Statistical and Evaluation Analyst (AI Safety Institute)

The AI Safety Institute is the first state-backed organisation focused on advanced AI safety for the public interest. We launched at the AI Safety Summit because we believe taking responsible action on this extraordinary technology requires a capable and empowered group of technical experts within government.

Our staff includes senior alumni from OpenAI, Google DeepMind, start-ups and the UK government, and ML professors from Oxford and Cambridge. We are now calling on the world’s top technical talent to build the institute from the ground up. This is a truly unique opportunity to help shape AI safety at an international level.

We have ambitious goals and need to move fast.

Develop and conduct evaluations on advanced AI systems
We will characterise safety-relevant capabilities, understand the safety and security of systems, and assess their societal impacts.

Drive foundational AI safety research
We will launch moonshot research projects and convene world-class external researchers.

Facilitate information exchange
We will establish clear information-sharing channels between the Institute and other national and international actors. These include stakeholders such as policymakers and international partners.

Our Inclusive Environment
We are building an inclusive culture to make the Department a brilliant place to work where our people feel valued, have a voice and can be their authentic selves. We value difference and diversity, not only because we believe it is the right thing to do, but because it will help us be more innovative and make better decisions.

What we Value:

• Diverse Perspectives: We believe that a range of experiences and backgrounds is essential to our success. We welcome individuals from underrepresented groups to join us in this crucial mission.
• Collaborative Spirit: We thrive on teamwork and open collaboration, valuing every contribution, big or small.
• Innovation and Impact: We are dedicated to making a real-world difference in the field of frontier AI safety and capability, and we encourage innovative thinking and bold ideas.

Job description

This is an exciting opportunity to carry out analysis in a rapidly developing and high-profile area of Government. The Analysis Team aims to establish a credible evidence base for AISI’s investments in frontier AI, make use of innovative methods to evaluate the impact of the AISI’s ground-breaking initiatives and keep abreast of developments in the frontier AI ecosystem. The team is made up of a Grade 6 Senior Economic Advisor, a G7 Economic Advisor, a G7 Social Researcher, and a HEO Economist. This is a brand-new role, and you’ll have lot of opportunity to shape the work.

A background in AI policy or technology is not essential. The adaptable nature of the team means you should be comfortable with and enjoy applying statistical techniques and research methods to a nascent and ever-changing policy area.

Across the three core missions of the AISI, some of the types of analytical work you may get involved in are:

• Support the Research Unit to set up randomised controlled trials (RCTs), providing crucial statistical advice on the experimental design, defining an appropriate population sample, running the study, and analysing results.
• Advise the Research Unit on the best way to report evaluation study outputs and capture data, so that they can be analysed using various statistical methods
• Support the management of an externally commissioned impact and process evaluation of AISI’s activities.
• Design and manage research into societal harms from frontier AI models, for example around bias and manipulation.

This role will involve working closely with the AISI Research Unit. This is a team of academics, AI research scientists and software engineers, including senior alumni from OpenAI, Google DeepMind, start ups, and machine learning professors from top universities, who conduct model evaluations and other research into the potential risks from frontier AI models. You can find out more about the AISI’s approach to technical model evaluations at the following link: https://www.gov.uk/government/...

Responsibilities  

The successful person will be responsible for but not limited to: 

Provide technical support on research and statistical methods for complex research projects. Within this, you will investigate different options and apply relevant research or statistical skills to do analysis and investigate options.

• Support on the design, commission, and delivery of research and evaluation projects. Within this you will balance timely delivery of findings against the need for reliable, good quality data on which decisions can be made.
• Support on the management of an externally commissioned evaluation, following Magenta Book principles
• Identify solutions to deliver evidence in a fast-paced environment
• Advise on the relative robustness of competing sources of evidence and the interpretation of findings.
• Collaborate effectively with other analytical, research, and programme delivery colleagues within AISI
• Communicate findings of research and evaluation projects, including to non-analysts

Person specification

Who are we looking for?

Your essential experience:

• Experience with advanced research and evaluation methodologies – e.g. randomised control trials, contribution analysis
• Experience with advanced statistical methods and approaches – e.g. multivariate regression, power calculations, interpretation of large complex datasets
• Statistical programme experience – R, STATA or Python are preferred, experience with these or the ability to learn quickly will be essential
• Experience of analysing a range of information and evidence to draw robust and defensible conclusions.
• Ability to clearly communicate data and statistics and their caveats and limitations to analytical and non-analytical audiences (policy officials and external stakeholders) verbally and in writing.
• Ability to manage multiple tasks to tight deadlines

Your desirable experience:

• A broad experience of designing and carrying out and/or commissioning research and/or evaluation with knowledge and experience of using both quantitative and qualitative research methods.
• Experience of using communication and collaboration skills to ensure analysis has an impact on decision-making.
• Experience of managing multiple complex tasks to tight deadlines in a high-pressure environment, with the ability to persuade and influence senior stakeholders.
• Experience working with others to define and deliver tasks that inform objectives, building strong relationships to ensure analysis has a strong and wide impact

Candidates should not be discouraged from applying if they do not have the desirable criteria.

Qualifications

You must be able to demonstrate that a) you satisfy the eligibility criteria and b) have the competence, at SEO grade for one of the analytical professions.

Eligibility requirements of the analytical professions can be found here:

GSR (Government Social Research) For qualification route, must hold an undergraduate degree at a minimum of 2:2. The degree must contain substantial social research methods training (comprising around one-third of modules taken on the course) including quantitative research methods.

For professional experience route, must hold an undergraduate degree at a minimum of 2:2, in any subject, or a degree equivalent (e.g. an advanced Certificate or Diploma from MRS), plus have at least 4 years’ social research practice experience. Social research practice experience consists of experience working in a research agency, market research agency or specialist research team.

GSG (Government Statistical Group) a first or second class honours degree in a numerate subject (2:1 minimum for Assistant Statistician and 2:2 minimum for Statistical Officer) containing at least 25% taught statistical content (e.g. Statistics, Mathematics, Economics, Sciences, Business Studies, Psychology, Geography or similar)

• a higher degree, e.g. a MSc or PhD, in a subject containing formal statistical training (e.g. Statistics, Mathematics, Economics, Sciences, Business Studies, Psychology, Geography or similar)
  
• have achieved or on track to achieve a level 4 apprenticeship including at least 2 years in a statistical/data analysis field in government and are able to demonstrate continuous professional development in statistics/data analysis.
  
• have worked in a statistical or data science field for a minimum of two years and are able to demonstrate Continuous Professional Development (via a logbook) in applying statistics or data science at the same level as a foundation degree or Higher National Diploma (level five), which demonstrates the expected skills as outlined in the GSG Competency framework. Find out more at https://gss.civilservice.gov.u...

Government Economic Service (GES): Must have a first or 2.1 honours bachelors degree (or international equivalent) with at least 50 per cent of economics in the two final years (or equivalent); or a Post Graduate Diploma (or equivalent) or a Masters degree in economics, including a substantial amount of taught macro and microeconomics.

Economic Adviser candidates must demonstrate substantial proven competence in applied economics as a professional economist.

GORS (Government Operational Research Service): A degree at 2.1 (or equivalent) or an MSc in a highly numerate subject, or equivalent experience, and be able to demonstrate relevant experience in your career in a relevant profession.

Apply before 11:55 pm on Sunday 7th April 2024