All Updates

Page 1822 of 1910

February 23, 2026

📰
The Verge182d ago

Dark Sky team launches Acme Weather

Ex-Dark Sky creators debuted Acme Weather for iOS, featuring main forecast plus alternates to address inaccuracy. Offers two-week free trial, then $25/year. Android version planned without timeline.

#weather-forecast#ios-app#ensemble-prediction
☁️
AWS Machine Learning Blog182d ago

Hexagon Boosts AI Models via SageMaker HyperPod

Hexagon collaborated with AWS to scale AI model production using SageMaker HyperPod. They pretrained state-of-the-art segmentation models on its infrastructure. This accelerates their AI development pipeline.

#model-training#segmentation-models#hyperpod
🇬🇧
BBC Technology182d ago

AI Exec Takes Microsoft Gaming Helm

AI executive Asha Sharma has been appointed to Microsoft's top gaming role, prompting fan backlash and debates online. Some fear it signals 'the end of Xbox' due to her limited gaming experience. The shake-up highlights tensions in Microsoft's gaming strategy.

#leadership-change#gaming-ai#fan-backlash
🔬
MIT Technology Review182d ago

Hidden Human Labor Powers Humanoid Robots

MIT Technology Review reveals the concealed human work essential for training and demonstrating humanoid robots. Nvidia CEO Jensen Huang proclaims the era of physical AI, extending beyond language models to real-world embodiment. The story originates from their weekly AI newsletter, The Algorithm.

#physical-ai#embodied-ai#robotics-labor
💼
VentureBeat182d ago

3x LLM Inference Speed Baked into Weights via MTP

Researchers developed Multi-Token Prediction (MTP) to enable LLMs to predict multiple tokens per forward pass, achieving 3x throughput without extra models. Unlike speculative decoding, it modifies training to focus on token sequence joints. Ideal for agentic workflows reducing single-user latency.

#inference-speedup#agentic-workflows
💰
TechCrunch AI182d ago

Particle Adds Podcast Clip Extraction

Particle's AI news app now extracts key moments from podcasts as short, relevant clips. Users can instantly play these alongside related stories. This feature enhances news consumption efficiency.

#podcast-ai#news-app#audio-extraction
💰
TechCrunch AI182d ago

Spotify AI playlists hit UK markets

Spotify is expanding its AI-powered Prompted Playlists feature to Premium users in the UK, Ireland, Australia, and Sweden. The tool generates custom playlists from text prompts. This continues Spotify's AI experimentation in music discovery.

#ai-music#playlist-generation#consumer-ai
🖥️
Computerworld182d ago

Apple's Storage Upsell Fuels AI Era Revenue

Apple generates major services revenue from iCloud subscriptions and device storage upsells, with 70% of US buyers paying for iCloud. 46% of iPhone buyers upgrade storage, boosted by Apple Intelligence needs. Strategy exploits ecosystem integration and photo usage.

#storage#services-revenue#ecosystem
🖥️
Computerworld182d ago

OpenAI Partners with Consultants for Frontier

OpenAI announced Frontier Alliances with Accenture, BCG, Capgemini, and McKinsey to deploy enterprise AI agents. Partners will integrate Frontier, an orchestration layer for AI agents, into client systems and workflows. Limited availability now, wider rollout soon.

#ai-agents#partnerships
📊
Bloomberg Technology182d ago

AI Scare Trade Crushes IBM Shares

AI disruption fears sparked a 'scare trade' hitting delivery, payments, and software stocks. IBM saw its worst plunge in 25 years amid broader concerns.

#market#disruption#stocks
📰
The Verge182d ago

Discord drops Persona amid verification backlash

Discord conducted limited UK test of Persona age verification, now ended due to user backlash over face scans and ID uploads. Company distances itself from provider. Global age checks start next month.

#age-verification#privacy-backlash#biometrics
🇬🇧
The Register - AI/ML182d ago

Regulators: AI Image Tools Face Privacy Scrutiny

Global privacy watchdogs warn generative AI firms that tools creating realistic people images must comply with data protection laws. Companies cannot ignore rules when training or deploying such models.

#privacy-regulation#data-protection#gen-ai
📰
The Verge182d ago

Big Tech slow on AI slop detection

Instagram's Adam Mosseri calls for cryptographic signing of real media to combat AI fakes. Progress on reliable deepfake labeling remains sluggish despite AI provider promises. Focus on chain-of-custody for authenticity.

#deepfakes#ai-slop#content-authenticity
☁️
AWS Machine Learning Blog182d ago

Agentic AI: smolagents on AWS Tutorial

Hugging Face smolagents is an open-source library for building agents with minimal code. Integrate it with AWS services to deploy a healthcare AI agent. Features multi-model deployment, vector RAG, and clinical decision support.

#agentic-ai#multi-model#healthcare-ai
💼
VentureBeat182d ago

Claude Code Security Launches Post 500+ Vulns

Anthropic's Claude Opus 4.6 identified 500+ high-severity vulnerabilities in open-source codebases, leading to the launch of Claude Code Security on Feb 20. The tool uses reasoning to analyze data flows and business logic beyond pattern-matching scanners like CodeQL. Now available to Enterprise and Team customers as a research preview.

#reasoning-scanning#code-security
⚛️
Ars Technica AI182d ago

LLMs Memorize and Regurgitate Full Novels

New research shows AIs can generate near-verbatim copies of entire novels from memorized training data. LLMs retain far more training content than previously estimated. This raises serious concerns over data extraction and copyright.

#memorization#copyright-risk#data-extraction
🌍
The Next Web (TNW)182d ago

OpenAI Launches Frontier Alliances for Enterprise AI

OpenAI announced Frontier Alliances, partnering with Boston Consulting Group, McKinsey & Company, Accenture, and Capgemini. The initiative helps enterprises shift AI from pilots to full production integration in workflows. It broadens OpenAI's support for large organizations.

#partnerships#enterprise-ai#consulting
Grammarly182d ago

Superhuman Go Adds Box, Gamma, Wayground Agents

Superhuman expanded its Superhuman Go AI agent ecosystem with three new partner agents from Box, Gamma, and Wayground (formerly Quizizz). These agents integrate enterprise knowledge, visual creation, and interactive learning directly into user workflows. The announcement was made on February 23, 2026.

#ai-agents#partnerships#productivity
🇬🇧
The Register - AI/ML182d ago

Gemini Chat Histories Vanish Post-3.1 Update

Users report months of Google Gemini conversations disappearing recently. Issues coincide with Gemini 3.1 rollout; Google attributes to temporary bug.

#chat-history#bug-report#model-update
💰
TechCrunch AI182d ago

Pentagon summons Anthropic over Claude

Defense Secretary Pete Hegseth summons Anthropic CEO Dario Amodei to discuss military use of Claude. Hegseth threatens to label Anthropic a supply chain risk. This escalates tensions over AI in defense applications.

#military-ai#supply-chain
Page 1822 of 1910