News archive.
Earlier AI news and analysis, kept for reference. Current coverage lives in News & Analysis.
Founder stops hiring junior engineers as AI coding tools cut junior roles
Eugenia Kuyda, CEO of Replika and Wabi, says AI has made junior engineer hires 'extremely expensive and completely unsustainable' — and she's restructuring her company around it.
Agents Can Now Search for Tools at Runtime Instead of Pre-Installing Them
Hugging Face launches Agentic Resource Discovery, a new open standard that lets agents find skills, MCP servers, and other capabilities on the fly. Available now via CLI, REST API, and MCP protocol.
Allen AI releases MolmoMotion to predict 3D object paths from video
MolmoMotion forecasts where objects will move in 3D space given a video frame, query points, and an action description. The model outperforms existing methods on a new 2.7K-clip benchmark and improves robot manipulation planning accuracy by 20 percentage points.
ChatGPT now runs tasks on a schedule. Here's what that means for your workflow
OpenAI added scheduled task automation to ChatGPT, letting you set conversations to run at specific times. We break down what the feature does, who needs it, and how to set it up.
Anthropic opens Seoul office as AI labs battle for Korea
Anthropic is establishing a Seoul office amid intensifying competition for talent and market share in South Korea's AI sector. Here's what the move signals about regional strategy.
OpenAI's LifeSciBench grades AI on 750 real research tasks
OpenAI released LifeSciBench, a 750-task benchmark using expert-written rubrics to evaluate how well AI models perform on actual life-science research work. Here's what the benchmark measures and why it matters for biotech teams.
OpenAI Deploys GPT-4 and o1 with Partners via Dedicated Field Engineers
OpenAI is rolling out frontier models to enterprise customers through a network of field deployment engineers and strategic partners. Here's what's shipping now and what it means for your stack.
OpenAI releases LifeSciBench to test AI on biology tasks
OpenAI introduced LifeSciBench, a benchmark for evaluating frontier models on life science research problems. The tool measures performance across biology-focused tasks to help teams assess model capabilities in the field.
Anthropic Opens Seoul Office, Signs Korean AI Partnerships
Anthropic is establishing a physical presence in South Korea and partnering with local AI companies. The move signals expansion beyond the US market.
OpenAI releases LifeSciBench for evaluating AI on real research tasks
OpenAI introduced LifeSciBench, a benchmark authored and reviewed by life science experts to assess how AI systems perform on actual research decisions. Here's what it measures and why it matters for biotech teams.
OpenAI's GPT-5.4 improved a stubborn drug-making reaction with near-autonomous chemistry
OpenAI and Molecule.one deployed GPT-5.4 to optimize a key medicinal chemistry reaction. The system worked with minimal human direction—signaling a shift in how AI can assist drug discovery workflows.
Google's AMIE AI Matches Primary Care Doctors on Disease Management
Nature study shows Google's medical AI system scored higher than human doctors on treatment plan precision and clinical guideline adherence. Google is now testing AMIE in real clinical settings.