arXiv Science⌕ Search

arXiv subjects

Greg Kestin

Publications and source records attributed to Greg Kestin.

7 recordsLinked to original sources
arXiv2026

Views on AI Existential Risk Before and After a Public Event at Harvard University

We report the results of identical pre- and post-event surveys given to attendees of a talk, two-sided conversation, and Q&A centered around the book If Anyone Builds It, Everyone Dies at Harvard University in March 2026, covering perceived probability of AI-caused extinction or severe disempowerment resulting from unimpeded AI development, confidence in those estimates, and global priority. Among the 89 matched participants, the post-event median estimate of the probability of existential risk from advanced AI was 70%, and 96% agreed that mitigating AI existential risk should be a global priority. Although these self-selected respondents' pre-event views were already high (50% and 93%, respectively) relative to results of similar surveys that were previously administered to experts and the general public, the event produced increases on all measures when considering the respondents in aggregate. The magnitudes of increases in risk probability were negatively correlated with prior familiarity with the topic: among attendees with little prior familiarity, 60% shifted upward and none shifted downward, whereas among self-described experts, no respondents shifted upward and 20% shifted downward. Self-reported confidence also increased significantly, and confidence shifts were positively correlated with probability shifts. These findings indicate that a structured public engagement event can meaningfully shift risk perceptions, particularly among newcomers to the topic.

cs.CY↗
arXiv2025

Non-orbital particle trapping in binary black holes through dynamic stability

We present an interdisciplinary comparison between binary black hole systems and Radio Frequency (RF) Paul Traps, modeling the gravitational binary system as a rotating saddle near its center. This analogy connects these seemingly unrelated systems through the concept of dynamic stability. The rotating saddle potential is analytically tractable, allowing us to prove the existence of bounded charged particle trajectories under certain conditions. By focusing on stellar-mass black holes with a weak electric charge-a feature consistent with specific astrophysical conditions that leaves the spacetime metric largely unaffected but can influence nearby particle interactions-we can neglect complicating factors such as magnetic fields from large accretion disks of heavier black holes or stellar winds. Our simulation results demonstrate that charged particles can exhibit stable, non-orbital trajectories near the center of a binary system with charged stellar-mass black holes, providing unique three-dimensional trapping primarily through gravity. This system is distinctive in the literature for its non-orbital trapping mechanism. While theoretically intriguing, this trapping relies on specific conditions, including nearly identical black hole masses. These types of non-orbital trapping mechanisms could potentially allow for longer-lived plasma configurations, enhancing our ability to detect electromagnetic signatures from these systems. The significance of this work lies in the novel comparison between a laboratory-scale quantum system and a larger astrophysical one, opening new avenues for exploring parallels between microscopic and cosmic phenomena across fourteen orders of magnitude in distance.

astro-ph.HE↗
arXiv2025

Two-Stage Final Exams: An Assessment Strategy for Enhanced Collaborative Learning and Reduced Student Stress

Two-stage exams, which pair a traditional individual exam with a subsequent collaborative, group exam, have been shown to enhance learning, retention, and student attitudes in midterm settings. This study investigates whether these benefits extend to final exam settings, compares the traditional final exam experience with a two-stage variation, and explores using an asynchronous group component to maintain the exam comprehensive nature. Within introductory courses at two different institutions, two formats for two-stage exams were implemented: one with a 2-hour individual exam followed by a 1-hour synchronous group exam and another with a 3-hour individual exam followed by an untimed, asynchronous group exam completed within 36 hours. Survey data comparing student experiences with two-stage midterm and final exams showed consistently positive responses at both institutions. When comparing two-stage final exams to traditional (individual-only) final exams, students reported increased learning and retention, decreased anxiety, and an overwhelming preference for the two-stage format. These findings suggest that extending the benefits of two-stage midterms to the final exam is possible without compromising the evaluative purpose of a final exam while also maintaining high student motivation and engagement.

physics.ed-ph↗
arXiv2025

Active Learning Through Flexible Collaborative Exams: Improving STEM Assessments

Two-stage exams, which pair a traditional individual exam with a subsequent collaborative exam, are widely popular with both students and faculty for fostering deep engagement, collaboration, and immediate feedback. Over the last decade, this assessment model has gained substantial traction in STEM college courses; however, holding both stages during class sessions often limits the full potential of two-stage exams, deterring many instructors from adopting them. To accommodate these constraints, instructors must either reduce the exam's content depth and coverage, or hold evening, in-person exam sessions at fixed times, which can often be impractical for students and faculty with external commitments. In this paper, we introduce an alternative asynchronous approach that allows the collaborative portion to be completed unsupervised and outside of regular class hours, within 48 hours of the individual exam. This method not only eases logistical constraints but also allows more time for collaboration, potentially enhancing feedback quality. Our findings show that students equate their engagement, collaboration, and feedback quality during the asynchronous collaborative exam with that of standard in-class exams and show signs of retention of learning. This cost-effective approach requires no extra time or resources and could promote widespread adoption of this effective assessment method, especially in online-only and general STEM courses.

physics.ed-ph↗
arXiv2016

Towards an Effective Field Theory on the Light-Shell

We discuss our work toward the construction of a light-shell effective theory (LSET), an effective field theory for describing the matter emerging from high-energy collisions and the accompanying radiation. We work in the highly simplified venue of 0-flavor scalar quantum electrodynamics, with a gauge invariant product of scalar fields at the origin of space-time as the source of high-energy charged particles. Working in this simple gauge theory allows us to focus on the essential features of LSET. We describe how the effective theory is constructed and argue that it can reproduce the full theory tree-level amplitude. We study the 1-loop radiative corrections in the LSET and suggest how the leading double-logs in the full theory at 1-loop order can be reproduced by a purely angular integral in the LSET.

hep-ph↗
arXiv2015

Color Fields on the Light-Shell

We study the classical color radiation from very high energy collisions that produce colored particles. In the extreme high energy limit, the classical color fields are confined to a light-shell expanding at $c$ and are associated with a non-linear $σ$-model on the 2D light-shell with specific symmetry breaking terms. We argue that the quantum version of this picture exhibits asymptotic freedom and may be a useful starting point for an effective light-shell theory of the structure between the jets at a very high energy collider.

hep-ph↗