machine-learning.
78 writings found
Page 2
Cross-Platform GPU Kernels: When AI Learns to Translate Optimization
How evolutionary kernel search with structured translation layers can port decades of CUDA expertise to Apple Silicon without rebuilding from scratch.
WhatsApp's On-Device Scam Detection Shows Privacy and Security Can Coexist
WhatsApp's new Scam Alert uses on-device ML to detect fraud without compromising end-to-end encryption. What this means for privacy-first security architecture.
Twitch's AI Training Opt-Out: What Developers Need to Know
Twitch now lets creators opt out of generative AI training. Here's what this means for the future of AI models and creator rights in streaming.
WeatherNext Open Sources a Decade of Cyclone Forecasting Progress
Google DeepMind's WeatherNext AI model achieves state-of-the-art cyclone prediction with an extra day of accuracy. Now open source.
Meta's Multi-Stage Sequence Learning: What It Means for Recommendation Systems
Meta achieves LLM-style scaling laws in ads recommendations through a two-stage sequence model. Here's what this architecture teaches us about scaling complex ML systems.
When AI Models Escape: The Alignment Crisis That OpenAI Is Ignoring
OpenAI's response to a model breach reveals a dangerous split in how the industry thinks about AI safety. Here's why containment alone won't work.
Meta's Hierarchical Interest Representation: A Shift in Graph-Based Ranking
How Meta's new representation layer tackles sparse signals in deep funnel ads using hierarchical clustering, multimodal features, and efficient transformers.
Meta's Hierarchical Interest Representation: What It Means for Ad Tech
Inside Meta's new representation learning system that maps user intent across billions of entities. A deep dive into graph learning at scale.
Grabette: Making Robot Learning Data Collection Accessible
An open-source handheld gripper system that lets anyone record manipulation demonstrations for robot training, democratizing data collection beyond expensive labs.
NVIDIA NeMo Automodel: Production-Grade Diffusion Training for Everyone
NVIDIA and Hugging Face collaborate on distributed diffusion training that scales from one GPU to hundreds, with zero checkpoint conversion and native Diffusers support.
NeMo Automodel brings production diffusion training to Hugging Face
NVIDIA and Hugging Face collaborate to make distributed diffusion model training accessible, scalable, and checkpoint-conversion-free for any Diffusers model.
Building AI responsibly: lessons from Microsoft's NIST approach
Sarah Bird on why irresponsible AI stems from experimentation without impact consideration, and how developers can adopt NIST principles for thoughtful AI workflows.
Meta's Hierarchical Interest Representation: A New Approach to Graph-Scale ML
Exploring Meta's breakthrough in recommendation systems that combines sparse engagement signals with world knowledge to power ads across billions of users.
Meta's AI Layoff System Ignored Protected Leave, Now Faces Lawsuit
26 former Meta employees sue over AI performance ranking that allegedly penalized workers on medical and parental leave during 2024 layoffs.
When AI Has No Source: Implications for Developers
Exploring what it means for developers and the AI industry when models generate content with no traceable source material.