Meta's V-JEPA AI Model Learns Physics from Video
Meta's V-JEPA system learns to understand physical world dynamics by analyzing ordinary videos, according to WIRED.
Sources · WIRED AI
The news desk
Follow what’s changing.
Recent developments in artificial intelligence, with the original sources one click away.
All news
1541–1560 of 1,902Google Announces Gemini 3 Pro with Advanced Multimodal Capabilities
Google introduces Gemini 3 Pro, positioning it as the leading model for multimodal AI capabilities with enhanced vision features.
Sources · Google AI Blog
New York Times Files Copyright Lawsuit Against AI Start-Up Perplexity
The New York Times sued Perplexity for copyright infringement, joining over 40 other disputes between copyright holders and AI companies.
Sources · NYT Technology / TechCrunch AI
OpenAI Declares 'Code Red' Amid AI Competition Pressures
OpenAI has called a 'Code Red' as the company faces mounting competitive pressure to maintain its lead in AI development.
Sources · NYT Technology
Organizations Struggle to Move AI Projects from Pilot to Production, MIT Technology Review Reports
Despite high AI investment, three-quarters of enterprises face challenges scaling AI pilots to full production deployment.
Sources · MIT Technology Review
Enterprises Struggle to Scale AI Beyond Pilot Phase Despite High Investment
Three-quarters of enterprises face challenges moving AI projects from pilot to production, according to MIT Technology Review report.
Sources · MIT Technology Review
Three New arXiv Papers Address AI Applications in Hardware Verification, Healthcare, and Security
Recent research explores LLMs for RTL verification, multi-agent systems for headache diagnosis, and security threat modeling for AI agents.
Sources · arXiv AI (cs.AI)
Three New arXiv Papers Explore LLM Applications in Optimization, Healthcare, and Psychological Analysis
Researchers publish papers on using genetic algorithms with LLMs, multi-LLM collaboration for medication recommendations, and analyzing psychological states.
Sources · arXiv AI (cs.AI)
Three New arXiv Papers Explore LLM Applications in Time Series, Gaming, and Medical Practice
New research addresses LLM limitations in time series forecasting, game decision-making, and medical alignment through novel approaches.
Sources · arXiv AI (cs.AI)
Three New arXiv Papers Explore LLM Enhancements for Forecasting, Decision-Making, and Medical Applications
Recent arXiv preprints address limitations in LLM applications for time series forecasting, game-based reasoning, and medical practice alignment.
Sources · arXiv AI (cs.AI)
Three New arXiv Papers Explore LLM Optimization, Healthcare Applications, and Psychological Analysis
Recent arXiv publications examine genetic algorithms for LLMs, multi-model medication recommendations, and emotion detection in conversations.
Sources · arXiv AI (cs.AI)
Three New Research Papers Address AI Applications in Hardware Verification, Healthcare, and Security
Recent arXiv preprints introduce AI systems for RTL verification, headache diagnosis, and security threat modeling for AI agents.
Sources · arXiv AI (cs.AI)
Three New Research Papers Address LLM Agent Safety, Context Handling, and Research Capabilities
Recent arXiv papers tackle kill switches for malicious web agents, long-context reasoning evaluation, and hierarchical research agent architecture.
Sources · arXiv AI (cs.AI)
Three New Research Papers Address Web-Based LLM Agent Capabilities and Safety
Recent arXiv papers examine security controls, long-context reasoning, and hierarchical architectures for LLM-based web agents.
Sources · arXiv AI (cs.AI)
Two New arXiv Papers Explore LLM Vulnerabilities and Reasoning Capabilities
Researchers introduce 'Doublespeak' attack method and investigate iterative reasoning improvements in large language models.
Sources · arXiv AI (cs.AI)
New Multi-Agent Systems Target Cooperation and GPU Optimization Challenges
Researchers introduce AsymPuzl for evaluating agent cooperation and Astra for optimizing GPU kernel performance using multi-agent approaches.
Sources · arXiv AI (cs.AI)
Three New arXiv Papers Address LLM Agent Capabilities and Safety Challenges
Recent arXiv preprints explore deep research agents, social interaction evaluation, and safer LLM instruction handling through hierarchical approaches.
Sources · arXiv AI (cs.AI)
Three New arXiv Papers Advance Federated Learning Efficiency and Decision Systems
Recent arXiv publications explore energy-efficient federated learning for medical imaging, wireless edge AI integration, and improved decision support systems.
Sources · arXiv AI (cs.AI)
Three New arXiv Papers Examine Foundation Models, World Modeling, and Adversarial Training
Recent research explores tabular foundation models, world modeling in transformers, and the relationship between robust models and attack transferability.
Sources · arXiv AI (cs.AI)
Three New Research Papers Address Privacy and Application Challenges in Large Language Models
Recent arXiv preprints explore membership inference attacks on RAG systems, econometric frameworks for LLM text analysis, and KV-cache privacy risks.
Sources · arXiv AI (cs.AI)