Audio versions of blogs and papers from BlueDot courses.

Original text: https://drive.google.com/file/d/19dxfYiUnrK4fR5EJW3Nb2zkpgvsEC28y/view?usp=sharing

Author(s): Roger Brent, T. Greg McKelvey, Jr., and Jason Matheny

A podcast by BlueDot Impact.

Audio versions of blogs and papers from BlueDot courses.

Original text: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7103698/

Author(s): Amesh Adalja and Thomas Inglesby

A podcast by BlueDot Impact.

Audio versions of blogs and papers from BlueDot courses.

ESM3 is an AI model which is trained on protein sequence, structure and function data. This allows it to follow prompts to design novel proteins. When reading, consider the ways in which this model might present opportunities and risks for pandemic preparedness.

Original text: https://www.evolutionaryscale.ai/blog/esm3-release

Author(s): Evolutionary Scale blog

A podcast by BlueDot Impact.

Audio versions of blogs and papers from BlueDot courses.

This paper introduces a framing of "differential technology development" – affecting the relative timing of new innovations to reduce a specific risk. For instance, it may be beneficial to delay or halt risk-increasing technologies and preferentially advance risk-reducing defensive, safety, or substitute technologies. As you read, consider how you might apply this framing to biosecurity risks associated with developments in AI and biotechnology (this week's pre-class exercise).

Original text: https://ora.ox.ac.uk/objects/uuid:b481e9ad-bc27-4550-87ca-f414354aeb35

Author(s): Jonas B. Sandbrink, Hamish Hobbs, Jacob L. Swett, Allan Dafoe, Anders Sandberg

A podcast by BlueDot Impact.

Audio versions of blogs and papers from BlueDot courses.

In this post, Kyle Fish presents five key goals for biosecurity technologies: to be fast, general, cheap, robust, and scalable. These will be useful heuristics for us to apply as we consider a range of interventions over the next three weeks. Kyle was previously the CTO at Alvea, a vaccine start-up which became the fastest startup to take a new drug from founding to a Phase 1 clinical trial.

Original text: https://forum.effectivealtruism.org/posts/rjFJYREBZZscvDwLu/a-framework-for-technical-progress-on-biosecurity

Author(s): Kyle Fish

A podcast by BlueDot Impact.

Audio versions of blogs and papers from BlueDot courses.

This resource provides an overview of the Biological Weapons Convention, a disarmament treaty that bans the development, production, acquisition, transfer, stockpiling and use of biological weapons. Despite 187 states being party to the treaty, there are substantial challenges including a lack of funding and difficulty verifying compliance.

Original text: https://disarmament.unoda.org/publications/the-biological-weapons-convention/

Author(s): United Nations Office for Disarmament Affairs

A podcast by BlueDot Impact.

Audio versions of blogs and papers from BlueDot courses.

The US State Department report is released each April and provides an assessment of states' adherence to arms control, non-proliferation and disarmament agreements and commitments. We see this as a useful resource giving some evidence of modern BW activity and for you to consider the role of intelligence agencies in assessing BWC compliance. However, we would also encourage reflection regarding the specific language used and how/what information is presented. The US has written this with the intention of publishing it openly, so it is worth considering their broader geopolitical motivations when reading.

Original text: https://www.state.gov/wp-content/uploads/2024/04/2024-Arms-Control-Treaty-Compliance-Report.pdf

Author(s): US Department of State

A podcast by BlueDot Impact.

Audio versions of blogs and papers from BlueDot courses.

By Samuel Hammond

Source:
https://www.secondbest.ca/p/ai-and-leviathan-part-i

A podcast by BlueDot Impact.

Audio versions of blogs and papers from BlueDot courses. 

By Vitalik Buterin

Ethereum founder Vitalik Buterin describes how democratic, defensive and decentralised technologies could distribute AI's power across society rather than concentrating it, offering a middle path between unchecked technical acceleration and authoritarian control.

Source:

https://vitalik.eth.limo/general/2025/01/05/dacc2.html

A podcast by BlueDot Impact.

Audio versions of blogs and papers from BlueDot courses.

By Sella Nevo et al.

In this report, RAND researchers identify real-world attack methods that malicious actors could use to steal AI model weights. They propose a five-level security framework that AI companies could implement to defend against different threats, from amateur hackers to nation-state operations.

Source:

https://www.rand.org/pubs/research_briefs/RBA2849-1.html

A podcast by BlueDot Impact.