MPHD at Berkeley: Math, Productivity, Happiness & Decision-making

 about the seminar | schedule and abstracts | how it started | usual format

Semester: Spring 2012
Organizers: Andrew Critch and Nisan Stiennon
Time: Wednesdays, 5:00 – 6:00pm
Location: 740 Evans Hall
Contact / join the mailing list: critch at math dot berkeley dot edu

On every second week, we go out for dinner after the seminar. 

About the seminar

Most of us want to be happierget things donehelp other peoplehave a sense of purpose, and achieve other such deeply human goals. In this seminar, we hope to take advantage of our analytic minds — our comparative advantage as mathematical thinkers — to address these goals, and find ways our analytic minds can learn more from our emotions and intuitions, too. Some of our guiding principles are:

  • Improving on purpose:  Life skills like time-management, emotional stability, motivation, and communication will affect all of our goals for the rest of our lives, so they’re probably worth improving on purpose instead of by random walk.
  • Learning from science:  Fields like psychology, neuroscience, education, and economics can tell us a lot about what a good strategy looks like and how to implement one on an organic human brain.
  • Reasoning under uncertainty:  Hating uncertainty won’t make it go away. We still have to make decisions constantly, so refining our mental and emotional probabilistic reasoning skills is likely to be worth a non-zero allocation of time and awareness. The skill of belief updating — noticing and internalizing evidence — is especially worth upgrading from decent to awesome, and heuristics from probability, statistics, and artificial intelligence can help us quantify what me mean by that.
  • Working together:  We can learn more about life if we work together than on our own. For that it helps to have a common language to understand each other, and to consciously formulate our beliefs and values in a way that can be communicated and analyzed.


Date Dinner Kick-off Speaker Topic announcement Links / materials
Jan 18 No Andrew Critch,
UC Berkeley, Mathematics
Overview / organizational meeting  
Jan 25 Yes Nisan Stiennon, Stanford, Mathematics Formulating beliefs and values  
Feb 1 No Andrew Dudzik,
UC Berkeley, Mathematics
Getting Things Done (with less stress) Wikipedia article on GTD
Software: Remember the Milk, for web and mobiles
Software: Evernote, for web, PC, Mac, and mobiles
Feb 8 Yes Andrew Critch,
UC Berkeley, Mathematics
Directed graphical causal models for our personal lives Paper: Easterday et al (2009) Constructing causal diagrams to learn delibration
Website: Matthew Easterday’s research
Handout: Using causal graphs and thinking hats
Worksheets: Be creative, Be curious, Be skeptical, and Be strategic
Feb 15 No Nick Hay,
UC Berkeley, Computer Science
Training work-focus with the Pomodoro Technique (and GTD update) Wikipedia article on the Pomodoro Technique
Software: Pomodroido, for Android
Feb 22 Yes Nisan Stiennon
Stanford, Mathematics
Confidence calibration: Make probabilities work for you Worksheets: Confidence calibration
Feb 29 No Andrew Critch,
UC Berkeley, Mathematics
Coordinatizing emotion-space (and feeling it) Wikipedia article on Affective neuroscience
Worksheets: Emotional awareness
Mar 7 Yes Anna Salamon,
Stop rationalizing: The what fooling yourself feels like game Paper: Taber, Lodge (2006) Motivated Skepticism in the Evaluation of Political Beliefs
Worksheets: Types of rationalization and how they feel
Mar 14 No Kaushik Krishnan
UC Berkeley, Economics
Spaced repetition learning and software Wikipedia article on spaced repetition learning
Software: Mnemosyne
Software: Anki
Mar 21 Yes Andrew Critch,
UC Berkeley, Mathematics
How and when to change beliefs: Internalizing Bayesian updating Wikipedia article on Bayes’ Rule
An Intuitive (and Short) Explanation of Bayes’ Theorem
Worksheet: the Really Getting Bayes game
Mar 28     No Seminar; Spring Break  
Apr 4 Yes Joseph Williams,
UC Berkeley, Psychology
Strategies for behavior change  
Apr 11 No Elliott Collins,
UC Berkeley, A&R Econonomics
Calibrating shyness/embarrassment with Rejection Therapy  
Apr 18 Yes Anna Salamon, SingInst Value of information and Fermi estimations in daily life  
Apr 25 Yes Andrew Critch,
UC Berkeley, Mathematics
The future of human reasoning  


Last year, Nisan and I attended a week-long workshop on “Rationality and Social Effectiveness” with 24 other people, mostly grad students and tech professionals, who all seemed to really like it. Now we want to recreate something similar in a seminar format, and the first step to getting good at something is to practice it once, so here we go! 

Usual format

Our default format is designed to give the audience time to optimize and filter which questions/issues get discussed with the room as a whole, and to ensure everyone experiences some amount of personally relevant conversation:

10-15 mins: kick-off A speaker presents a quick topic introduction, setting an intuitive focal point for the rest of the seminar. To keep this part short, audience members ask few or no questions, and instead jot down self-reminders of questions and thoughts they want to pursue later.
0-5 mins: audience contributions The audience can contribute quick facts or comments they think will be concrete and useful for the whole room, but not yet questions. Please contribute experience and expert knowledge here!
5-10 mins: small group
Small groups form, and people with relevant experience or knowledge are asked to distribute among the groups. Participants ask each other their stored questions from the kick-off, and try to clarify things for each other. In this way, people intuitively settle on what their most important remaining ideas and questions are. Some ideas to start conversation:
  • Questions, reservations, or criticisms
  • Related personal experiences
  • How we can use the kick-off ideas in our own lives
  • How we can test the kick-off ideas in our own lives
10-15 mins: unified Q&A discussion
The room again engages in a single conversation, much like the “any questions” part of other seminars, but with a moderator summarizing the discussion on the board as it progresses, both as a record and a subtle social signal to stay on topic. Audience members should feel free to answer each other’s questions instead of channeling all communication through the speaker when it seems inefficient to do so. The kick-off speaker is not expected to help moderate the conversation, but should feel welcome to.

Sometimes, in place of the small group conversations, there will be specific exercises to stimulate (hopefully) interesting thoughts, and/or to intuitively clarify the “spirit” of that day’s topic.