Phys.org March 8, 2024 Scientists are enthusiastically imagining ways in which artificial intelligence (AI) tools might improve research. A team of researchers in the US (Yale University, Princeton University) developed a taxonomy of scientists’ visions for AI, observing that their appeal comes from promises to improve productivity and objectivity by overcoming human shortcomings. But proposed AI solutions can also exploit our cognitive limitations, making us vulnerable to illusions of understanding in which we believe we understand more about the world than we actually do. Such illusions obscure the scientific community’s ability to see the formation of scientific monocultures, in which […]
Tag Archives: Artificial intelligence
AI models struggle to identify nonsense, says study
Phys.org September 14, 2023 Neural network language models appear to be increasingly aligned with how humans process and generate language, but identifying their weaknesses through adversarial examples is challenging due to the discrete nature of language and the complexity of human language perception. An international team of researchers (USA – Columbia University, Israel) turned the models against each other by generating controversial sentence pairs where two language models disagreed about which sentence is more likely to occur. Considering nine language models (including n-gram, recurrent neural networks, and transformers), they created hundreds of controversial sentence pairs through synthetic optimization or by […]
2021’s Top Stories About AI Spoiler: A lot of them talked about what’s wrong with machine learning today
IEEE Spectrum December 27, 2021 Many of this year’s top articles grappled with the limits of deep learning and spotlighted researchers seeking new paths. Here are the 10 most popular AI articles that Spectrum published in 2021. Several came from Spectrum’s October 2021 special issue on AI, The Great AI Reckoning … read more.
AI system identifies buildings damaged by wildfire
Phys.org September 16, 2021 Existing technologies lack accuracy and ability to scale to effectively aid disaster relief and recovery. Even today, most wildfire event inspectors visit sites and manually classify building damage using before and after images of the buildings. A team of researchers in the US (Stanford University, California Polytechnic State University) has developed DamageMap, an artificial intelligence-powered post-wildfire building damage classifier. It is a binary classifier. Unlike existing solutions DamageMap relies on post-wildfire images alone by separating the segmentation and classification tasks. The model has an overall accuracy of 98% on the validation set (five wildfire events all […]
AI taught to rapidly assess disaster damage so humans know where help is needed most
Asia Research October 1, 2020 Using convolutional neural network (CNN) a team of researchers in Japan trained an AI using post-disaster aerial images to accurately determine how battered the buildings are. It works by classifying buildings as collapsed, non-collapsed, or blue tarp-covered based on the seven damage scales (D0-D6) used in the 2016 Kumamoto earthquakes. Based on the photos used to train the AI, they found that the blue tarp-covered category predominantly represented D2-D3 levels of devastation. When the system was tested on post-disaster aerial images of the September 2019 typhoon that hit Chiba, results showed that damage levels of […]
Landmark recommendations on development of artificial intelligence and the future of global health
Science Daily May 19, 2020 In this review article, a team of researchers in the US (Columbia University, Johns Hopkins University) suggests that AI-driven health interventions fit into four categories relevant to global health researchers: (1) diagnosis, (2) patient morbidity or mortality risk assessment, (3) disease outbreak prediction and surveillance, and (4) health policy and planning. However, much of the AI-driven intervention research in global health does not describe ethical, regulatory, or practical considerations required for widespread use or deployment at scale. Despite the field remaining nascent, AI-driven health interventions could lead to improved health outcomes in Low and Middle […]
Global AI Agenda
MIT Technology Insights April 22, 2020 This report is part of “The global AI agenda,” a thought leadership program by MIT Technology Review Insights examining how organizations are using AI today and planning to do so in the future. Featuring a global survey of 1,004 AI experts. From China to Japan, Singapore to India, policymakers across Asia have developed national-level plans for how AI can be used to enhance domestic and regional competitiveness, which include public and private sector collaboration. The results of the Asia-Pacific survey underscore the readiness of bringing AI to the fore when it comes down to driving […]
AI learns to design
Science Daily November 6, 2019 A team of researchers in the US (Carnegie Mellon University, Pennsylvania State University) has implemented a two-step framework that learns to imitate human design strategies from observation to generate designs without any explicit information about objective and performance metrics. It is trained to imitate a set of human designers by observing their design state sequences without inducing problem-specific modeling bias or extra information about the problem. It is designed to interact with the problem through a visual interface as humans did when solving the problem. The designs generated by a computational team of these agents […]
Is AI the next big climate-change threat? We haven’t a clue
MIT Technology Review July 29, 2019 Some predict that in the absence of significant innovation in materials, chip manufacturing and design, data centers’ AI workloads could account for a tenth of the world’s electricity usage by 2025 while others expect data center energy consumption to remain relatively flat over the next few years, in spite of a spike in AI-related activity. These widely diverging predictions highlight the uncertainty around AI’s impact on the future of large-scale computing and the ultimate implications for energy demand. But pessimistic forecasts ignore several important developments that could limit AI’s power grab. One of them […]
With Squad X, Dismounted Units Partner with AI to Dominate Battlespace
DARPA News July 7, 2019 DARPA’s Squad X Experimentation program aims to demonstrate a warfighting force with artificial intelligence as a true partner. In a recent field test, the program worked with U.S. Marines at the Air Ground Combat Center in Twentynine Palms, California, to track progress on two complementary systems that allow infantry squads to collaborate with AI and autonomous systems to make better decisions in complex, time-critical combat situations. With the conclusion of third experiment, the CACI system is moving into Phase 2, which includes an updated system that can remain continuously operational for five or more hours. […]