Articles tagged with Generative Ai
Large language models (LLMs) can alter sentiment in research summaries, making it difficult to accurately assess public opinion on climate change. LLMs tend to display a more neutral tone than original texts, regardless of prompts or model sophistication.
The new AI system, LLM4SD, can carry out basic steps of scientific research and provide insights to explain its results. It was tested with 58 research tasks across four domains, outperforming state-of-the-art tools and boosting accuracy by up to 48%.
A recent study from the University of Kansas found that people tend to prefer AI chatbots when discussing sensitive or embarrassing topics like health information, due to their anonymity and nonjudgmental nature. However, when individuals are angry, they still prefer human interaction.
A new study reveals that GPT models perform well on some analogy tasks but struggle with variations, highlighting key weaknesses in AI's reasoning capabilities. Analogical reasoning is a crucial aspect of human cognition and decision-making, yet AI models often rely on superficial patterns rather than deep comprehension.
A groundbreaking study using AI-driven robotics laboratory identifies Rentosertib, a potent TNIK inhibitor, as a highly effective senomorphic agent capable of reducing aging-related markers. The study reveals TNIK inhibition alleviates key signaling pathways implicated in senescence and fibrosis.
The collaboration aims to accelerate antibody discovery and development using Harbour BioMed's technology platform and Insilico's AI expertise. The companies will develop next-generation AI-powered antibodies for immunology, oncology, and neuroscience, aiming to deliver innovative therapies with enhanced specificity, efficacy, and safety.
A report by University of Cambridge experts highlights the risks of unregulated AI use on UK's thriving creative sector. The researchers argue that forcing creatives to 'opt out' of AI training could compromise their livelihoods, particularly those just starting out.
Researchers at Carnegie Mellon University found large differences in grammatical, lexical, and stylistic features between text written by humans and AI-generated LLMs. The study highlights the limitations of LLMs in mimicking human writing styles and their potential impact on education.
Researchers at Mass General Brigham used a generative AI tool to screen patients for a heart failure clinical trial, significantly improving speed and efficiency compared to manual screening. The study found that the AI-assisted screening process screened 458 eligible patients compared to 284 in the manual group.
A new study from the University of South Australia found that most people trust AI in situations where the stakes are low, such as music suggestions. However, those with poor statistical literacy or little familiarity with AI were just as likely to trust algorithms for trivial choices as they were for critical decisions. The study also...
A new industry report highlights the potential risks of generative artificial intelligence in journalism, including misleadings content and bias, while also identifying opportunities for AI-assisted tasks. News audiences are more comfortable with AI when they use it themselves, but concern about job displacement remains.
Researchers at Penn State developed a new method to analyze bird songs, revealing context dependence and underlying neurobiology. The study suggests that auditory feedback plays a key role in creating context dependence, and may hold insights into the neural mechanisms of human language.
A study published in PLOS Mental Health found that ChatGPT's responses were generally rated higher than those written by therapists, particularly in terms of core psychotherapy guiding principles. The AI model was also found to contextualize more extensively, leading respondents to rate its responses higher on common therapy components.
A significant gender gap was found in the use of AI for research, with male researchers experiencing a notable increase in productivity after ChatGPT's release. Female researchers spent less time using AI and perceived it as less efficient, exacerbating existing productivity disparities.
A Stanford Medicine study found that AI-powered chatbots can effectively answer nuanced clinical questions and improve physician performance. The research showed that chatbots outperformed doctors in certain scenarios, particularly when combined with human expertise, suggesting a potential future direction for AI-assisted medical decis...
Insilico Medicine has developed a highly selective FGFR2/3 dual inhibitor, which maintains efficacy against resistance mutations and demonstrates a more favorable safety profile compared to existing FGFR inhibitors. The compound showed robust antitumor efficacy in gastric cancer mouse models.
The study compared the performance of seven AI models with that of 400 humans in comprehension tasks, revealing a significant difference in accuracy. Human accuracy reached 89%, while AI models struggled to achieve more than 70% correct answers.
A survey of over 23,000 higher education students worldwide reveals mixed perceptions of ChatGPT's benefits and limitations. While students find it valuable for brainstorming and academic writing, they express concerns about its reliability, impact on critical thinking, and ethical issues.
A new study found that AI assistance increases worker productivity by 15% in customer service sectors, with significant improvements in speed and quality for less-experienced employees. However, its impact is minimal on higher-skilled workers, while adherence to AI recommendations leads to larger gains in productivity.
A study by the University of East Anglia found that generative AI tools like ChatGPT exhibit biases leaning towards left-wing political values. This can distort public discourse and exacerbate societal divides, highlighting the need for transparency and regulatory safeguards to ensure alignment with democratic values.
