I am giving a talk at my advisor’s workshop on the Future of Analytics and Operations Research which will be well attended by MIT PhD students, professors, and alums from the operations research center (many of which work on healthcare analytics and were involved in COVID forecasting/response). I plan to talk about “Strategic Priorities in Health Security” as well as circulate a document which gives an overview of biosecurity and the problems best suited for people in my academic community (data science/analytics/OR; see why EA needs OR for additional context).
Goal and Questions
My specific goal with this post is to solicit feedback from the biosecurity community to maximize the chance that this talk is able to be an entry point to bring new, talented researchers into impactful work on biosecurity. Feedback of any kind related to this goal is appreciated, but some specific questions are below:
What is the most inspiring / scary / accurate / overall best way to help this audience quickly understand GCBRs and the work being done to mitigate them?
Do you have any suggestions for potential projects that could benefit in a material, mission-critical way from world-class analytics / OR research support[1], or would be particularly good motivating case studies?
Is any of the existing OR research on biosecurity (examples here and here) particularly useful? What would you like to see more or less of?
What is the best way to bring interested researchers further into the community? What’s the best “single next action” that someone interested in engaging more should take?
Do you work with or know of a team/org that needs research support of this kind now (or will soon) that could immediately benefit from collaborators at MIT?
Do you see any key risks or red flags that could result in this talk becoming infohazardous or otherwise counterproductive?
My tentative thoughts for my talk and followup document are given below. If you would rather your answers be private, please email me.
Talk Outline
I only have 8 minutes so I have to be concise. The overarching message is “you should be scared shitless of engineered pandemics–here are some relevant problems and how they relate to things you have expertise in.”
Tentative outline:
Historical examples of near GCBRs
Investment in pandemic prevention/preparedness is woefully inadequate (especially in comparison to other forms of defense spending)
This is especially alarming when considering advances in biotech, eg:
Dual use AI models
DNA sequencing/synthesis costs coming down at Moore’s law speed
Gain of function research
Introduce defense-in-depth and use as an organizational structure to introduce key problems for rest of talk
Prevention
Far-UVC: optimize allocation, model effectiveness vs. cost
Monitoring/detection
NAO design, operations, and data analysis
Rapid response+preparedness
Targeted testing, containment of affected areas
Sustained response
Vaccine manufacturing, allocation, and distribution
Worst case resilience
Facility location for refuges
Guide Outline
This will be provided at the end of the talk. The goal is to give readers additional context and a large sample of relevant problems with enough detail and references to get people thinking in the right direction.
Describe the motivation for biosecurity
Introduce key concepts
Defense-in-depth
Dual-use research
Infohazards
Relevant problems organized by depth (with more detailed problem statements forthcoming)
Prevention
Far-UVC: optimize placement, model effectiveness vs. cost
Optimize lab/facility inspection schedules
Monitoring/detection
NAO design, operations, and data analysis
Better models for disease spread
Rapid response+preparedness
Facility location and inventory management of strategic stockpiles
Integrated data sources to enable rapidly deploying medical countermeasures to areas of need
Targeted testing
Containment of affected areas
Sustained response
Hardened supply chains for testing, PPE, food
Inventory management
Warm manufacturing capacity
Vaccine manufacturing, allocation, and distribution
Clinical trial site identification
Worst case resilience
Facility location for refuges
Food stockpiles and supply chain hardening of most basic survival goods
Emphasize discretion and infohazards one last time
Links to relevant orgs in the space and further list of references
OR is primarily an applied discipline, and therefore OR academia is unusually supportive of purely applied research projects. A good (non biosec) recent example is here where an academic OR collaboration saved the UN World Food Program $150 million/year in logistics costs.
What should MIT analytics professors know about biosecurity?
Context
I am giving a talk at my advisor’s workshop on the Future of Analytics and Operations Research which will be well attended by MIT PhD students, professors, and alums from the operations research center (many of which work on healthcare analytics and were involved in COVID forecasting/response). I plan to talk about “Strategic Priorities in Health Security” as well as circulate a document which gives an overview of biosecurity and the problems best suited for people in my academic community (data science/analytics/OR; see why EA needs OR for additional context).
Goal and Questions
My specific goal with this post is to solicit feedback from the biosecurity community to maximize the chance that this talk is able to be an entry point to bring new, talented researchers into impactful work on biosecurity. Feedback of any kind related to this goal is appreciated, but some specific questions are below:
What is the most inspiring / scary / accurate / overall best way to help this audience quickly understand GCBRs and the work being done to mitigate them?
Do you have any suggestions for potential projects that could benefit in a material, mission-critical way from world-class analytics / OR research support[1], or would be particularly good motivating case studies?
Is any of the existing OR research on biosecurity (examples here and here) particularly useful? What would you like to see more or less of?
What is the best way to bring interested researchers further into the community? What’s the best “single next action” that someone interested in engaging more should take?
Do you work with or know of a team/org that needs research support of this kind now (or will soon) that could immediately benefit from collaborators at MIT?
Do you see any key risks or red flags that could result in this talk becoming infohazardous or otherwise counterproductive?
My tentative thoughts for my talk and followup document are given below. If you would rather your answers be private, please email me.
Talk Outline
I only have 8 minutes so I have to be concise. The overarching message is “you should be scared shitless of engineered pandemics–here are some relevant problems and how they relate to things you have expertise in.”
Tentative outline:
Historical examples of near GCBRs
Investment in pandemic prevention/preparedness is woefully inadequate (especially in comparison to other forms of defense spending)
This is especially alarming when considering advances in biotech, eg:
Dual use AI models
DNA sequencing/synthesis costs coming down at Moore’s law speed
Gain of function research
Introduce defense-in-depth and use as an organizational structure to introduce key problems for rest of talk
Prevention
Far-UVC: optimize allocation, model effectiveness vs. cost
Monitoring/detection
NAO design, operations, and data analysis
Rapid response+preparedness
Targeted testing, containment of affected areas
Sustained response
Vaccine manufacturing, allocation, and distribution
Worst case resilience
Facility location for refuges
Guide Outline
This will be provided at the end of the talk. The goal is to give readers additional context and a large sample of relevant problems with enough detail and references to get people thinking in the right direction.
Describe the motivation for biosecurity
Introduce key concepts
Defense-in-depth
Dual-use research
Infohazards
Relevant problems organized by depth (with more detailed problem statements forthcoming)
Prevention
Far-UVC: optimize placement, model effectiveness vs. cost
Optimize lab/facility inspection schedules
Monitoring/detection
NAO design, operations, and data analysis
Better models for disease spread
Rapid response+preparedness
Facility location and inventory management of strategic stockpiles
Integrated data sources to enable rapidly deploying medical countermeasures to areas of need
Targeted testing
Containment of affected areas
Sustained response
Hardened supply chains for testing, PPE, food
Inventory management
Warm manufacturing capacity
Vaccine manufacturing, allocation, and distribution
Clinical trial site identification
Worst case resilience
Facility location for refuges
Food stockpiles and supply chain hardening of most basic survival goods
Emphasize discretion and infohazards one last time
Links to relevant orgs in the space and further list of references
OR is primarily an applied discipline, and therefore OR academia is unusually supportive of purely applied research projects. A good (non biosec) recent example is here where an academic OR collaboration saved the UN World Food Program $150 million/year in logistics costs.