News/Science
How AI Is Detecting Hidden Heart Attack Risk in CT Scans
New artificial intelligence technology has demonstrated a groundbreaking ability to identify individuals at risk of heart attacks within the next decade by analyzing CT scans for previously undetectable inflammation. This innovative approach could significantly enhance early intervention strategies for cardiovascular disease prevention. Technological breakthrough in cardiac risk assessment: The AI model developed by Oxford University spinout Caristo Diagnostics can detect heart inflammation invisible to the human eye, potentially revolutionizing preventive cardiology. The AI analyzes CT scans to identify subtle signs of inflammation in coronary arteries, which is strongly associated with future cardiac events. In a study of 40,000 patients, those...
read Aug 3, 2024Why Some Experts Believe AI Will Never Achieve Consciousness
The debate surrounding artificial intelligence (AI) and consciousness has long captured the imagination of scientists, philosophers, and science fiction enthusiasts alike. Recent discussions have reignited this conversation, exploring the fundamental differences between biological brains and artificial computing systems. The AI consciousness hypothesis: Some scientists and philosophers posit that artificial intelligence could potentially achieve consciousness, drawing parallels between the complexity of the human brain and advanced computing systems: This idea has been popularized in science fiction, such as Philip K. Dick's novel "Do Androids Dream of Electric Sheep?" and its film adaptation, "Blade Runner." Proponents argue that if the brain, as...
read Aug 1, 2024AI is Enabling Faster, More Accurate Results in Animal Health
AI is enabling faster and more accurate diagnoses for improved animal health outcomes. Pioneering AI use in veterinary medicine: The Royston Veterinary Centre in England has successfully employed AI to diagnose and treat several pets, showcasing the technology's potential to transform veterinary practices: AI analyzed samples from Blossom, a cat with aggressive cancer, and two dogs, Scooby Doo and Tara, with skin conditions, leading to accurate diagnoses and effective treatments. By analyzing prepared slides under a microscope, the AI flags areas of concern for surgeons to review, streamlining the diagnostic process. Speeding up diagnoses and expanding capabilities: AI implementation significantly...
read Aug 1, 2024Mayo Clinic Is Using AI for Early Dementia Detection
The Mayo Clinic is leveraging AI to enhance the detection of early signs of dementia in EEG tests, potentially changing how medical professionals approach the diagnosis and treatment of cognitive disorders. AI-enhanced EEG analysis: The Mayo Clinic Neurology AI Program (NAIP) has developed an AI-based method to analyze electroencephalogram (EEG) tests with increased speed and precision, enabling the detection of subtle abnormalities that may indicate cognitive issues: Researchers used machine learning to distill complex brain wave patterns into six specific features, allowing the AI to focus on indicators of cognitive problems such as Alzheimer's disease and Lewy body dementia. The...
read Aug 1, 2024Researchers Are Exploring How AI Could Help with End-of-Life Decisions
A new AI tool, which hasn't been built, but whose potential benefits are under careful consideration, aims to help with end-of-life decisions. The challenge of end-of-life decision-making: When patients are unable to make their own medical decisions, the responsibility falls on surrogates, often family members, which can be emotionally distressing and lead to disagreements: In the case of Sophie, a woman in her mid-50s who suffered a hemorrhagic stroke, her family members couldn't agree on whether to continue medical treatments or let her die peacefully. Surrogates often struggle with the burden of making life-or-death decisions on behalf of their loved...
read Jul 31, 2024Deep Learning Enables Eavesdropping on Digital Video Displays
Eavesdropping on digital video displays through electromagnetic emanations: Researchers have developed a deep learning-based system called DEEP-TEMPEST that can effectively eavesdrop on digital video displays, such as HDMI, by analyzing the unintentional electromagnetic waves emanating from cables and connectors: The digital case, particularly HDMI, poses a greater challenge compared to analog (VGA) due to a 10-bit encoding that results in a larger bandwidth and a non-linear mapping between the observed signal and pixel intensity. Existing eavesdropping systems designed for analog video obtain unclear and difficult-to-read images when applied to digital video, necessitating a new approach. Deep learning as a solution:...
