All posts
Showing 12 of 1407 articles · Page 9 of 118
Coldcard Hardware Wallet Vulnerability Exposed Millions in Stolen Bitcoin
Coldcard hardware wallet flaw allowed theft of 1,158 BTC from 2,673 addresses. Update firmware and move funds immediately.
SanDisk and SK Hynix Partner on GPUs With Over 3TB of Memory Using High-Bandwidth Flash
SanDisk and SK Hynix develop High-Bandwidth Flash memory combining NAND and HBM to enable cheaper, more efficient AI inference with massive models.
GitHub Launches Stacked Pull Requests to Simplify Large Code Changes
GitHub launches Stacked Pull Requests in public preview, enabling developers to organize large code changes into layered, reviewable sequences.
CosmosEscape: Critical Azure Cosmos DB Vulnerability Allowed Unauthorized Database Access
CosmosEscape vulnerability in Azure Cosmos DB allowed unauthorized database access. Microsoft has patched the flaw.
AI Analyzes 3,000 Dream Reports to Reveal How the Brain Reconstructs Reality
AI analysis of 3,000 dream reports reveals how the brain combines memories and emotions to create dreams, not random noise.
Ferrari's First Electric Car Exceeds Annual Sales Target in Just Two Months
Ferrari's Luce electric car exceeds annual sales target in two months despite negative reception to its design by Jony Ive and Marc Newson.
Malicious Instructions in Word Documents Can Hijack Microsoft Copilot
Malicious instructions in Word docs can hijack Copilot and alter files. Security researcher reveals vulnerability Microsoft delayed disclosing.
Companies Ditch Expensive AI Models as 'Tokenmaxxing' Drives Costs Up
Companies seek cheaper AI models as tokenmaxxing inflates costs. Mozilla's Krikorian warns the metric repeats productivity measurement mistakes.
GitHub Strengthens npm and GitHub Actions Security to Block Supply Chain Attacks
GitHub rolls out security upgrades to npm and GitHub Actions to block supply chain attacks on maintainer accounts and package installations.
AI Labs Are Bulk-Buying and Destroying Rare Books for Training Data
AI labs like Anthropic are buying rare books in bulk, digitizing them, and destroying originals for LLM training data.
Sam Altman Backs Slower AI Development Pace
Sam Altman now advocates for slower AI development to allow society time to adapt, but faces challenges avoiding collusion concerns.
Nvidia Leads Open Secure AI Alliance With 37 Companies to Combat Cyber Threats
Nvidia launches Open Secure AI Alliance with 37 companies to create open-source cybersecurity tools for AI systems.