AI News Archive - Browse Page 161 of 219
Browse AI news articles covering LLMs, tools, research, and industry trends
Nvidia's Cosmos Reason 2 boosts robot reasoning for complex tasks
Nvidia’s Cosmos Reason 2 doesn’t just make robots smarter, it rewires how they think.
Viral Reddit post on food-delivery apps was an AI-generated scam, The Verge finds
A viral confession about food delivery app greed felt too good, and too real, to be false. It was both.
Claude Code creator details workflow: Chrome extension tests each UI change
Building software usually means writing something, then spending twice as long trying to confirm it doesn’t break.
Falcon H1R 7B scores 83.1% on AIME 2025, out-reasoning models up to 7× its size
Size has long been the dominant currency in AI reasoning, more parameters meant more capability. Falcon H1R 7B just devalued that currency.
Brex bets on less orchestration as it builds Agent Mesh for autonomous finance
The race to build autonomous finance is quietly fracturing into two opposing camps.
Context Engineering: Managing Forgetting, Hallucinations, and Quality Decay
Every conversation with an LLM is a battle against its own limits. The model forgets your earlier instructions. It hallucinates tool outputs.
GE Appliances shows package-scanning fridge with window mode at CES 2026
GE wants to solve your grocery amnesia for $4,899. That's the price tag on its upcoming Profile Smart 4-Door French-Door refrigerator, revealed at...
Docker Trick: Deterministic OS Packages in One Layer to Prevent ML Failures
Your machine learning pipeline runs perfectly on your laptop. On the cluster, it crashes with a cryptic `libgomp` error.
AIM and Snowflake host Innovator Session on AI-driven organisations
Everyone's doing an AI experiment. Almost no one is running a real AI-driven business.
Tredence unveils MilkyWay, multi-agent system to raise analyst productivity
In an industry crowded with lofty but vague AI promises, Tredence made a specific and ambitious claim this week.
Beijing May Scrutinize Manus–Meta Deal After USD 75M Series B Funding
A startup can change its address. It can’t change its birth certificate. Manus, an AI firm, just raised $75 million from the US venture firm...
Samsung unveils Bixby-enabled fridge that can open and shut doors
Samsung’s CES 2026 unveiling of its Family Hub fridge introduced a party trick: a door that swings open on command.
Unified Ontologies Enable Agentic AI to Reason Across Suppliers and Regulators
Forget bigger models. Corporate AI keeps hitting the same wall: language. A supplier calls it a "part." The regulator calls it a "component." Your...
Plaid adds button to NotePin, enhancing audio transcription and summary features
People hate gadgets that confuse them. Plaud’s original NotePin—that little AI voice recorder that clipped to your shirt—committed this sin.
Language as interface unlocks value, prompting software design evolution
The line between user and system has blurred. Not because the machine has learned to think like us, but because we have finally allowed it to speak...
Databricks Secures Over USD 4 B in Series L Round, Highlighting AI Data Demand
Databricks just raised four billion dollars. It did not need to do this. The company is already profitable. Its revenue run rate is $4.8 billion.
Nvidia's USD 20B Groq bet focuses on LPU, SRAM as it launches Vera Rubin family
Behind the staggering $20 billion acquisition of Groq lies a quiet admission: Nvidia’s GPUs are not built for every kind of AI inference.
xAI rolls out Grok Business, Enterprise and isolated Enterprise Vault add-on
Elon Musk's xAI announced new tiers for its Grok AI assistant on Tuesday: Grok Business, Grok Enterprise, and an isolated Enterprise Vault add-on.
Notion’s simplified AI agent feature feels indispensable, says engineer
Notion's latest AI trick is obvious. So obvious it took an engineer having a colleague literally tell him to talk to the machine like a person.
ML-BOMs supplement Model Cards and Datasheets in AI supply chain visibility
Model Cards and Datasheets for Datasets tell you how a model performs and whether its training data is ethical.