read Jul 31, 2024A New AI Blood Test Predicts Male Infertility
The AI-powered blood test shows promise in predicting male infertility risk, potentially making initial screening more accessible without the need for specialized labs, though it is not a replacement for comprehensive semen analysis. Key Takeaways: Researchers have developed an AI system that can predict male infertility risk with 74% accuracy by measuring hormone levels in blood samples associated with sperm production. The AI model can also predict non-obstructive azoospermia, a severe form of infertility where sperm are absent from semen, with 100% accuracy. While the AI-enhanced blood test could make initial infertility screening more accessible in primary care settings, it...
read Jul 31, 2024Why We May Be Closer Than We Think to AI Reading Our Minds
The rapid advancement of AI and its potential to read our minds is becoming a reality, with significant implications for the future of imagination, privacy, and human-AI interactions. Decoding brain activity at MIT: Researchers are making progress in understanding the complex patterns of brain waves and their associations with various cognitive processes and neurological conditions: Scientists have identified distinct interactions between beta and gamma waves across the cortex that target the processing of specific stimuli, rather than being generic in nature. Synchronous brain waves have been linked to epilepsy, while impaired predictive coding in the brain has been associated with...
read Jul 26, 2024How Generative AI Is a Metaphor for the Human Genome
Recent research suggests that the human genome functions similarly to generative AI models, providing a compressed set of information that can be used to create new organisms with significant potential for variation. Key takeaways from the genome-AI comparison: The genome-as-generative-AI model offers a new perspective on how our genetic code works, moving beyond traditional metaphors like blueprints or recipes: Like generative AI models that create new images from compressed data, the genome contains compressed information from evolutionary "training data" that can be used to create new organisms. This model allows for substantial variation in the resulting organisms, just as generative...
read Jul 22, 2024AI-Powered Weather Model Promises Accurate, Efficient Forecasts and Climate Insights
Google researchers have developed NeuralGCM, a new weather prediction model that integrates AI with traditional physics-based methods, potentially providing accurate forecasts more efficiently and cost-effectively. Bridging the divide between AI and traditional weather modeling: NeuralGCM combines the strengths of machine learning and conventional weather prediction techniques: The model uses a conventional approach to calculate large-scale atmospheric changes and incorporates AI to correct errors that accumulate on smaller scales, such as cloud formations and regional microclimates. This selective injection of AI aims to enhance the accuracy of predictions while maintaining the reliability of traditional physics-based models. Potential for long-term climate modeling...
read Jul 22, 2024DOE Launches $12B AI Initiative for Science, Energy, and Security Breakthroughs
The Department of Energy (DOE) has unveiled a $12 billion, five-year road map to develop integrated scientific AI systems, aiming to advance scientific discovery, energy solutions, and national security applications. Key focus areas of the FASST initiative: The Frontiers in Artificial Intelligence for Science, Security and Technology (FASST) project will establish a network of AI research clusters utilizing DOE's computing infrastructure, focusing on four main pillars: Data: Developing methods, platform protocols, and tools for efficient and safe aggregation and distribution of AI training datasets. Computing Infrastructure and Platforms: Developing and deploying next-generation AI computing platforms and infrastructure such as testbeds...
read Jul 21, 2024AI May Enable Interstellar Communication — So What Should We Tell the Aliens?