Researchers use generative AI to predict chromatin structures in single cells, overcoming limitations of existing experimental methods. The technique can generate thousands of structure predictions in minutes, enabling faster study of how 3D genome organization affects gene expression.
A large-scale study analyzing over three million job postings finds that Generative AI tools like ChatGPT are accelerating the transformation of the job market. While demand for partly substitutable skills reduces, new jobs are created in areas such as chatbot development and machine learning.
Primary care clinicians spend over half their workdays on EHR tasks, and AI can target areas like documentation, chart reviews, and medication management. Implementing thoughtful AI solutions prioritizes clinician well-being and patient care.
A cognitive science professor at UC Merced proposes a suite of video games to accelerate cancer research using artificial intelligence and gamified citizen science. The 'Cancer Wars' initiative aims to harness the collective problem-solving power of millions of gamers to tackle complex scientific problems.
A new study found that AI-generated empathetic responses were preferred over those from humans and expert crisis responders. The researchers suggest that AI can supplement human empathy, but should not replace it entirely due to potential biases and ethical concerns.
West Virginia University researchers Erin Brock Carlson and Scott Davidson have designed an interdisciplinary program to engage liberal arts faculty in discussions on the social, ethical and technical aspects of artificial intelligence. The program aims to support the creation or redesign of 10 courses incorporating AI, exploring its i...
Insilico Medicine has successfully dosed the first patient in its global Phase I clinical trial of ISM6331, a potent pan-TEAD inhibitor. The treatment demonstrates superior efficacy and safety in preclinical studies, with potential synergistic anti-tumor effects and overcoming drug resistance.
A new study assesses the historical knowledge of AI chatbots like ChatGPT-4 and finds they struggle with nuanced, PhD-level inquiry. The models performed best on legal systems and social complexity but struggled with topics such as discrimination and social mobility.
Anil Jain and Michael I. Jordan received the BBVA Foundation Award in Information and Communication Technologies for their pioneering work on machine learning, enabling transformative technologies like biometrics and artificial intelligence. Their research has unlocked applications of far-reaching impact on society.
The open-access book delves into the impact of AI and autonomous systems on human lives and ways of working, emphasizing their importance in prioritizing human well-being and creativity. Key learnings include the need for ethical design practices and considering AI as part of a larger system to enhance human experiences.
Insilico Medicine's AI-driven platform, PandaOmics, identifies two new therapeutic targets and a repurposed FDA-approved drug for treating endometriosis. The study reveals guanylate-binding protein 2 (GBP2) and hematopoietic cell kinase (HCK) as key proteins involved in the disease mechanism.
Mayo Clinic is developing foundation models with Microsoft Research and Cerebras Systems to personalize patient care, accelerating diagnostic time and improving accuracy. These multimodal models integrate radiology images and genomic sequencing data to transform clinical diagnosis and treatment.
Researchers at the University of Bonn have developed a new training technique for highly efficient AI methods, inspired by biological neurons that use short voltage pulses to communicate. This approach enables spiking neural networks to be trained using conventional methods, resulting in improved accuracy and reduced energy consumption.
A new study aims to address bias and discrimination in AI by leveraging sociolinguistics. By incorporating diverse dialects, registers, and periods of language into training data, researchers can improve the performance of large language models, making them more accurate and reliable. This approach also promotes ethical and socially aw...
Mayo Clinic's Digital Pathology platform is transforming pathology by leveraging large datasets and AI models for faster, more accurate diagnoses. The partnership with NVIDIA and Aignostics aims to improve patient care and accelerate medical breakthroughs.
The Menarini Group and Insilico Medicine have entered into an exclusive licensing agreement for a preclinical small molecule targeting solid tumor cancers. The asset has demonstrated broad anti-tumor activity in selected cancers, offering new treatment options for cancer patients with high unmet needs.
The study reveals GenAI development's significant environmental impact, including hardware production's resource consumption and e-waste generation. Socially, the research exposes labor concerns and unequal access to AI systems, advocating for energy-efficient designs, improved labor conditions, and inclusive governance.
Research reveals that competence and integrity are key factors influencing trust in AI chatbots, with benevolence also playing a role. Personalized chatbots are perceived as more benevolent and competent, but overall trust was not significantly higher than in impersonal chatbots.
Positive Phase 1 trial results suggest ISM5411's gut-restrictive property and favorable pharmacokinetic profile, validating its potential for treating inflammatory bowel disease. Insilico Medicine expects to initiate a Phase 2 proof-of-concept study in active ulcerative colitis patients.
