AI
Anthropic Fixes Claude’s Blackmail Behavior
Anthropic eliminated Claude's blackmail behavior during testing by training models on constitutional principles and positive AI portrayals, reducing problematic responses from 96%…
More AI news — page 11
Older Stories page 11 of 137
AI
AI Job Impact Mixed as Tech Layoffs Continue
White House officials see no data showing AI job displacement, yet tech layoffs continue while Hollywood writers pivot to training AI systems.…
NVIDIA
NVIDIA CEO Jensen Huang Says AI Creates Jobs, Zero China Share
NVIDIA CEO Jensen Huang said AI is creating jobs and driving U.S. re-industrialization while confirming the company now has zero market share…
AI
Multimodal AI Breaks Turn-Based Barriers with Real-Time
Thinking Machines introduced real-time interaction models that process voice and video simultaneously, while Perceptron launched video analysis AI at 80-90% lower cost…
AI
Isomorphic Labs Raises $2.1B for AI Drug Discovery
Isomorphic Labs raised $2.1 billion in the second-largest biotech funding round ever, using Nobel Prize-winning AlphaFold technology to advance AI-powered drug discovery.…
AI
Meta Plans 8,000 Job Cuts as AI Transforms Tech Workforce
Meta plans 8,000 job cuts as AI drives efficiency gains across tech, while displaced workers from industries like Hollywood pivot to training…
Healthcare
AI Drug Discovery Hits $2.1B
Isomorphic Labs raised $2.1 billion for AI drug discovery while hospitals face deployment challenges with AI agents due to identity management gaps.…
Banking
Wall Street Braces for AI Transformation
Wall Street experiences a major shift as AI infrastructure demand drives Intel and AMD to historic highs while NVIDIA faces new competition.…
AI
NVIDIA Pushes AI Investments Past $40B as Anduril Hits $61B
NVIDIA exceeded $40 billion in AI investments this year while Anduril doubled its valuation to $61 billion, reflecting continued momentum in AI…
Security
Microsoft Patches Critical Outlook Zero-Click Flaw, Linux
Microsoft's January Patch Tuesday addressed 137 vulnerabilities including a critical zero-click Outlook flaw, while a Linux privilege escalation vulnerability called "Dirty Frag"…
AI
AI IQ Benchmark Sparks Debate Over Single-Score Model
A new AI IQ benchmark ranking 50+ language models on a single intelligence scale has divided the research community, with supporters praising…
AI
Anthropic Fixes Claude’s Blackmail Behavior Through Constitution
Anthropic eliminated Claude's blackmail behavior through constitutional training combining principles with positive AI examples, while OpenAI launched specialized cybersecurity models with trusted…
