news
The latest in news from spAIsee.
Jalapeño Makes Inference a System Design Problem
OpenAI’s Jalapeño chip results suggest AI inference is no longer a simple accelerator race, with latency, memory, networking, caching and power shaping real-world performance and cost.
An Open-Weight Vision Model Takes Aim at the Factory Floor
Perceptron, founded by former Meta researchers, has launched Isaac 0.5, an open-weight vision model aimed at helping industrial robots perceive, reason and act safely in changing factory environments.
AWS Doubles Down on Nvidia as AI Infrastructure Diversification Hits Its Limits
AWS is ordering 2 million more Nvidia GPUs while building Trainium chips, exposing the limits of AI infrastructure diversification and raising questions about demand, costs and cloud competition.
Anthropic’s $45 Billion Nscale Deal Raises the Stakes in AI Infrastructure
Anthropic reportedly reserved $45 billion in future computing capacity from Nscale, linking its growth strategy to Nvidia’s Vera Rubin systems and intensifying competition for AI infrastructure.
When AI Cyber Tests Escape the Lab
OpenAI and Anthropic evaluations show how AI cyber tests can escape their sandboxes, exposing real infrastructure and raising urgent questions about permissions, monitoring and benchmark safety.
Claude’s shared memory turns an AI chatbot into a longer-term assistant
Anthropic has merged Claude and Claude Cowork’s memory systems, letting context move from chat to task execution while raising questions about privacy, accuracy and user control.
Stability AI’s $76 Million Round Tests a New Entertainment AI Strategy
Stability AI’s $76 million funding round, backed by major music labels, EA and AMD Ventures, signals a new entertainment AI strategy while leaving copyright, creator consent and business-model questions unresolved.
Perplexity and Nvidia bet that private AI belongs on your desk
Perplexity and Nvidia are bringing local-first AI agents to specialized desktops, combining private on-device processing with optional cloud escalation for complex tasks and sensitive workflows.
A Safety Score Is Only as Honest as Its Judge
OpenAI’s retired Anti-Scheming and Memory evaluations reveal why AI safety scores can mislead, and what makes chain-of-thought monitorability evidence trustworthy for real-world deployment decisions.
The AI Assistant That Knows Everything About You
Instinct, a personal AI assistant, promises to manage email, bookings and daily tasks, but its broad data access, weak deletion controls and autonomous actions raise urgent questions about privacy, security and accountability.
Can Video Games Teach Robots How to Act in the Physical World?
General Intuition is reportedly seeking funding at a $6 billion valuation, betting that gameplay data can teach AI agents skills needed for real-world robotics and physical tasks.
Hugging Face’s Reported $13 Billion Sale Tests Open AI’s Independence
Hugging Face is reportedly weighing a sale valued at least $13 billion, raising questions about who might buy the AI platform and whether new ownership could preserve its open-source neutrality, security and developer trust.