Data
$397 to $13 million: the public record of software prices
PriceProof turns USAspending awards into a checkable record of what US agencies paid for enterprise software, from Tableau at $397 to a $13 million contract.
Data
10 stories on data from the Data Today newsroom.
Data
Border surveillance towers missed 1,050 deaths in plain view
Border surveillance AI failed to prevent over 1,050 deaths within tower range from 2015 to 2026. MIT Tech Review mapped 4,000 remains against 600 towers, exposing detection gaps in the $6.2 billion virtual wall.
Data
TikTok mobile API scrape: 4.5 billion videos open-sourced
A reverse-engineered TikTok mobile API scraped 5.94 billion videos and 3.23 billion profiles in three weeks. 4.5 billion video records are now on Hugging Face as 289 GB of open data.
Data
Agentic MCP attacks bypass SOTA guardrails 58% of the time
MCP attack chains bypass SOTA guardrails more than half the time because text classifiers miss composed tool-call exploits. The agentic safety gap is architectural, not a tuning problem.
Data
MIRA world model simulates Rocket League without a physics engine
From pixels and actions alone, the 5B-parameter MIRA world model simulates Rocket League with stable infinite rollouts at 20 fps on one B200.
Data
How long would it take to brute force your password with today's computing power?
A random 12-character password can outlast every computer on Earth, but an 8-character one on a weak hash falls in hours. Here is the brute-force math, mapped.
Data
What an AI prompt really costs in energy, water and CO2
A typical AI text prompt uses roughly 0.24 to 0.34 watt-hours and a fraction of a milliliter of water, not a bottle per email. Training is huge but rare, and inference is what adds up.
Data
EVA-Bench Data 2.0 makes voice agents measurable
EVA-Bench Data 2.0 is a 213 scenario dataset for testing enterprise voice agents across airline, IT, and healthcare HR workflows.
Data
Three-quarters of the world's AI compute sits in one country
The United States holds about 75 percent of global GPU cluster performance. That concentration shapes pricing, latency, and policy for everyone building elsewhere.
Data
Open-weight models closed the gap to 1.7 points
Open-weight models went from a budget compromise to a default choice. On some benchmarks the gap to the best closed models shrank from 8 points to 1.7 in a year.