#synthetic data
9 stories taggedsynthetic data.
Falcon-Emirati-7B Can Actually Speak Emirati Arabic, Not Just Translate It
The Technology Innovation Institute has released a 7-billion-parameter model trained specifically on the Emirati dialect, scoring 84.83% on a purpose-built benchmark where models far larger fall short.
The Best AI Agents Still Fail at Most Real Office Work. ServiceNow Built a Tool to Fix That.
ServiceNow's AutoSynthData spots where an AI agent breaks down on the job, then automatically generates practice tasks to fix the gap. The backdrop: the strongest model tested cleared barely a third of realistic enterprise tasks.
Nvidia's Kumo Tabular Reads a Spreadsheet Cold and Beats Every Tuned Model Doing It
A new open AI model from Nvidia makes predictions on business data tables without any training on your data, no setup, and no specialist required, and it just topped every major leaderboard.
Iceland's Treble Raises $18 Million to Stress-Test the Voice AI Boom
The acoustic simulation startup, which already benchmarks speech models on Hugging Face, now has $40 million total to push into robotics, smart glasses, and physical AI.
A $10 Million Bet on AI That Streams Worlds Forever, Not Just Clips
Visko's Orbis model generates live, interactive 4K video indefinitely, and its Stanford-trained founder says it could supply the training footage that robots headed to your home will need.
AI Training Data Startup Micro1 Hit $500 Million Gross Run Rate in Eight Months
Demand for human-verified AI training data is minting fast-growing startups. Micro1's revenue explosion shows the market is big enough for several players, and the ethical debates are heating up.
Inside the Simulators Teaching Robots to Move, Grip and Learn
Training a real robot costs a fortune and breaks things. A new generation of GPU-powered physics simulators is changing that, one virtual arm at a time.
Apple researchers taught AI to practise using apps without needing a real app to practise on
A new technique lets AI agents learn to make software calls by simulating the whole environment inside another AI, cutting one of the biggest bottlenecks in building helpful AI assistants.
Train the Robot in a Fake World First, Then Send It Into the Real One
Before a robot ever touches a warehouse shelf or a factory floor, it can spend thousands of hours learning inside a computer-built copy of that environment. Here is why that matters.