Vidur Kapur

Quick Assessment

Overview

Vidur KapurPersonVidur KapurVidur Kapur is a superforecaster and AI policy researcher involved in multiple forecasting organizations and the Sentinel early warning system, contributing to AI risk assessment and EA Forum discu...Quality: 38/100 is a superforecaster and AI policy researcher affiliated with multiple forecasting organizations including Good JudgmentOrganizationGood Judgment (Forecasting)Good Judgment Inc. is a commercial forecasting organization that emerged from successful IARPA research, demonstrating that trained 'superforecasters' can outperform intelligence analysts and predi...Quality: 50/100, Swift CentreOrganizationSwift CentreSwift Centre is a UK forecasting organization that provides conditional forecasting services to various clients including some AI companies, but is not primarily focused on AI safety. While they de...Quality: 50/100, SamotsvetyOrganizationSamotsvetyElite forecasting group Samotsvety dominated INFER competitions 2020-2022 with relative Brier scores twice as good as competitors, providing influential probabilistic forecasts including 28% TAI by...Quality: 61/100, and RAND, while also working as an AI Policy Researcher at ControlAIOrganizationControlAIControlAI is a UK-based advocacy organization that has achieved notable policy engagement success (briefing 150+ lawmakers, securing support from 100+ UK parliamentarians) while promoting direct in...Quality: 63/100¹. His work focuses on existential risks from AI, biological threats, and geopolitical instability, with particular emphasis on early detection and rapid response to global catastrophes.

Kapur is a key member of the SentinelOrganizationSentinel (Catastrophic Risk Foresight)Sentinel is a catastrophic risk foresight organization co-founded by Nuño Sempere and Rai Sur, operating informally from around 2023 and incorporated as a US 501(c)(3) nonprofit in or after late 20...Quality: 39/100 team, a project dedicated to early detection and response for global catastrophes². The project emphasizes rapid foresight on timescales of days to weeks, positioned as increasingly critical as AI capabilities integrate into society and potential dangers escalate. His forecasting work spans AI timelines, catastrophic risk scenarios, and geopolitical events, contributing to organizations that collectively shape risk assessment in the effective altruism and AI safety communities.

Beyond forecasting, Kapur maintains an active presence on the Effective Altruism Forum, where he contributes posts and commentary on utilitarianism, political bias, ethics, and cause prioritization³. His dual focus on technical forecasting and philosophical engagement positions him as a bridge between quantitative risk assessment and broader EA community discussions.

Forecasting Work

Superforecasting Organizations

Kapur works with several prominent forecasting platforms and organizations. As a superforecaster with Good Judgment, he contributes to probabilistic forecasting on geopolitical and technological developments⁴. His work with Samotsvety, a group of elite forecasters, has included high-stakes AI risk assessments. Notably, Samotsvety estimated a 30.5% risk of AI catastrophe killing the vast majority of humanity by 2200⁵.

Sentinel Project

The Sentinel project represents Kapur's most prominent contribution to existential risk mitigation. Sentinel focuses on early detection and response for global catastrophes, emphasizing rapid observation-orientation cycles to identify emerging threats⁶. The project improved content quality and distribution following a November 2024 fundraise, with Kapur listed as a key team member⁷.

Sentinel's work includes forecasting AI developments and geopolitical risks. For example, Kapur participated in Sentinel Minutes podcast episodes discussing OpenAIOrganizationOpenAIComprehensive organizational profile of OpenAI documenting evolution from 2015 non-profit to Public Benefit Corporation, with detailed analysis of governance crisis, 2024-2025 ownership restructuri...Quality: 62/100's plans for an automated research intern by September 2026 and a true AI researcher by March 2028⁸. He has also contributed forecasts on drone attacks, regime change scenarios (such as Iranian political instability), and AI model capability timelines⁹.

In a podcast on Iranian regime change, Kapur provided probability estimates of 10% by the end of August and 15% by year-end, analyzing base rates, internal and external factors, economic suffering, corruption, and public disillusionment¹⁰. This work exemplifies Sentinel's approach of combining geopolitical analysis with structured forecasting methodologies.

AI Safety and Existential Risk

Kapur's engagement with AI safety encompasses both technical forecasting and conceptual analysis of catastrophic scenarios. He provided comments on a draft post ranking AI existential risk scenarios by "embarrassment level," discussing how predictive models might enable catastrophic terrorist attacks due to insufficient safeguards and offense-biased scientific understanding¹¹. This commentary reflects concern about dual-use AI capabilities and the asymmetry between offensive and defensive applications.

His forecasting work directly addresses AI timelines and capability milestones. Sentinel's projections on automated AI researchers and model doubling times contribute to the broader AI safety community's understanding of when transformative AI capabilities might emerge¹². Kapur's role bridges the gap between abstract x-risk scenarios and concrete near-term forecasts that can inform policy and research prioritization.

Effective Altruism Forum Contributions

Kapur maintains an active profile on the EA Forum, contributing posts, comments, and "Quick Takes" on various topics¹³. His contributions often challenge conventional EA thinking or introduce methodological refinements.

"EAs are not perfect utilitarians"

In a notable post, Kapur argued that EA participants have diverse motivations beyond pure utilitarianism, including prestige, loyalty, novelty-seeking, and burnout avoidance¹⁴. He urged caution against over-rationalizing non-utilitarian choices, suggesting that acknowledging human limitations leads to more realistic and sustainable engagement with effective altruism. Community responses noted this aligns with utilitarian self-awareness, as perfect utilitarians would account for human cognitive and motivational constraints.

Political Debiasing

Kapur co-authored a post on political debiasing and the Political Bias Test, addressing criticisms such as ceiling effects and the ease of gaming bias assessments¹⁵. The work discussed pre-tests on Amazon Mechanical Turk and strategies for inferring bias from belief patterns. This methodological contribution reflects his interest in improving how the EA community identifies and corrects for cognitive biases.

Other Engagements

Kapur has commented on EA's public perception, noting that while the community welcomes internal criticism, this may not be apparent to outsiders¹⁶. He has also engaged in discussions on estimating future human value, recommending analyses by Michael Dickens and What We Owe The Future, while acknowledging uncertainties like s-risks and the trajectory of human altruism¹⁷. His "Quick Takes" include critiques of evidence quality in animal welfare interventions, such as questioning evidence on electric stunning for shrimp welfare¹⁸.

Community Reception

Within the EA Forum, Kapur is viewed as a constructive contributor whose posts spark discussion on practical EA limitations and methodological improvements¹⁹. His work on political debiasing has drawn methodological feedback, with commenters refining ideas around gaming effects and pre-test design²⁰. No major controversies or criticisms have been noted; interactions focus on collaborative refinement of ideas rather than adversarial debate.

His forecasting work, particularly with Sentinel, has been positively received in EA contexts focused on existential risk. The project's emphasis on rapid foresight aligns with growing concerns about AI acceleration and the need for early warning systems.

Key Uncertainties

Several aspects of Kapur's work and influence remain unclear from available sources:

• Quantitative impact: Specific forecasting track records, accuracy metrics, or Brier scores are not publicly documented, making it difficult to assess predictive performance relative to other superforecasters.
• Policy influence: The extent to which Kapur's forecasts inform actual policy decisions or institutional strategies at organizations like RAND is not specified.
• Relationship to other forecasters: Collaborative dynamics with other prominent superforecasters (e.g., Samotsvety members) and specific division of labor within Sentinel are not detailed.

Sources