Software Engineer – Scalable Alignment

Engineering       ·      London, UK       ·     Full time

At DeepMind, we value diversity of experience, knowledge, backgrounds and perspectives, and harness these qualities to create extraordinary impact. We are committed to equal employment opportunity regardless of sex, race, religion or belief, ethnic or national origin, disability, age, citizenship, marital, domestic or civil partnership status, sexual orientation, gender identity, pregnancy, maternity or related condition (including breastfeeding) or any other basis as protected by applicable law. If you have a disability or additional need that requires accommodation, please do not hesitate to let us know.

Snapshot

At DeepMind, we’ve built a unique culture and work environment where long-term ambitious research can flourish. Our special interdisciplinary team combines the best techniques from deep learning, reinforcement learning and systems neuroscience to build general-purpose learning algorithms. We have already made a number of high profile breakthroughs towards building artificial general intelligence, and we have all the ingredients in place to make further significant progress over the coming year!

About us

We’re a dedicated scientific community, committed to “solving intelligence” and ensuring our technology is used for widespread public benefit.

We’ve built a supportive and inclusive environment where collaboration is encouraged and learning is shared freely. We don’t set limits based on what others think is possible or impossible. We drive ourselves and inspire each other to push boundaries and achieve ambitious goals.

We constantly iterate on our workplace experience with the goal of ensuring it encourages a balanced life. From excellent office facilities through to extensive manager support, we strive to support our people and their needs as effectively as possible

Our list of benefits is extensive, and we’re happy to discuss this further throughout the interview process.

The team

The goal of the Scalable Alignment Team (SAT) is to make highly capable agents do what humans want, even when it is difficult for humans to know what that is.  This means we want to remove subtle biases, factual errors, or deceptive behaviour even if they would normally go unnoticed by humans, whether due to reasoning failures or biases in humans or due to very capable behaviour by the agents.

To achieve this, we ask humans what they want and train agents to do that, assisting humans in judgements by providing evidence, outlining arguments, and pointing out subtleties. As language is a key medium for human communication, much of SAT’s work revolves around large language models (LLMs) such as Chinchilla, fine-tuning these models using techniques such as human preference RL, debate, citing evidence, and LM red teaming.  We view LLMs both as a tool for safety by enabling human-machine communication and as examples of ML models that may cause both near-term and long-term harms. Since our goal is to do what humans want, the uncertainties involved are about humans, not just ML; we need to carefully design the interaction between humans and machines to achieve answers humans would endorse after careful reflection.

We view human interaction as only one component of safety, and work with many other teams at DeepMind to build a unified overall strategy, including Alignment, Ethics and Society, and Strategy and Governance.

The Role

Software Engineers in the Scalable Alignment Team work across our software stack to enable researchers in SAT and collaborating teams to make fast progress towards our mission. Software design cannot be disconnected from active research, so this means tight collaboration with Research Scientists and Research engineers on human interaction and model fine-tuning experiments.  Building on significant experience in software architecture and high performance computing, you will share leadership of all SAT software components that influence research progress, including code architecture and design, human processes around maintenance and evolution, testing methodologies, and performance.

Familiarity with large language models and related deep learning algorithms are not a requirement, but you should be excited to learn!

Key responsibilities

• Take the lead and assume responsibility for solving technical problems. You drive multiple stages of software projects – from requirements gatherings to launch and validation. You look for what can be improved and have a proactive attitude in helping others.
• Work well within teams. You participate in technical discussions, ask insightful questions, and work with the team to pull together scoping, schedules, and solutions to issues. You have experience leading and organising collaborative teams.
• Communicate effectively.  You clearly and effectively communicate your ideas, your work, and your results to team members, team leaders and senior managers.

About you

Essential:

• Bachelor’s degree in computer science, a related technical field, or equivalent practical experience.
• Software engineering experience in large, shared codebases, including the human aspects of software engineering.
• Ability to communicate technical ideas effectively, e.g. through discussions, whiteboard sessions, written documentation.
• Experience with either high performance distributed computation or robust, high performance servers.

Nice to have:

• Experience with Python, C++, and Javascript.
• Experience with multi-host distributed systems.
• Knowledge of mathematics, statistics, and machine learning concepts.
• A passion for making AGI go well.

Please click here to apply.