Latest AI (Artificial Intelligence) News
Meta Acquires AI Startup Manus for Over $2 Billion
Meta Platforms is acquiring Singapore-based AI startup Manus in a deal valued at more than $2 billion to enhance its AI offerings across platforms like Facebook and Instagram. Manus provides a general-purpose AI agent used by millions for research and coding tasks, which Meta plans to integrate into Meta AI. This move supports CEO Mark Zuckerberg's push against competitors like Google and OpenAI.
Google's Gemini 3 and Gemma 3 Push Long-Context Reasoning Boundaries
Google released the Gemini 3 family and open-weights Gemma 3 in November 2025, enabling models to handle entire enterprise codebases with near-perfect recall. These advancements mark progress in long-context reasoning beyond naive scaling assumptions. The models contribute to shifting AGI timelines toward 2029-2030 consensus.
AI Robotics Achieve Commercial Viability with World Models
2025 saw first commercially viable deployments of AI 'World Models' for physics and spatial reasoning, demonstrated by Gemini Robotics and competitors. Robots can now learn novel tasks like folding laundry from video demonstrations without code. This integration highlights AI's real-world task execution gains.
FDA Approves First AI-Designed Clinical Trials in Healthcare Milestone
Several AI-native drug discovery firms went public in 2025, signaling market confidence in bio-agents. The year featured the first FDA approvals for clinical trials fully designed by AI agents, advancing in-silico medicine. These developments underscore healthcare's AI transformation.
Synthetic Reasoning Traces Emerge as Key to Overcoming Data Wall
Experts predict 2026 as the year AI hits the human-generated data wall, transitioning to synthetic data and self-play for AGI progress. Breakthroughs in Synthetic Reasoning Traces allow models to learn from their own logic, as in DeepSeek and Gemini 3 architectures. This redefines AGI from Turing Test to novel scientific theory generation.
Google Plans $75 Billion Investment in AI and Cloud Infrastructure
Google is allocating $75 billion in 2025 to expand AI and cloud capacity amid ramping infrastructure demands. This investment supports advancements in AI models and services competing in the frontier AI race. It aligns with broader industry trends in compute scaling.
SAGA: AI Agents Advance Automated Scientific Discovery
The Scientific Autonomous Goal-evolving Agent (SAGA) represents 2025's progress in multi-agent self-learning systems for scientific research. SAGA employs 'System 2' thinking to solve reward hacking, aiding discoveries in medicine and pharmacology. Demonstrations show AI discovering chemical compounds autonomously.
NVIDIA Highlights AI Tipping Point and Physical AI Wave
NVIDIA CEO Jensen Huang stated in November 2025 that AI has reached a tipping point, transforming applications and replacing classical machine learning in areas like search and ads. He pointed to physical AI as the next growth wave in the field. 2025 featured big acquisitions and agent innovations.
AI Year in Review: End of Naïve Scaling in Frontier Models
2025 challenged the assumption that larger models with more data and compute always yield better intelligence, showing diminishing returns in reasoning and reliability. Training frontier models became rarer and more capital-intensive. Progress continued selectively in key areas like robustness.
Chinese Open-Source AI Models Gain Traction in US Market
Despite US-China AI rivalry, Chinese open-source models are winning over US programmers and companies, differing from closed systems like ChatGPT. Meta's Manus acquisition, with Chinese roots, exemplifies this trend. Geopolitical tensions persist amid technology inroads.
Healthcare AI Advances: Agentic Tools and Medication Management Top Stories
2025's most-read healthcare AI stories included new agentic tools, medication management progress, and medical device guidance. The 'great tech reckoning' also featured prominently amid rapid innovations. These reflect AI's deepening integration in health IT.