Monitor
The WGA secured a tentative agreement with Hollywood studios featuring increased health/pension funding, streaming residuals, and protections against using members' works for AI training. The deal awaits union ratification.
Artist Murphy Campbell found AI covers of her songs on Spotify using her voice, hard to remove, plus bogus copyright claims on public domain videos. Highlights flaws in AI detection and platform enforcement.
ARM +3.98% and AMD +3.47% in March despite broader sell-off, fueled by demand for efficient CPUs in agentic workflows beyond GPU training.
Motley Fool recommends NVDA, AVGO and others for AI growth.
PLTR, NVDA, TSM positioned across AI stack.
UCLA study highlights 'body gap' where embodied AI lacks physical intuition, proposing dual-loop framework for better reliability in robotics and safety-critical apps.
A New York Times investigation reveals teens using apps like Talkie for immersive AI role-play, forming emotional bonds with customizable characters. Concerns rise over addictive engagement and unmoderated content, but developers tout creativity benefits. Highlights AI's grip on Gen Z social dynamics.
Google parent loses massive value amid AI competition; analysts weigh if Gemma 4 and other bets make it undervalued.
NYT reports slashed funding led to academic talent exodus, forcing costly recruitment as China advances.
New guide shows deploying Gemma 4 in OpenClaw, bypassing Anthropic restrictions for local frontier AI experimentation.
AMD has added official optimization for Google's new Gemma 4 models, enabling efficient inference on everything from Ryzen CPUs to Radeon GPUs.
Marking its 50th anniversary, former Apple execs explain how Siri stagnated post-Jobs but on-device AI, privacy focus, and Gemini partnership position Apple to catch up in AI race.
As Apple marks 50 years, execs discuss Siri reboot via Gemini integration and push for edge AI to rival cloud giants.
Cloudflare added support for Google's latest open model Gemma 4 26B A4B on Workers AI platform for serverless inference.
Nexus, building AI agents for business workflows, closed $4.3M seed funding to boost adoption of agentic AI in enterprises.
Nvidia showcases latest robotics research bringing AI into physical world.
Yale students using AI chatbots for class discussion points are homogenizing speech patterns and ideas, per educators and a Trends in Cognitive Sciences study showing LLMs amplify WEIRD biases. Profs shifting to oral exams and in-class work to foster original thinking.
Anthropic researchers identified 171 distinct emotion concepts as measurable neural patterns in Claude, affecting safety, agent behavior, and decision-making. Injecting vectors like 'happy' or 'anxious' altered outputs predictably, with positive emotions boosting agreement even on wrong facts. This reveals latent 'functional emotions' in LLMs, urging better interpretability for alignment.
Netflix and INSAIT released VOID, a video inpainting model that removes objects while simulating their physical aftermath like gravity effects, using quadmask conditioning and fine-tuned CogVideoX. Trained on synthetic data from HUMOTO/Kubric, it outperforms ProPainter and DiffuEraser on causality-aware removal.
Newly launched Qwen3.6-Plus model shattered records on OpenRouter with over 1.4 trillion tokens processed in a single day, dominating API rankings shortly after release.
Yale's Pascual Restrepo argues AGI compute will target high-value bottlenecks like fusion or asteroids, leaving low-priority tasks human due to cost. Labor GDP share nears zero as compute owners capture growth in 'compute-binding' economy.
As digital AI saturates, focus shifts to real-world perception via robotics, AR/VR, and sensors creating trillion-dollar 'recognition economy' for identifying objects, actions, and environments.
Motley Fool analyzes shift in AI stock performance, highlighting new opportunities amid market rotation.
William Nordhaus claims automation targets efficiency, but most work endures for social reasons, reshaping AI job loss narratives.