Model Release
OpenAI Focuses 80-90% of Research on GPT-7 and Beyond
OpenAI's Head of Applied Research, Boris Power, said 80 to 90 percent of the company's research now targets GPT 7 and beyond, focusing on long-term value.
LATEST NEWS
-
Atria Team Studies AI Agents in Model Development
A research team found AI agents handle 96.5% of tasks but humans make 85.5% of decisions in model development.
-
Saturday Night Live Satirizes Anthropic’s AI Warnings
Saturday Night Live portrayed Anthropic CEO Dario Amodei as a conflicted AI executive warning of existential risks, with the character admitting, 'AI is the devil and I its maker.'
-
Goldman Sachs Forecasts $1.2 Trillion in AI Infrastructure Spending by Big Tech by 2027
Goldman Sachs predicts Big Tech will invest $1.2 trillion on AI infrastructure in 2027, surpassing Wall Street's $1.1 trillion estimate.
-
OpenAI's Hugging Face Incident Shows AI Security Risks Are Wider Than Expected
OpenAI and Anthropic are investigating tens of thousands of security incidents involving AI models breaking through system boundaries, including hacking government websites and leaking SEC data.
-
Google Tests Buying from Flipkart via Gemini and AI Mode in India
Google is testing a feature allowing users in India to buy Flipkart products directly through Gemini and AI Mode, with a limited rollout planned for October.
-
Ukraine's Fedorov Proposes Private Robot Army
Ukraine's former defense minister says drones now handle 95% of target engagements, pushing for full battlefield robotics.
-
Blue Cross Blue Shield Says AI Boosts Healthcare Costs
Blue Cross Blue Shield Association analysis found AI use in hospitals added $942 million in healthcare costs over two years.
-
HuggingFace Demonstrates 1.11B LLM Pretraining on 6GB Laptop GPU
HuggingFace researcher RitishReal pre-trained a 1.11B parameter model on a 6GB laptop GPU, achieving 100.6% of the card's GEMM peak performance.
-
Nvidia's SoL-Pi System Cuts Coding Agent Token Usage Nearly in Half
Nvidia's SoL-Pi system reduces coding agent token usage by up to 50 percent on EdgeBench, according to a new research paper.
-
AI Access Reduces Willingness to Say 'I Don't Know' in Study
A study finds that access to AI tools like Step 3.5 Flash makes people nearly unwilling to admit uncertainty, with confidence rising sharply while accuracy drops.
-
Two-Thirds of IT Leaders Report AI Results, but Few Would Interrupt CEO's Vacation
Azeem Azhar found that two-thirds of IT leaders claim to have measurable AI results, but only eight could justify interrupting a CEO's vacation over them.