1+ day, 14+ hour ago   (294+ words) The new adapter lets Fetch.ai's autonomous agents communicate with external AI agents through Google-linked A2A protocol Fetch.ai has shipped a tool that could quietly reshape how autonomous AI agents talk to each other across platforms. The A2A Outbound Adapter, part…...

3+ day, 7+ hour ago   (331+ words) Back-to-back launches from the two AI giants push the proprietary lead in coding benchmarks to nearly 1,800 points, leaving open-source alternatives further behind than ever. Anthropic’s latest model ships with a 1 million-token context window and a 128K max output, numbers that let…...

3+ day, 7+ hour ago   (238+ words) The TikTok parent's research team reworked a DeepSeek technique to align visual generation models with human preferences, posting benchmark gains of up to 181%. The work, developed by ByteDance’s Seed team in collaboration with researchers at the University of Hong Kong,…...

3+ day, 8+ hour ago   (282+ words) The new model conditions directly on original audio performances, preserving tone, pacing, and emotion across more than 90 languages. ElevenLabs just shipped what might be the most significant upgrade to AI-powered dubbing since the technology left the research lab. Dubbing v2, the…...

3+ day, 11+ hour ago   (325+ words) The Chinese AI startup's latest model activates a fraction of its total parameters to deliver performance rivaling GPT-5.6 at dramatically lower cost DeepSeek just dropped a model that processes a million tokens of context while activating fewer parameters than some…...

3+ day, 12+ hour ago   (430+ words) The decentralized AI project is deploying autonomous agents at schools like San José State and Cal State Long Beach, with plans to expand through 2026. Fetch.ai’s Innovation Lab has been quietly rolling out custom AI agents across universities in the…...

4+ day, 11+ hour ago   (399+ words) The KARL agent matches frontier model accuracy at a third less cost by learning to stop searching when it has enough information The hardest thing to teach a search system is restraint. Most AI agents, when tasked with finding information,…...

5+ day, 3+ hour ago   (379+ words) New API access, an AI security plugin, and a public hook registry give v4 developers a more complete toolkit for building custom pool logic. Think of hooks as modular plug-ins for Uniswap’s liquidity pools. They’re external smart contracts that can inject…...

5+ day, 5+ hour ago   (417+ words) A lightweight diffusion adapter lets large language models generate multiple tokens in parallel without sacrificing output quality Getting a large language model to think faster usually means making it think worse. A new method called Uno claims to break that…...

5+ day, 7+ hour ago   (342+ words) AWS used its own agentic AI coding tool to overhaul its inference engine, a project originally scoped for 30-40 engineers over 18 months Six engineers. Seventy-six days. That’s what it took AWS to rebuild the inference engine behind Amazon Bedrock, its flagship…...