Match score not available

Technical Governance and Responsible AI Researcher, Cohere For AI

Remote:

Hybrid

Contract:

Full time

Work from:

San Francisco (US), London (GB), New York (US), Toronto (CA)

Offer summary

Qualifications:

Computer science or economics degree, Strong scientific background and rigor, Experience in governance topics, Comfort in translating technical concepts.

Key responsabilities:

Conduct technical governance research and contribute to reports
Run experiments to improve understanding of AI

Cohere Scaleup https://cohere.com/

51 - 200 Employees

See more Cohere offers

Job description

Why this role?

Cohere For AI is a state-of-the-art research lab that sets progress at the frontier of machine learning. We have a proven track record of top-tier publications and a commitment to cross-institutional high-impact collaboration. Learn more about our work here.

In addition to fundamental research, we contribute technical perspectives on questions at the core of AI development. This involves combining tools across fields, such as computer science economics and deep learning, to provide rigorous landscape analysis on open questions about responsible development in AI. The goal of this role is to help cultivate a better understanding of AI futures, and to support effective AI policy.

Our team works to provide a technically grounded perspective to shape advanced recommendations for the governance of artificial intelligence, informing how we can improve risk identification and mitigation throughout the model development and deployment lifecycle.

Please Note: We have offices in Toronto, San Francisco, New York, and London but embrace being remote-first! There are no restrictions on where you can be located for this role.

As a technical governance researcher, some of your core responsibilities will include:

Executing on technical governance research, contributing to reports that provide technical perspectives on the state of AI.

Run and analyze scientific experiments to advance our understanding of large language models and the ecosystem which large language models exist in.

Build models, collect data and interpret results about specific aspects relating to the the state of AI, including varied questions like: inference time compute, traceability of open weights and watermarking, parsing evidence to-date to motivate prioritization of risks.

Launch initiatives to steward public awareness of AI capabilities and their limitations through whitepapers, guides, workshops, gatherings, and courses.

Communicate scientific findings with a wide audience that includes policymakers.

Some examples of past technical governance projects that have come out of Cohere For AI:

Understanding of the viability of compute thresholds

Assessing AI Biorisk - An evidence based assessement on prioritization of Biorisk

Policy Primer - The AI Language Gap

Open Problems in Technical AI Governance

The Data Provenance Initiative: A Large Scale Audit of Dataset Licensing & Attribution in AI

You may be a good fit if you have:

Technical contributions to governance topics, evidencing rigour and thorough evaluation.

Executed experiments with rigour and have a strong scientific background. This role will require comfort in translating technical concepts into policy recommendations and discussing technical concepts with a general audience.

An understanding of global public policy issues particularly with respect to AI policy.

Ability to identify and analyze public policy developments, put them into context and identify scientific problems underlying policy discourse.

Computer science degree, economics degree or equivalent technical experience.