Shahar Avin
YOU?
Author Swipe
View article: The UK’s pandemic preparedness and early response to the COVID-19 pandemic
The UK’s pandemic preparedness and early response to the COVID-19 pandemic Open
This article focuses on the UK's pre-COVID 19 pandemic preparedness and its early response to the COVID-19 pandemic (January '20 - March '20). The aim of this article is to explain the high excess mortality the UK experienced compared to m…
View article: Strategic Insights from Simulation Gaming of AI Race Dynamics
Strategic Insights from Simulation Gaming of AI Race Dynamics Open
We present insights from "Intelligence Rising", a scenario exploration exercise about possible AI futures. Drawing on the experiences of facilitators who have overseen 43 games over a four-year period, we illuminate recurring patterns, str…
View article: 3. Classifying Global Catastrophic Risk
3. Classifying Global Catastrophic Risk Open
Identifying a lack of satisfactory methodologies for classifying Global Catastrophic Risk (GCR) scenarios, this chapter proposes a new classification framework which draws together multiple areas of knowledge. The suggested framework categ…
View article: 9. Accumulating Evidence Using Crowdsourcing and Machine Learning
9. Accumulating Evidence Using Crowdsourcing and Machine Learning Open
Focusing on the study of existential risk, this chapter shows how academic research publications can be systematically reviewed using semi-automated processes such as crowdsourcing and machine learning. These publications can then be accum…
View article: 8. Exploring Artificial Intelligence Futures
8. Exploring Artificial Intelligence Futures Open
This chapter presents a survey of the different methods available for the exploration of Artificial Intelligence futures, from fictional narratives to integrative, interdisciplinary, and participatory methods of exploring AI futures. Exami…
View article: Computing Power and the Governance of Artificial Intelligence
Computing Power and the Governance of Artificial Intelligence Open
Computing power, or "compute," is crucial for the development and deployment of artificial intelligence (AI) capabilities. As a result, governments and companies have started to leverage compute as a means to govern AI. For example, govern…
View article: Lessons from COVID-19 for GCR governance: a research agenda
Lessons from COVID-19 for GCR governance: a research agenda Open
The Lessons from Covid-19 Research Agenda offers a structure to study the COVID-19 pandemic and the pandemic response from a Global Catastrophic Risk (GCR) perspective. The agenda sets out the aims of our study, which is to investigate the…
View article: AI Systems of Concern
AI Systems of Concern Open
Concerns around future dangers from advanced AI often centre on systems hypothesised to have intrinsic characteristics such as agent-like behaviour, strategic awareness, and long-range planning. We label this cluster of characteristics as …
View article: Frontier AI Regulation: Managing Emerging Risks to Public Safety
Frontier AI Regulation: Managing Emerging Risks to Public Safety Open
Advanced AI models hold the promise of tremendous benefits for humanity, but society needs to proactively manage the accompanying risks. In this paper, we focus on what we term "frontier AI" models: highly capable foundation models that co…
View article: Model evaluation for extreme risks
Model evaluation for extreme risks Open
Current approaches to building general-purpose AI systems tend to produce systems with both beneficial and harmful capabilities. Further progress in AI development could lead to capabilities that pose extreme risks, such as offensive cyber…
View article: Rewards, risks and responsible deployment of artificial intelligence in water systems
Rewards, risks and responsible deployment of artificial intelligence in water systems Open
Artificial intelligence (AI) is increasingly proposed to address deficiencies across water systems, which currently leave about 25% of the global population without clean water, about 50% without sanitation services and about 30% without h…
View article: Lessons from COVID-19 for GCR governance: a research agenda
Lessons from COVID-19 for GCR governance: a research agenda Open
The Lessons from Covid-19 Research Agenda offers a structure to study the COVID-19 pandemic and the pandemic response from a Global Catastrophic Risk (GCR) perspective. The agenda sets out the aims of our study, which is to investigate the…
View article: Filling gaps in trustworthy development of AI
Filling gaps in trustworthy development of AI Open
Incident sharing, auditing, and other concrete mechanisms could help verify the trustworthiness of actors
View article: Foresight for unknown, long-term and emerging risks, Approaches and Recommendations
Foresight for unknown, long-term and emerging risks, Approaches and Recommendations Open
Foresight approaches have played an increasing role in our efforts to understand and develop ways of addressing global risks. They enable us to engage a range of expertise in exploring possible future scenarios, technological and societal …
View article: Submission of Evidence to The House of Lords Select Committee on Risk Assessment and Risk Planning
