Taras Morozov

133 karmaJoined Mar 2022


Research raising sanity waterline

It seems that teaching the general public rationality tools may cause more polarisations. That is because many ideas seem to be used primarily for argument-winning instead of truth-seeking.

There is a risk that some ideas will make people less rational when teaching rationality. For example, Eliezer Yudkowsky wrote an article Knowing About Biases Can Hurt People

Scott Alexander uses the term Symmetric and Asymmetric Weapons for a similar idea: Some thinking-tools are more useful for winning arguments than truth-seeking. 

Therefore, there is a need for research into improving the general public's rationality without causing harm and even more polarisation.

Create an organization doing literature reviews and research on demand

Values and Reflective Processes,  Effective Altruism

Create a research organization that will offer literature reviews and research to other EA organizations. They will focus on questions that are not theory-heavy and can be approached by a generalists without previous deep knowledge of the field. Previous examples of such research are publications of AI Impacts or literature reviews by Luke Muehlhauser.
Besides research itself, this is useful also for:
- it frees up the time of senior researchers
- It can be a good training place for junior researchers.
- it may enable a larger infusion of valuable ideas from academia.

Research differential technological progress and trajectory changes
Research That Can Help Us Improve, Values and Reflective Processes

The idea of Differential technological progress (DTP) may be a crucial consideration for many at-first-glance good ideas like:

- improving scientific publishing

- increasing GDP

- increasing average intelligence


But given its importance, there hasn`t been much research and publications on GTP.

Central question for research is how to use DTP to prioritize interventions. Examples of subquestion to research are:

- when in the past there were intentional trajectory changes.

- what subgoals seem to be good when DTP is considered.

- and so on.

Find good ways to distribute books to people with high potential

Epistemic Institutions, Effective Altruism

This project has two parts:
1) find people with high potential, especially students.
2) find a good way to distribute books on world problems to them.

Ad 1: Examples:
- students in low and medium income countries may have a higher demand for English books
- participants on STEM olympiads
- people with SAT scores > x
- students in selective schools

Ad 2: It is important to do it in a nice, non-preaching way.
One possible implementation is a book club that sends out a book every two months, with regular online meetups for its readers.

Create and curate educational materials on EA-related topics

 Effective Altruism

EA Fellowship and EA Handbook took existing resources and curated them into a good introduction to EA. Do something similar with different formats and subjects.
I.e., create: 
- Fellowships
- Reading lists.
- record existing courses in academia
- and so on

With a goals to:
- make it easy to take up new fields.

In fields like:
- Rationality
- Bioweapons
- Forecasting
- and so on.

To prove the point: ATM the most upvoted comment is also the oldest one - Pablo's Retrospective grant evaluations.

Offer paid sabbatical to people considering changing careers

Empowering Exceptional People

People sometimes are locked-in in their non-EA careers because while working, they do not have time to:

  • Prioritize what altruistic job would fit them best
  • Learn what they need for this job

Create an organization that will offer paid sabbaticals to people considering changing careers to more EA-aligned jobs to help this transition. During the sabbatical, they could be members of a community of people in a similar situation, with coaching available.

Provide personal assistants for EAs

 Empowering Exceptional People

Many senior EAs spend way too much with busywork because it is hard to get a good personal assistant. This is currently so because: 

  1. There is no obvious source of reliable, vetted assistants.
  2. If an EA wants to become an assistant, it is harder for them to find a job for EA or on EA-related projects.
  3. Assistants have an incentive to have many clients, to avoid loss of income if they would lose their client. This leads to assistants having less time per client, and thus more time is spent on communication and less on work itself.
  4. Assistants tend to be paid personally by EAs instead of by their employers. That leads to using them less than would be optimal.
  5. There is no community of assistants that would be sharing knowledge and helping each other.

All these factors would be removed if an agency managed personal assistants.

Look for UFOs

Space Governance

In recent years, there has been an upsurge in reports by the military on sightings of UFOs including detecting the same object with multiple modalities at once (examples: 12). 

Avi Loeb proposes to create a network of high-resolution sensors, (just as the military have). But compared to the military their results will not be classified and can be openly analyzed by scientists. The cost of doing this is in the order of millions of dollars. 

Knowing if there are aliens has many consequences, including for the longtermism, since if the universe is densely populated, its expected moral value does not depend solely on our decisions.

Write encyclopedias (esp. Wikipedia), then translate them (esp. to Russian and Chinese)

Epistemic Institutions

Create a team of people who will write articles on Wikipedia on subjects related to EA. Why this is important is described here

Besides writing articles on English Wikipedia, they can also:

  1. Create good illustrations (somehow, medical articles tend to have much beter pictures than other areas)
  2. Translate these articles to other languages (especially Russian and Chinese)
  3. Topics that are not notable enough for Wikipedia can be described in a separate encyclopedia.
Load more