo and may implicate loss of Ras ubiquitylation in developmental disorders and cancer.

Rictor/TORC2 regulates fat metabolism, feeding, growth, and life span in Caenorhabditis elegans.

Rictor is a component of the target of rapamycin complex 2 (TORC2). While TORC2 has been implicated in insulin and other growth factor signaling pathways, the key inputs and outputs of this kinase complex remain unknown. We identified mutations in the Caenorhabditis elegans homolog of rictor in a forward genetic screen for increased body fat. Despite high body fat, rictor mutants are developmentally delayed, small in body size, lay an attenuated brood, and are short-lived, indicating that Rictor plays a critical role in appropriately partitioning calories between long-term energy stores and vital organismal processes. Rictor is also necessary to maintain normal feeding on nutrient-rich food sources. In contrast to wild-type animals, which grow more rapidly on nutrient-rich bacterial strains, rictor mutants display even slower growth, a further reduced body size, decreased energy expenditure, and a dramatically extended life span, apparently through inappropriate, decreased consumption of nutrient-rich food. Rictor acts directly in the intestine to regulate fat mass and whole-animal growth. Further, the high-fat phenotype of rictor mutants is genetically dependent on akt-1, akt-2, and serum and glucocorticoid-induced kinase-1 (sgk-1). Alternatively, the life span, growth, and reproductive phenotypes of rictor mutants are mediated predominantly by sgk-1. These data indicate that Rictor/TORC2 is a nutrient-sensitive complex with outputs to AKT and SGK to modulate the assessment of food quality and signal to fat metabolism, growth, feeding behavior, reproduction, and life span.