Zoe Lu
Pro-sentient AGI through cooperation and competition / Advocacy in a post-ASI world
Independent
Bio
I'm the first product hire at HRT focusing on the R&D platform to enable algo developers to build the best (often GPU/neural-powered) trade strategies in the industry. Previously, I worked as a product manager at Two Sigma. I also have experiences working with startups and VCs. During my undergraduate years, I conducted research in AI safety and regulation regarding self-driving cars while double majoring in Computer Science and Mechanical Engineering, with a minor in Philosophy, at Massachusetts Institute of Technology.
Mentee must-haves/nice-to-haves
Must-have: proactive/agentic Want: overlapping background/interest would be nice in terms of building general rapport
Mentee role
They will shape the emphasis of the project, research on the topics, and draft the output.
❓ Sample mentee tasks
1. Modeling AI cooperation/competition and its impact on overall alignment towards sentience wellbeing - this can be as broad-stroke as outlining the most likely scenarios and as detailed as using simulation software to tweak different parameters and offering an interactive tool for the audience to explore this idea themselves 2. Modeling super human intelligence timeline and outcomes - similarly, this can be as broad-stroke as outlining the most likely scenarios and as detailed as using simulation software to tweak different parameters and offering an interactive tool for the audience to explore this idea themselves
Depending on mentees' interest, more time/focus can be spent on building a simulation tool vs. writing up the takeaways.
Mentor support
Scoping/shaping direction, project planning, facilitating connections, giving advice and feedback (on the specific project or more generally on advocacy, career, life etc.), encouragement and accountability, help to prepare/rehearse presentations/pitches (if relevant)
Questions for applicants
In a few bullets: (1) Why are you drawn to this topic/project? (2) Which part(s) of this topic/project is the most interesting to you? (3) What are your relevant skills and experiences? (4) What skills and experiences (if any) would you like to build during this project?
Mentor-led project
Project 1: AGI sentience alignment through cooperation and competition / Project 2: Sentience advocacy in a world of superhuman intelligence
The two main questions I'm most interested in answering are: 1. Given the current progression in AI development, it's likely (or at least possible) that multiple companies will develop several differently-aligned AGIs and these AGIs will need to cooperate/compete with each other. Will these AGIs align on a future that more or less prioritizes wellbeing for all sentience compared to the current human alignment? What can we do to facilitate favorable alignment during this cooperation/competition phase?
The expected output will be a research paper/blog post with action items. Depending on mentees' interest, we can emphasize different parts of this topic (e.g., spend more time modeling AI cooperation/competition vs. brainstorming intervention/evaluation strategies).
2. Besides developing machine intelligence, there are also increasing investments/efforts (e.g., Neuralink, CRISPR) to enhance human intelligence/abilities. The focus so far in researching alignment is on machine intelligence. In a world with abundant super human intelligence (and perhaps near-immortality), will we prioritize wellbeing for other sentience more or less? How do current animal/AI advocacy interventions fare when humans are much more intelligent?
The expected output will be a research paper/blog post with suggested action items. Depending on mentees' interest, we can emphasize different parts of this topic (e.g., spend more time researching super human intelligence timeline vs. animal/AI advocacy efficacy in this world).