EA - Announcing: SERI Biosecurity Interventions Technical Seminar (BITS) by James Lin

Link to original articleWelcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Announcing: SERI Biosecurity Interventions Technical Seminar (BITS), published by James Lin on January 12, 2023 on The Effective Altruism Forum.1. IntroductionThe biosecurity space has promising technical interventions, such as UV-C and metagenomic sequencing. Often, students in college and community-builders are interested in biosecurity, but don’t have an in-depth understanding of the solutions. This is concerning for 3 central reasons:Poor explanations are presented for biosecurity solutions, and may contain inaccuraciesThe reputation of the biosecurity community is damagedPublic perception of biosecurity interventions will be off the mark relative to the intervention’s true effectiveness (e.g. “far UVC will end pandemics by 2025”)As a consequence, people are turned away from biosecurityPromising people transition to other fields of workWhen someone doesn’t understand the specifics, it’s highly unlikely that they’ll pursue direct work, e.g. building a startup for that technologyInstead, they’ll likely transition to community-building or traditional career pathsBiosecurity culture becomes diluted and can’t attract technically bright researchers or entrepreneursCompetence is magnetic, and promoting a culture of technical learning and deep dives is crucial to doing impactful work in a domain as challenging as biosecurityA technical biosecurity program can serve as a way for students to become much more knowledgeable about biosecurity solutions.2. ContextThere have been biosecurity discussion groups run in the past, but most of these groups focus on the top-of-the-funnel content—the introductory information and context that helps to form a foundation. To our knowledge, there doesn’t yet exist a deep dive group, one that specifically aims to equip students with technical understanding of various biosecurity interventions.When it comes to precedent, there have been similar discussion and research mentorship programs in the field of AI alignment, including AGISF, SERI MATS, MLSS, and the CERI Fellowship for biosecurity, nuclear risk, climate change, and AI safety. These programs have been largely successful, providing hundreds of students with hands-on technical experience or in-depth discussions about particular domains of AI safety.3. ProgramAboutOur proposal is to establish an online 8-week-long technical deep dive group. This group will explore many crucial aspects of technical biosecurity interventions while ensuring that we maintain an emphasis on defensive interventions. Examples of session topics include far-UVC, SotA for PPE, and the NAO (syllabus). The program will culminate in a research presentation for a chosen aspect of related topics and interventions (more details here). Each group will meet once per week and will have ~5 people. The expected weekly time commitment is 2-3 hours.There will be a virtual and in-person pilot of this program. Everyone is eligible to apply for the virtual program. The in-person program seminar group will be held at Stanford, though depending on demand we may also run an in-person group at Berkeley.There are a few reasons for running a primarily virtual program:It’s important to have an experienced biosecurity person facilitating, which usually limits these interactions to virtualVirtual programs are much more scalable and flexibleIt’s easier to reach a critical mass of studentsIf you would like to join, please feel free to apply! The application deadline is Friday, Jan 20th, and the program is expected to start in 2-3 weeks. We expect to start by running a few groups in parallel at different times during the week to accommodate those in different time zones.SyllabusThis is the syllabus with the content we will be covering. There will also be an optional research poster to be submitted 2 weeks after the...