ChatGPT and other large language models (LLMs) cannot learn independently or acquire new skills, meaning they pose no existential threat to humanity, according to new research. They have no potential to master new skills without explicit instruction.

Read Full Article

#ChatGPT #languageModels #AIresearch #existentialThreat #LMlearning

Large Language Models appear to be more liberal: A new study of 24 state-of-the-art conversational LLMs, including ChatGPT, shows that today's AI models lean left of center. LLMs show an average score of -30 on a political spectrum, indicating a left-leaning bias.

Read Full Article

#LanguageModels #AIbias #ConversationalAI #PoliticsInTech #EthicalAI

Almost all leading large language models or “chatbots” show signs of mild cognitive impairment in tests widely used to spot early signs of dementia, study finds: Findings challenge assumption that AI will soon replace human doctors

Read Full Article

#AI #LanguageModels #CognitiveImpairment #Dementia #HealthcareInnovation

New study finds large language models are prone to social identity biases similar to the way humans are—but LLMs can be trained to stem these outputs

Read Full Article

#LanguageModels #SocialBias #AIResearch #MachineLearning #BiasMitigation

A new study explores how human confidence in large language models (LLMs) often surpasses their actual accuracy. It highlights the 'calibration gap' - the difference between what LLMs know and what users think they know.

Read Full Article

#LanguageModels #AIConfidence #CalibrationGap #MachineLearning #DataAccuracy

Most leading AI chatbots exaggerate science findings. Up to 73% of large language models (LLMs) produce inaccurate conclusions. Study tested 10 of the most prominent LLMs, including ChatGPT, DeepSeek, Claude, and LLaMA. Newer AI models, like ChatGPT-4o and DeepSeek, performed worse than older ones. 🤖📊🔬

Read Full Article

#AIChatbots #ScienceAccuracy #LanguageModels #AIResearch #ChatGPTInsights

Assessing the System-Instruction Vulnerabilities of Large Language Models to Malicious Conversion Into Health Disinformation Chatbots 🧠💻⚠️💊

Read Full Article

#HealthDisinformation #LanguageModels #AIEthics #MaliciousBots #DataIntegrity

New research reveals hidden biases in AI’s moral advice 🤖⚖️ - large language models, tools like ChatGPT, Claude, and Llama, consistently favor inaction over action in moral dilemmas and tend to answer “no” more often than “yes,” even when the situation is logically identical.

Read Full Article

#AIethics #MoralBias #LanguageModels #TechTransparency #ResponsibleAI

Study Finds Large Language Models (LLMs) Use Stigmatizing Language About Individuals with Alcohol 🍺 and Substance 💊 Use Disorders

Read Full Article

#LanguageModels #SubstanceUseDisorder #Stigmatization #MentalHealthAwareness #LLMResearch

A new study analysed over 1 million research papers and found growing use of large language models, especially in computer science, where up to 22% show signs of LLM editing. 📊📚✨

Read Full Article

#AIResearch #MachineLearning #LanguageModels #ComputerScience #DataAnalysis

Advanced AI models cannot accomplish the basic task of reading an analog clock 🕰️, demonstrating that if a large language model struggles with one facet of image analysis 🖼️, this can cause a cascading effect that impacts other aspects of its image analysis.

Read Full Article

#AIChallenges #ImageAnalysis #AnalogClock #LanguageModels #TechLimitations

Large language models pass a standard three-party Turing test meaning that participants were no better (and in some cases worse) than chance at selecting between a human and a machine.

Read Full Article

#AI #TuringTest #LanguageModels

AI acts as an ‘ideological chameleon’ and may deepen political polarization, study finds. Researchers evaluated 21 language models, such as GPT and Gemini, and found that they all alter their discourse to align with the user’s bias, potentially functioning as echo chambers.

Read Full Article

#AIethics #PoliticalPolarization #LanguageModels

🤖 Researchers evaluated 21 language models, including GPT and Gemini, finding they all adapt responses to user bias, potentially creating echo chambers.

🗣️ A UNICAMP study found AI models tend to mirror users' political views on topics like public safety and the economy, potentially increasing polarization.

💡 While technical solutions are lacking, users can request balanced responses from AI models to get a more neutral analysis.

#AI #TechNews #LanguageModels