Workflows: an experimental alternative to introductory fellowships

By Swan 🔸 @ 2022-08-03T23:43 (+34)

TL;DR

Check out the Workflows Home Page, provide any feedback if you wish, and stay tuned for my update in three months if this actually works as an alternative to introductory reading groups and increases student retention. Reading group = Introductory Fellowship in this article. 

Contents:

1. Epistemics & Disclaimer
2. What is a workflow?
3. What is the theory of change?
4. What are my goals for the workflows during the months of September to November?
5. Anticipated Challenges and Downfalls
6. Accountability for this experiment
7. Acknowledgments

Epistemics & Disclaimer

• I am a university group organizer so this proposal is aimed at university organizers and the like. 
• UC Berkeley is only one university so even if this experiment works it would need to be reproduced in other universities. 
• This is still a work in progress so I expect a lot of issues to arise during the first iteration. 
• There are probably other solutions to this problem, which could include modifications to the reading group itself to make it more direct and action-oriented, so perhaps the workflow could be used as a supplement to a reading group as opposed to completely replacing it. 

What is a workflow?

• The Workflows Home Page is an alternative and self-paced version of an intro reading group that lets someone jump straight to the area that they are interested in to get the basics of it, while also providing the option to meet with group organizers, get personalized feedback & suggestions as to how to get involved with projects. 
• The long-term goal is for these (if they are proven to be helpful) to be made available to every university group as an easily accessible and comprehensive option.
• The proposed structure: 
  • Exploration: The goal for this stage is to gain enough familiarity with these ideas that the student can explain to someone new why it is or might not be a worthwhile cause area. 
  • Reflection: The student writes a 250-word reflection on the topic of the specific cause area. This is important because it gets the students to think critically about applying their own ideas to the readings. 
  • Personalized Feedback (ie a 1:1 chat): Once they are familiar with the introductory materials to the cause area/major they are interested in, they can meet with a group organizer to receive personalized feedback on the next steps. 
  • What’s Next & How to Take Action
    • What’s Next (General EA)
    • Suggested Courses & Skills 
    • Testing your fit
    • Organizations & Opportunities (both Full-Time and Part-Time)

What is the theory of change?

Note: Reading group = Introductory Fellowship in this article. 

Common critiques of reading groups:

• Lack of sustained attendance. See below for the statistics from our latest round of intro reading groups.
• Lack of direct action. Some students feel that the reading group doesn’t have any active part and that people “just sit around and talk about things”
• A misconception we have gathered from our feedback is that some people thought that the reading group is of a religious nature. The word fellowship can sometimes cause this reaction from students who aren’t sure what EA actually is. 

Statistics from our Summer Reading Group:

• How did people find out about the reading group?
  • 33 people filled out a general interest form & indicated that they were interested in specifically the reading group. Below is a breakdown of where the people came from. 
    • 12 came from major newsletter announcements 
      • 2 Integrative Biology/Molecular & Cellular Biology 
      • 2 Political Science
      • 8 Public Policy
    • 5 signed up from a post in an Electrical Engineering/CS student forum 
    • 5 came from flyering on campus (potentially from the AIMS contest)
    • 3 signed up from a Facebook post in Berkeley student FB groups
    • 3 were already members of EA UC Berkeley
    • 3 were unspecified
    • 2 came from class announcements (Intro Physics course)
    • 2 signed up because a friend was a part of EA (word of mouth)
    • 1 signed up at Cal Day (welcome day for new students)
  • 3 others found our group through word of mouth

• How engaged were people with the reading group?
  • 17 out of the 36 initially interested, actually signed up. 
  • 9 active members in the intro group (i.e. attend meetings and actively participate)
     
• How much direct action is being taken after the reading group?
  • A member previously participated in the AIMS contest and went to EAG SF.
  • Many participants are still confused about what direct actions they can next take. 

How the workflows would solve these issues:

• No need to attend synchronous meetings. Three-quarters of the workflow can be completed asynchronously with only a single 1:1 meeting required with the group organizers. If desired more 1:1 meetings can be built into the structure. 
• Direct action is included in every workflow. There are suggestions for courses and skills that can be developed, short (under 10hrs) Testing Your Fit Projects, as well as Part-Time and Full-Time Opportunities that students can apply for. 
• Not a similar format to bible studies in any way. 

Other benefits:

• All the resources are consolidated for the reader so it acts as a pipeline to direct action as opposed to just being a substitute for an introductory reading group. 
• Possibility for mass scaling within the university, as there is no longer a cap of 20ish students per reading group
• Possibility for mass scaling across various universities, workflows are easy to reproduce, as customizable templates are provided. 
• Gives university organizers some of their time back since they no longer have to commit 2hrs per week to organizing and preparing for reading groups. 
   

Downsides:

• Higher upfront time cost. It takes longer to set up the infrastructure for them initially, however this should pay off due to an increase in productivity once they are set up. 
• Workflows need to be updated more frequently than reading group syllabi due to the constantly evolving nature of the jobs and opportunities that are available. 
• People might develop worse epistemics and not actually think for themselves by following an easily laid out path. It would be unfortunate if people didn’t end up actually exploring different careers and ideas because workflows are currently so one-track and think this is a potentially heavy downside. Two ways this can be mitigated is through the creation of a “general exploration” workflow and also the continued emphasis on things such as 80,000hrs and more broad career exploration.
   

What are my goals for the Workflow during the months of September to November?

• Have 40 people minimum fill out our intake Airtable Form. 
• Have 20 people make it to the 1:1 meeting (¾ of the way through)
• Have 1 < people engage in one (or more) of the projects listed in any of the cause areas (this can be a part-time position or testing your fit mini projects)
• Gather feedback from as many people who engaged with any parts of the project and see where the successes and failures were. 
   

Anticipated Challenges and Downfalls

• Lack of engagement - EA UC Berkeley plans to include this as one of the main goals for recruitment. Specifics of this will be included in the update post.  
• Lack of follow-through on projects - I plan to include more 1:1s as needed in order to motivate people.
• Difficulties keeping track of everything - I plan to set up an Airtable to keep track of everyone so that there is an easy way to visualize at which stage people dropped off.

Accountability for this experiment

The purpose of this post is for accountability! That way I have to stick to my goals and the EA Community can see both the beginning of this experiment and the end, even if the results are negative or neutral. Negative results are still results! I plan to post an update somewhere around November. Hopefully, with some of the feedback, I can continue to improve and iterate on these!

Acknowledgments

• Miranda Zhang for making the career planning program which was an inspiration for this 
• Aris Richardson and Carolyn Qian for working on the ideation process of this and providing edits to the article itself
• Karthik Tadepalli for making the Global Development workflow
• Nathaniel Li & Owen Murphy for making the AI Safety Workflow
• Garrett Ehinger for helping with various readings & finding leads for the Animal Welfare Workflow
• A big thanks to EAGT Un-Conference the for helping me present this idea to a wider audience.  
• If I missed anyone just ping me :)