Human oversight of erroneous clinical AI and decision support a systematic review and meta analysis - Nature
This story deserves to be shared
Send it to someone who'd care, or post it on your networks.
We’re sharing this article early to provide faster access to peer-reviewed, accepted research. It is citable and carries a permanent DOI. This version is subject to further edits and will be replaced automatically by the final Version of Record. All legal disclaimers apply.
AbstractArtificial intelligence (AI) increasingly supports clinical decisions, but human oversight is protective only if clinicians recognize and resist erroneous output. We systematically reviewed controlled and real-world studies of clinician responses to incorrect automated advice. Forty-six reports represented 41 independent cohort clusters, including 40 core cohorts. Across three adjusted cohorts, clinician correctness was substantially higher with correct than incorrect automated advice (pooled OR 10.19, 95% CI 7.43–13.98; I² = 0%); a ≥ 2020 sensitivity analysis was similar but imprecise (OR 9.22, 95% CI 2.06–41.13). In four heterogeneous real-world computerized ECG studies, 26.6% of erroneous interpretations were not corrected or recognized (95% CI 6.4–65.6%; I² = 95.8%). Contemporary AI and large language model studies also documented harmful reliance but were too heterogeneous for pooling.
This article is an excerpt. To read the full content, visit the original source. Nature
Related articles
PHOTOS: Rare scientific artifacts on display at University of Calgary - Calgary Journal
Julia Margaret Cameron, photographic portrait of Charles Darwin, 1868. From the D. Hall Darwin-Wallace Collection. Rare books, documents, letters and images from famous evolutionary biologists are on display this fall at the U of C’s Nickle Galleries.
MEGATRON simulations connect the first stars with the universe's smallest galaxies - The Brighter Side of News
MEGATRON simulations offer a possible explanation for the similar iron content of many of the smallest galaxies. Radiation delayed early star formation until galaxies could retain elements from powerful stellar explosions. The results connect ancient stars with early-universe observations, but depend on uncertain assumptions about the first stars. Some of the universe’s smallest galaxies contain a surprisingly similar proportion of iron, despite holding different amounts of stars. Computer simulations now offer an explanation: a single explosion from a very massive first-generation star could have supplied much of that iron. The finding gives astronomers a testable way to investigate stars that disappeared billions of years ago. One paper follows the chemical legacy of the first stars into hundreds of simulated dwarf galaxies. Its central finding concerns an iron pattern already measured in nearby galaxies, rather than a direct observation of the earliest stellar explosions. Other MEGATRON studies investigate galaxy spectra and the gas within and around galaxies.
Single X-ray pulse reveals plasma nanostructures across scales with one camera - Phys.org
Schematic of an observational experiment using an X-ray free-electron laser to study the plasma formation dynamics of a microscale rod assembly target under a high-intensity femtosecond laser. Credit: National Institute for Fusion Science Understanding instabilities and turbulence in plasmas is one of the key challenges across a wide range of plasma research fields, including nuclear fusion, astrophysics and laser-plasma physics. In plasmas, slight disturbances can grow due to instabilities, and energy is transferred between structures of various sizes due to turbulence. These phenomena give rise to structures ranging from very small to large, which significantly influence the overall motion and energy flow of the plasma.
From the Rewrite Desk: Neanderthal Man Was Not a Stupid Brute - Walter Bradley Center for Natural and Artificial Intelligence
An ideologue should hate paleontology. It is always turning up something new. And sometimes the new stuff is really inconvenient. At one time, Neanderthal man was our dumb cousin. His stupidity caused him to lose the Darwinian battle for survival. National Geographic had the story back then.
0 comments
No comments yet. Be the first to share your thoughts!