Connor Leahy - Wikipedia

reference

Wikipedia·en.wikipedia.org/wiki/Connor_Leahy

Credibility Rating

3/5

Good(3)

Good quality. Reputable source with community review or editorial standards, but less rigorous than peer-reviewed venues.

Rating inherited from publication venue: Wikipedia

Useful background reference for understanding a key figure in the AI safety ecosystem, particularly relevant when reading about EleutherAI, Conjecture, or AI governance advocacy efforts.

Metadata

Importance: 30/100wiki pagereference

Summary

Wikipedia biography of Connor Leahy, co-founder of EleutherAI and CEO of Conjecture, who is a prominent AI safety researcher and advocate. Leahy is known for his work on open-source large language models and his public advocacy for AI safety and existential risk from advanced AI systems.

Key Points

•Co-founded EleutherAI, an open-source AI research collective known for GPT-Neo and other large language models
•Founded Conjecture, an AI safety company focused on alignment research and making AI systems safer
•Prominent public voice on AI existential risk, frequently appearing in media and policy discussions
•Advocates for strong AI governance and has testified to policymakers about AI safety concerns
•Represents a bridge between open-source AI development and AI safety advocacy communities

1 FactBase fact citing this source

Entity	Property	Value	As Of
Connor Leahy	Notable For	CEO of Conjecture; co-founder of EleutherAI; prominent AI safety advocate; testified before UK Parliament and EU on AI risks	—

Cached Content Preview

HTTP 200Fetched May 10, 20267 KB

# Connor Leahy

Connor Leahy

German-American artificial intelligence researcher

**Connor Leahy** is a German-American[\[1\]](https://en.wikipedia.org/wiki/Connor_Leahy#cite_note-1) [artificial intelligence](https://en.wikipedia.org/wiki/Artificial_intelligence "Artificial intelligence") [researcher](https://en.wikipedia.org/wiki/Research "Research") and [entrepreneur](https://en.wikipedia.org/wiki/Entrepreneurship "Entrepreneurship") known for cofounding [EleutherAI](https://en.wikipedia.org/wiki/EleutherAI "EleutherAI")[\[2\]](https://en.wikipedia.org/wiki/Connor_Leahy#cite_note-:1-2)[\[3\]](https://en.wikipedia.org/wiki/Connor_Leahy#cite_note-:2-3) and being CEO of [AI safety](https://en.wikipedia.org/wiki/AI_safety "AI safety") research company Conjecture.[\[4\]](https://en.wikipedia.org/wiki/Connor_Leahy#cite_note-:0-4)[\[5\]](https://en.wikipedia.org/wiki/Connor_Leahy#cite_note-5)[\[6\]](https://en.wikipedia.org/wiki/Connor_Leahy#cite_note-6) He has warned of the [existential risk from artificial general intelligence](https://en.wikipedia.org/wiki/Existential_risk_from_artificial_general_intelligence "Existential risk from artificial general intelligence"), and has called for regulation such as "a moratorium on frontier AI runs" implemented through a cap on compute.[\[7\]](https://en.wikipedia.org/wiki/Connor_Leahy#cite_note-7)

## Career

In 2019, Leahy reverse-engineered [GPT-2](https://en.wikipedia.org/wiki/GPT-2 "GPT-2") in his bedroom, and later co-founded [EleutherAI](https://en.wikipedia.org/wiki/EleutherAI "EleutherAI") to attempt to replicate [GPT-3](https://en.wikipedia.org/wiki/GPT-3 "GPT-3").[\[2\]](https://en.wikipedia.org/wiki/Connor_Leahy#cite_note-:1-2)

Leahy is skeptical of [reinforcement learning from human feedback](https://en.wikipedia.org/wiki/Reinforcement_learning_from_human_feedback "Reinforcement learning from human feedback") as a solution to the [alignment problem](https://en.wikipedia.org/wiki/Alignment_problem "Alignment problem"). “These systems, as they become more powerful, are not becoming less alien. If anything, we’re putting a nice little mask on them with a smiley face. If you don't push it too far, the smiley face stays on. But then you give it \[an unexpected\] prompt, and suddenly you see this massive underbelly of insanity, of weird thought processes and clearly non-human understanding.”[\[8\]](https://en.wikipedia.org/wiki/Connor_Leahy#cite_note-8)

He was one of the signatories of the 2023 [open letter](https://en.wikipedia.org/wiki/Pause_Giant_AI_Experiments:_An_Open_Letter "Pause Giant AI Experiments: An Open Letter") from the [Future of Life Institute](https://en.wikipedia.org/wiki/Future_of_Life_Institute "Future of Life Institute") calling for "all AI labs to immediately pause for at least 6 months the training of AI systems more powerful than GPT-4."[\[9\]](https://en.wikipedia.org/wiki/Connor_Leahy#cite_note-9)[\[10\]](https://en.wikipedia.org/wiki/Connor_Leahy#cite_note-10)

In November 2

... (truncated, 7 KB total)

Resource ID: kb-bdf7458184e1e789 | Stable ID: sid_c5a4gZFbYa