Artificial intelligence may enable real-time communication with extraterrestrial civilizations, prompting us to consider what we should tell them about humanity. The power of large language models: Recent advancements in AI, particularly the development of transformer neural network architecture and large language models (LLMs), have the potential to revolutionize communication with extraterrestrial intelligence: LLMs, trained on vast datasets, contain a wealth of human knowledge and are already poised to significantly impact global employment, according to the IMF. By transmitting a well-curated LLM to extraterrestrial civilizations, we could enable them to indirectly converse with us and learn about humanity without being hindered...
read Jul 18, 2024A Short History of AI, As Told By MIT
Despite its widespread use, there is little consensus on what artificial intelligence actually is, with opinions ranging from it being akin to magic to simply complex mathematics. Historical context and competing definitions: The very term "artificial intelligence" was controversial from the start, with many of John McCarthy's colleagues in 1955 preferring alternative names like "automata studies" or "complex information processing": AI has since attracted zealous fandoms and detractors, with camps often talking past each other and engaging in heated debates with significant real-world consequences. How AI is defined and explained by industry leaders and scientists to policymakers and the public...
read Jul 18, 2024Neural Network Learns Spatial Mapping from Visual Experience, Mirroring Brain’s Cognitive Maps
The predictive coding neural network constructs an implicit spatial map of its environment by assembling information from local exploration into a global representation within its latent space. Key takeaways: The network, trained on a next-image prediction task while navigating a virtual environment, automatically learns an internal map that quantitatively reflects spatial distances: The network's latent space encodes accurate spatial positions, enabling the agent to pinpoint its location using only visual information. Distances between image representations in latent space correlate strongly with actual physical distances in the environment. Individual latent space units exhibit localized, overlapping receptive fields akin to place cells...
read Jul 12, 2024A Startup Spun out of Meta Has a Massive AI Model that Speaks the Language of Proteins
A startup spun out of Meta has unveiled a massive AI model that speaks the language of proteins, creating new fluorescent molecules in an impressive proof-of-principle demonstration. EvolutionaryScale debuts protein language model ESM3: EvolutionaryScale, launched by former Meta scientists, announced its new protein language model ESM3 this month alongside $142 million in new funding to apply the model to drug development, sustainability, and other areas: ESM3 was trained on over 2.7 billion protein sequences and structures, as well as data on protein functions, allowing it to design proteins to user specifications. The model is seen as a frontier in the...
read Jul 11, 2024Researchers are Using AI to Decipher Sperm Whale ‘Language’
Researchers have uncovered complex structures in sperm whale communication that bear similarities to those found in human language, potentially revolutionizing our understanding of these enigmatic ocean giants. AI reveals sperm whale "phonetic alphabet": Using artificial intelligence to analyze thousands of sperm whale recordings, researchers identified 156 distinct codas (rhythmic click sequences) and their basic building blocks, akin to phonemes in human language: The AI detected subtle variations in coda speed, rhythm, and ornamentation, suggesting a richer information-carrying capacity than previously thought. The sperm whale "phonemes" can be combined to create a vast repertoire of vocalizations, hinting at the possibility of...
read Jul 11, 2024How AI Is Being Used to Change Scientific Research
Dr. Bradley Love, a professor of cognitive and decision sciences at University College London, is using AI to fundamentally change the way scientific research is conducted, focusing on predictions and cross-domain pattern recognition to overcome the limitations of siloed, human-driven research. Building BrainGPT to predict the future of neuroscience: Dr. Love and his team have developed BrainGPT, an AI language model that assists in neuroscientific research by processing vast amounts of data across different domains to find patterns and make predictions about new situations: BrainGPT aims to create a "collective mind" by treating individual research papers as incomplete contributions to...
read Jul 10, 2024OpenAI and Los Alamos Pioneer Real-World AI Lab Experiment
OpenAI and Los Alamos National Laboratory are collaborating to study the potential benefits and risks of using generative AI in a real-world laboratory setting, marking an important step in understanding the technology's impact on scientific research. Pioneering experiment in an active lab: OpenAI and Los Alamos are conducting what they believe to be the first experiment of its kind, using AI to assist in performing basic biomedical tasks within a functioning laboratory: The initial experiment involves using AI to help someone without specialized molecular biology skills genetically engineer E. coli bacteria to produce insulin. This real-world setting will provide valuable...
