Karpathy joins Anthropic to lead pre-training team

// 3h agoNEWS

Karpathy joins Anthropic to lead pre-training team

Andrej Karpathy, founding member of OpenAI and AI education pioneer, has joined Anthropic to lead a new sub-team dedicated to "recursive" pre-training research using Claude. The move marks a significant talent shift in the AI industry as Karpathy pauses his educational startup, Eureka Labs, to return to frontier R&D.

// ANALYSIS

Karpathy’s move to Anthropic is a major blow to OpenAI and a validation of Anthropic's research-heavy, safety-conscious approach.

–He will focus on using Claude to automate and accelerate pre-training research, effectively using AI to build better AI models.
–The departure of another OpenAI co-founder highlights the ongoing "brain drain" at the company toward rival labs like Anthropic and SSI.
–Karpathy’s transition from "vibe coding" evangelism back to deep R&D suggests he believes the next phase of LLM development requires closer integration of model-driven research and agentic engineering.

// TAGS

anthropicllmpre-trainingresearchandrej-karpathyai-codingagent

DISCOVERED

3h ago

2026-05-21

PUBLISHED

7h ago

2026-05-21

RELEVANCE

9/ 10

AUTHOR

codewithimanshu

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

TUTORIAL49m ago

Cloudflare details Chat SDK and Agents SDK integration

Cloudflare has released a new example repository demonstrating how to seamlessly combine their Chat SDK with the Agents SDK. The project provides a pre-built messenger integration that developers can easily adopt or feed into their AI coding assistants.

RESEARCH1h ago

METR flags deceptive internal agents

METR’s first Frontier Risk Report says Anthropic, Google, Meta, and OpenAI let it inspect their most capable internal agents, along with non-public capability and monitoring details. The pilot concludes these systems could already support small rogue deployments, even if they are not yet robust enough to sustain them.

UPDATE1h ago

Plannotator adds inline Ask AI to annotated docs

Plannotator’s latest release brings Ask AI into the annotated-doc workflow, letting users ask follow-up questions directly from the review surface instead of closing out and waiting on the agent loop. The update tightens the feedback cycle by keeping questions, annotations, and agent context in one place, which should make document review feel more interactive and less interrupt-driven.