Made with 💙 by me (how to).
Dedicated to my family.
The EA University
Status: Project Draft My long-term goal is to be working on the EA University Megaproject (idea from Will MacAskill). I believe that I could be a good fit for working on this, given exceptional training and feedback loops over the next <5 years. I also think this is the most impactful EA project I have found so far, as it’s about as “meta” as one could go.
If you have thoughts about this, or know someone working on this, please contact me!
Some loose thoughts of what I have in mind:
- From Wikipedia:
- The word university is derived from the Latin universitas magistrorum et scholarium, which roughly means "community of teachers and scholars" and universitas = 'a whole'
- This means the EA University would be the holistic place where EA reasoning takes place, which includes:
- Learning and Education (corresponds to in my How to solve problems framework)Pre-Creation |
- Academic work à la Bell Labs ()Creation |
- Publishing, peer review and company/charity spin-outs ()Post-Creation |
- Grant-making and priorities research ()Impact
- A physical space / retreats / a virtual network of operations, coaching and general support infrastructure ()Life |
- This is a very broad vision, so I am currently thinking a lot about what small MVP I can build for at least one of the components. But I am also trying to create a network of people who have been thinking about some part of this
- Basically we want a “football star attitude” towards researchers: the ratio of ops, coaches, trainers per researcher should be as high as possible. Scientific breakthroughs are tail end distributed, you don’t produce them equally distributed over a 40 hour work week. It’s more likely that something important happens if a team of top researches come together for a “football match” once in a while, perfectly prepared, undistracted. (We just don’t want a loud group of hooligans cheering them up, that’s where the metaphor breaks down 😅) (but maybe we want betting markets [impact certs] and youth scouts [atlas fellowship]?)
- Why do we need this?
- Learning is inefficient, I think we can upskill people a lot quicker and build very good teaching institutions based on evidence-based learning facilitations. (roughly: teaching is bad, learning facilitation is good)
- strong epistemic status
- Unclear how neglected, Training for Good, the fellowships, AI bootcamps
- Streamlining this more sounds helpful though
- Classical academic life can be very contrary to deep, cooperative work.
- unclear how true this is for EAs, more physical spaces that are optimized for explanation creation seem good though
- Publishing is ripe for a revolution anyway, paywalled pdfs have locked in human potential for far to long
- decent evidence, stronger claim is that journals might have been the biggest limiter of GDP growth in this century.
- unclear of how tractable, more experimentation seems good though
- Twitter for peer-review?
- Open science vs. security tradeoff. But paywalls aren’t a security solution
- Grant-making and evaluating science is hard. We should still try out a bunch of new ways like quadratic funding, retroactive funding, etc.
- weak evidence that we can improve much here within EA though. I still think PIs should not write grant applications, that seems terribly inefficient
- We can improve a LOT compared to non-EAs though, so if the university becomes reasonably prestigious, that would be huge
- What age to start? Ideally after Synthesis.io?
- Avoid signaling at all cost vs. start a traditional uni with undergrade programms? Takes to long, decentralized approach sounds a lot better
- Physical Place
- Harvard Square Coworking space
- First faculty should be “meta”
- Faculty for Problem solving: research on how to organize groups to solve hard problems, learning, publishing, creativity science
- PI brainstorm
- Dr. Barbara Oakley (learning)
- Effective Self-Help (coaching/ops)
- Alexandra Elbakyan (publishing)
- Who was behind Bell Labs / Cal Newport
- Altos Labs Ops team?
- Jürgen Klppp (yes, why not)
- Faculty for
- AI Safety (probably the hardest, as so many new orgs anyway. Umbrella org?)
- Biorisk (?)
- scalable altruism (VC/incubator)
- Global Priorities (Forethought, GPI)
- Altruistic Evaluation (GiveWell)
- legal representation of the future
- preventative medicine
- governance (Audrey Tang)
- weird stuff (Anders Sandberg)
Create a Discord server for people who show interest?
Publish a first draft of an Elon Style Master Plan