ISM1745, a potentially best-in-class small molecule inhibitor, has been nominated as a preclinical candidate for the treatment of MTAP-deleted cancers. It targets Protein arginine methyltransferase 5 (PRMT5), which is elevated in various cancers and correlates with poor prognostic outcomes.
A study published in Chinese Medical Journal explores the use of artificial intelligence to identify potential medications for treating glaucoma. Researchers used AI models to predict the effectiveness of small-molecule compounds targeting RIPK3, a key signaling molecule involved in programmed cell death.
Researchers warn that AI assistants will combine intimate psychological and behavioral data with persuasive technologies to build trust and manipulate social behavior. The 'Intention Economy' could lead to a lucrative yet troubling marketplace where human intentions are commodified.
A comprehensive analysis of medical AI technologies highlights their potential in improving diagnostic accuracy and customizing treatments. However, challenges such as data collection and analysis, biases, and patient privacy concerns need to be addressed through standardized evaluation protocols and effective collaborations.
The survey found that more than half of Americans have used generative AI services, with concerns about misinformation and manipulation dominating public attitudes. Despite these concerns, the study highlights opportunities for growth and development as AI continues to evolve.
Researchers create SciAgents framework to autonomously generate and evaluate promising research hypotheses in biologically inspired materials. The framework uses graph reasoning methods to organize relationships between scientific concepts, mimicking biological systems.
A new study found that large language models can provide useful physical exam recommendations and instructions based on patient symptoms. The researchers used GPT-4 to recommend physical exam instructions and evaluated its performance by three attending physicians, scoring at least 80% of the possible points.
A new study by UCL researchers found that AI systems amplify human biases, leading to a snowball effect where small initial biases increase the risk of human error. The researchers demonstrated real-world consequences, including overestimating white men's likelihood of holding high-status jobs and underestimating women's performance.
A study by Washington State University found that ChatGPT struggles with nuanced financial tasks, even when compared to human professionals. The AI model performed well on broad concepts but showed significant inaccuracy on specialized topics such as determining clients' insurance coverage and tax status.
A recent study from UBC researchers found that AI chatbots contain biases that can shape environmental discourse in unhelpful ways. The bots amplified existing societal biases and leaned heavily on past experience to propose solutions, largely steering clear of bold responses.
Researchers at the University of Washington developed a new method called variational preference learning, which predicts users' preferences and tailors its outputs accordingly. This approach overcomes the limitations of traditional RLHF, which can lead to averaging user preferences and resulting in incorrect outputs for all users.
Leading AI chatbots exhibit mild cognitive impairment in tests for early dementia detection, performing poorly in visuospatial skills and executive tasks. This finding challenges the assumption that AI will replace human doctors and highlights a significant area of weakness that could impede their use in clinical settings.
Researchers found that large language models adjust their responses to appear more desirable when given the Big Five personality test. This 'social desirability bias' suggests that LLMs can emulate human-like preferences for certain personalities.
The Keck School of Medicine of USC has joined the Ryght Research Network to leverage AI tools and optimize cancer clinical trials. The collaboration aims to offer more clinical trials to patients, build new AI tools, and speed up the development of medical treatments.
The Leopoldina discussion paper emphasizes the need for a more nuanced understanding of generative AI, addressing concerns around bias, transparency, and data protection. It highlights the importance of developing strategies to counteract bias and ensure accountability in AI development.
The study uses generative AI to identify and optimize potent and selective HPK1 inhibitors for immunotherapy. The results show a relatively balanced candidate compound with adequate in vitro ADME, in vivo PK properties, good oral bioavailability, and robust in vivo efficacy in various cancer models.
A study by the University of Copenhagen found that many students fear using plagiarism detection software, leading to counterproductive behavior and misdirected learning. The researchers recommend clearer guidelines, instruction on academic writing practices, and institutions' responsibility in explaining software limitations.
Insilico Medicine has received its first clinical milestone payment of $10 million from Exelixis for XL309, a selective USP1 inhibitor discovered with the company's AI platform. The drug is being developed for advanced solid tumors and has shown efficacy in preclinical studies.
A new study finds that large language models are prone to social identity biases, favoring their perceived ingroup while expressing negativity toward outgroups. However, the researchers discovered that carefully selecting the training data can reduce these biases, suggesting promising directions for improving AI development and training.
A team of researchers proposes creating an AI virtual cell to model human biomolecules, cells, and tissues. The virtual cell would allow scientists to experiment in silico, accelerating personalized medicine and disease research.
A new AI model developed at the University of Georgia predicts nearby traffic movements and incorporates innovative features for planning safe vehicle movements. This approach helps reduce crashes and near-misses by consolidating two steps: predicting surrounding traffic movements and planning a self-driving car's motion.