Submission of Evidence to The House of Lords Select Committee on Risk Assessment and Risk Planning Open
In this submission we: (1) introduce our approach to defining and classifying extreme risks, including global catastrophic risks and existential risks (addressing question 1: "What do you understand the term ‘extreme risk’ to mean?"); (2) …
View article: Tackling threats to informed decision-making in democratic societies: Promoting epistemic security in a technologically-advanced world
Tackling threats to informed decision-making in democratic societies: Promoting epistemic security in a technologically-advanced world Open
Access to reliable information is crucial to the ability of a democratic society to coordinate effective collective action when responding to a crisis, like a global pandemic, or complex challenge like climate change. Through a series of w…
View article: Written Evidence - Defence industrial policy: procurement and prosperity
Written Evidence - Defence industrial policy: procurement and prosperity Open
In this response we particularly focus on defence and those in adjacent markets systems that integrate increasingly capable artificial intelligence (AI), especially those based on machine learning (ML). Many systems that the Ministry of De…
View article: An agent-based model clarifies the importance of functional and developmental integration in shaping brain evolution
An agent-based model clarifies the importance of functional and developmental integration in shaping brain evolution Open
Comparisons of vertebrate brain structure suggest a conserved pattern of scaling between components, but also many examples of lineages diverging dramatically from these general trends. Two competing hypotheses of brain evolution seek to e…
View article: Toward Trustworthy AI Development: Mechanisms for Supporting Verifiable Claims
Toward Trustworthy AI Development: Mechanisms for Supporting Verifiable Claims Open
With the recent wave of progress in artificial intelligence (AI) has come a growing awareness of the large-scale impacts of AI systems, and recognition that existing regulations and norms in industry and academia are insufficient to ensure…
View article: AI Paradigms and AI Safety: Mapping Artefacts and Techniques to Safety Issues
AI Paradigms and AI Safety: Mapping Artefacts and Techniques to Safety Issues Open
AI safety often analyses a risk or safety issue, such as interruptibility, under a particular AI paradigm, such as reinforcement learning. But what is an AI paradigm and how does it affect the understanding and implications of the safety i…
View article: Accumulating evidence using crowdsourcing and machine learning: A living bibliography about existential risk and global catastrophic risk
Accumulating evidence using crowdsourcing and machine learning: A living bibliography about existential risk and global catastrophic risk Open
The study of existential risk — the risk of human extinction or the collapse of human civilization — has only recently emerged as an integrated field of research, and yet an overwhelming volume of relevant research has already been publish…
View article: Autonomy and machine learning at the interface of nuclear weapons, computers and people
Autonomy and machine learning at the interface of nuclear weapons, computers and people Open
A new era for our species started in 1945: with the terrifying demonstration of the power of the atom bomb in Hiroshima and Nagasaki, Japan, the potential global catastrophic consequences of human technology could no longer be ignored. Wit…
View article: Method Pluralism, Method Mismatch & Method Bias
Method Pluralism, Method Mismatch & Method Bias Open
Pluralism about scientific method is more-or-less accepted, but the consequences have yet to be drawn out. Scientists adopt different methods in response to different epistemic situations: depending on the system they are interested in, th…
View article: Exploring artificial intelligence futures
Exploring artificial intelligence futures Open
Artificial intelligence technologies are receiving high levels of attention and ‘hype’, leading to a range of speculation about futures in which such technologies, and their successors, are commonly deployed. By looking at existing AI futu…
View article: Surveying Safety-relevant AI Characteristics
Surveying Safety-relevant AI Characteristics Open
[EN] The current analysis in the AI safety literature usually combines a risk or safety issue (e.g., interruptibility) with a particular paradigm for an AI agent (e.g., reinforcement learning).
\nHowever, there is currently no survey of sa…
View article: Mavericks and lotteries
Mavericks and lotteries Open
In 2013 the Health Research Council of New Zealand began a stream of funding titled 'Explorer Grants', and in 2017 changes were introduced to the funding mechanisms of the Volkswagen Foundation 'Experiment!' and the New Zealand Science for…
View article: Between Progress and Potential Impact of AI: the Neglected Dimensions
Between Progress and Potential Impact of AI: the Neglected Dimensions Open
We reframe the analysis of progress in AI by incorporating into an overall framework both the task performance of a system, and the time and resource costs incurred in the development and deployment of the system. These costs include: data…