read Jul 3, 2024AI Accelerates Disease Diagnosis: Earlier Detection, Novel Biomarkers, and Personalized Insights
Artificial intelligence is revolutionizing disease diagnosis by enabling earlier detection and uncovering novel disease markers, but challenges remain in ensuring accuracy, reducing bias, and establishing appropriate guardrails. AI's diagnostic potential: AI algorithms are sifting through vast amounts of data to identify disease patterns and biomarkers that traditional tests might miss, enabling earlier and more personalized diagnoses: Researchers found that facial temperature patterns detected by thermal cameras and AI are associated with various chronic illnesses like diabetes and high blood pressure. An AI-assisted study identified a high-risk subtype of endometrial cancer that would have gone unrecognized by traditional diagnostic methods. A...
read Jul 3, 2024AI Models Used to Analyze Medical Images May Be Biased, Study Finds
A new study has revealed that AI models used to analyze medical images can exhibit biases in their diagnostic performance across different demographic groups, with the most accurate models showing the largest fairness gaps. The key findings and implications are: AI models' surprising capabilities linked to fairness gaps: The AI models that were most accurate at predicting patient demographics like race and gender from chest X-rays also had the biggest discrepancies in diagnostic accuracy between groups, suggesting the models may be relying on "demographic shortcuts" that lead to incorrect results for women, Black patients, and other groups. MIT researchers previously...
read Jul 2, 2024AI Breakthrough: One-Size-Fits-All Exoskeleton Improves Walking and Running
Researchers at North Carolina University's Lab of Biomechatronics and Intelligent Robotics have developed an exoskeleton that adapts to new users without the need for individual adjustments, thanks to AI-generated control policies. The AI was trained using digital models of the human musculoskeletal system and an exoskeleton robot, eliminating the need for lengthy human trials for each user. This breakthrough could significantly reduce the cost of exoskeletons from around $200,000 to between $2,000 and $5,000 within the next few years. Record-breaking performance: Exoskeleton outperforms competitors in walking, running, and stair climbing; The AI-powered exoskeleton achieved record-breaking metabolic rate reductions in various...
read Jul 1, 2024AI and Genomics Breakthroughs Amid Geopolitical Tensions and Energy Challenges
Gloomy data centers, sunny solar power: The rapid growth of AI could strain US power grids, with data centers potentially consuming 8% of total electricity by 2030. However, this gloomy projection may be overstated, as renewable energy is rapidly expanding, with California already generating over 100% of its electricity from renewables at times. The data center boom and its energy demands will likely be a temporary issue, as energy mix, efficiency improvements, and evolving economics of both data centers and energy production are not fully accounted for in linear projections. Ultimately, the computational power of these data centers will be...
read Jun 27, 2024How AI Helped Solve One Of Biology’s Biggest Problems, And Why It May Just Be the Start
AlphaFold's breakthrough: A new era for protein science begins In December 2020, John Jumper from Google DeepMind presented AlphaFold2, an AI tool that could predict protein structures with over 90% accuracy - solving a 50-year-old challenge known as the protein folding problem and marking the biggest AI breakthrough in science to date. The protein folding problem takes center stage: For decades, scientists sought to crack the code of how a protein's amino acid sequence determines its folded 3D shape, which in turn defines its biological function: In the 1950s, Christian Anfinsen hypothesized that a protein's amino acid sequence contains the...
read Jun 27, 2024AI’s Growing Role in Science Raises Questions About Academia’s Future
The rise of AI-assisted scientific research raises important questions about the future of academia and knowledge creation: Key Takeaways: Large language models (LLMs) are now advanced enough to help write scientific papers, potentially speeding up the drafting process, especially for non-native English speakers. However, LLMs also come with risks, such as reproducing biases and generating plausible-sounding nonsense, which could compromise the integrity of scientific literature if not properly addressed. Prevalence of AI-assisted research: A recent analysis suggests that AI may have already contributed to at least 10% of published scientific papers, highlighting the rapid adoption of these tools in academia:...
read