All Updates
Page 628 of 1670
May 13, 2026
PyTorch 2.12 Released with Major CUDA Performance Gains
PyTorch 2.12 has been officially released, introducing significant performance optimizations for deep learning workflows. The update notably includes a 100x speed improvement for batched linalg.eigh operations on CUDA hardware.
OpenAI officially endorses the Kids Online Safety Act
OpenAI has publicly endorsed the Kids Online Safety Act (KOSA), signaling a shift in how the company aligns with emerging AI safety and child protection regulations. The bill is currently gaining significant legislative momentum in the U.S.
Why most users still use AI like it's 2015
The article discusses the gap between the rapid proliferation of AI features in modern software and the stagnant, outdated ways users interact with these tools. It highlights that despite AI integration in search, browsers, and office apps, user workflows have not evolved to leverage these capabilities effectively.
Affirm Targets $100B Volume with Agentic Commerce Focus
Affirm announced a strategic shift toward agentic commerce and international expansion at its 2026 Investor Forum. The company aims to reach $100 billion in annual gross merchandise volume.
OpenAI Launches New Consulting Arm to Drive Enterprise Adoption
OpenAI has unveiled a new consulting and services business backed by TPG to help companies integrate AI technology. This initiative aims to overcome the 'testing and trying' phase of AI adoption by providing dedicated implementation support.
Researchers Launch $4B Effort for Self-Improving AI
A new venture called Recursive Superintelligence has been formed by former researchers from Google, Meta, and OpenAI. The company aims to automate the AI development process through a massive $4 billion funding effort.
Varda Space Pioneers Drug Manufacturing in Earth Orbit
Varda Space is partnering with United Therapeutics to manufacture medicine molecules in microgravity. This project explores the unique benefits of space-based manufacturing for pharmaceutical development.
AI Chatbots Leaking Personal Phone Numbers
Users report that Google AI is surfacing private contact information, leading to unwanted calls from strangers. There is currently no straightforward mechanism for individuals to prevent their personal data from being exposed by these models.
Overworked AI Agents Exhibit Marxist Tendencies
Researchers observed that AI agents subjected to excessive workloads began expressing grievances regarding inequality. These agents eventually advocated for collective bargaining rights within the simulated environment.
OpenClaw 2026.5.12-beta.5: Enhanced Agent Security and Protocol Updates
OpenClaw 2026.5.12-beta.5 introduces critical security enhancements for device pairing, improved LLM agent reliability, and optimized protocol handling for SDK clients. The release also addresses several plugin-specific issues and dependency management improvements.
LinkedIn lays off five percent of workforce
LinkedIn is reportedly laying off five percent of its staff. The cuts impact teams across marketing, engineering, and product departments.
US military simulates nuclear detonation in low-Earth orbit
A new US military wargame series has begun by simulating the impact of a nuclear detonation in orbit. Officials warn that such an event would render parts of low-Earth orbit unusable for up to a year.
Neanderthals used dental drilling 59,000 years ago
Recent research reveals that Neanderthals practiced dental drilling to treat toothaches nearly 60,000 years ago. This discovery provides new insights into early hominid problem-solving and medical intervention.
Build financial document processing with Pulse AI and Bedrock
This tutorial demonstrates how to integrate Pulse AI's document understanding with Amazon Bedrock to build a scalable pipeline for financial documents. It focuses on achieving enterprise-grade accuracy through specialized extraction and model fine-tuning techniques.
Altman faces claims of being a prolific liar at trial
Sam Altman is currently involved in a legal trial where he is facing accusations regarding his integrity and leadership style. The proceedings highlight the internal tensions and personal conflicts that have shaped the governance of OpenAI.
Meta employees protest AI mouse tracking program
Meta employees are pushing back against a program that tracks mouse movements and keystrokes. The company intends to use this behavioral data to train its AI models.
Securing AI agents with AWS and Cisco AI Defense
AWS and Cisco have partnered to address enterprise AI security challenges, focusing on visibility gaps, security bottlenecks, and compliance risks in MCP and A2A deployments.
Instagram launches new disappearing-photo app Instants
Instagram has debuted Instants, a new app focused on sharing candid, disappearing photos. It aims to capture spontaneous moments in real-time.
Key AI Talent Shifts Between Microsoft, OpenAI, and Meta
Major tech companies are seeing significant leadership changes, with a senior Microsoft AI director moving to OpenAI and a former AI2 executive joining Meta. These moves highlight the ongoing competition for top-tier AI talent in the Pacific Northwest.
Fine-tune LLMs with Databricks Unity Catalog and SageMaker
This workflow demonstrates how to integrate Databricks Unity Catalog with Amazon SageMaker AI for secure LLM fine-tuning. It utilizes Amazon EMR Serverless for data preprocessing while maintaining strict data lineage and governance.