Uncommon Sense Summer Seminar

Play with cutting-edge artificial intelligence and try to figure out its impact on the world.
Challenge your ideas.
Compete in the The Economist's World in 2021 forecasting challenge.

Everyone, from students to world leaders, makes countless decisions each day such as:

Should you be taking vitamin D to prevent COVID (one study shows it significantly reduced COVID-19 hospitalizations but another study shows it doesn’t help patients who are already hospitalized)?

What career will make you happiest and have the best impact on the world (a doctor typically only saves as many lives as a quantitative trader donating 1% of their income)?

How should the US tackle climate change (a plastic bag ban may actually increase emissions, though it would decrease litter)?

Too often, these decisions are automatic, or influenced by our cognitive biases, or dictated by social norms. Even people’s most strongly held opinions are often just a product of their environment, not something they’ve actually thought through and come to their own conclusion on ⁠— you can predict someone’s political party with, 70% accuracy based on how their parents voted.

What is Uncommon Sense? Our program is a free online seminar for high schoolers offered in partnership with Stanford Effective Altruism that gives inquisitive high schoolers the opportunity to master decision-making techniques based in philosophy, economics, game theory, cognitive science, and statistics. Our teaching team of Stanford students provides students with valuable critical thinking skills in a unique discussion-based format.

Students will use the skills they learn in Uncommon Sense to compete in global forecasting competitions like The Economist's World in 2021 forecasting challenge for the chance to win the title of Superforecaster. In these contests, students will practice answers to questions such as “How many COVID-19 vaccines will be approved and/or authorized for emergency use by the U.S. FDA as of 31 March 2021?” and “What will be China's year-on-year GDP growth rate for the second quarter of 2021?”

In addition to forecasting, accepted students will learn to build models of complex systems, create Bayesian priors, examine and combat their cognitive biases, and make more accurate predictions. Classes are project-based, incorporating interactive activities such as a cold war nuclear simulation, an iterated prisoner’s dilemma tournament, betting with fake currency, and much more.

Above all, Uncommon Sense will help you reconsider the countless decisions you make each day, like:

Should you be taking vitamin D to prevent COVID (one study shows it significantly reduced COVID-19 hospitalizations but another study shows it doesn’t help patients who are already hospitalized)? What career will make you happiest and have the best impact on the world (a doctor typically only saves as many lives as a quantitative trader donating 1% of their income)? How should the US tackle climate change (a plastic bag ban may actually increase emissions, though it would decrease litter)?

Graduates of Uncommon Sense will gain the skills they need to come to their own conclusions about critical problems, make more informed choices, and better understand the levers and gears in the world around them.

Each day between the hours of 10 am and 8 pm PT, students will delve into lectures, discussions, games, and activities. All activities will occur weekdays from June 28th to July 9th. On average, there are about 1 hour of required real-time activities, 2 hours of optional real-time activities, and 2 hours of asynchronous work each day.


About the program

What, exactly, do you teach?

Great question! The things we teach fall into two broad categories:

  1. Instrumental Rationality: the skills you need to achieve your goals. What do you value for its own sake? What are the things you want to achieve above all else, the things your other actions are stepping-stones towards? These terminal goals could be happiness, improving the world, becoming famous, anything. Now ask yourself what you did this past hour. Was it the best possible action you could have taken to achieve your goals? We help you achieve your goals by giving you tools that help you become accustomed to making plans and habits that actually work and inch you in the direction you want to go.
  2. Epistemic rationality: the use of evidence and reason to form accurate beliefs about the world. Specific skills we will teach include Bayesian reasoning, statistics, epistemology, cognitive biases, forecasting, and building models of complex systems. We teach epistemic rationality by offering interesting and complex ideas for you to discuss, debate, and form beliefs about. We also equip you with useful ideas that help you think clearly about uncertainty and complex systems.

What do you mean by "beliefs"?

When we say “belief,” we mean anything you think is true about the world. Something as simple as “the sky is blue” is a belief. So is something as controversial as “GMOs are safe to eat.” Beliefs are not indisputable Truths about the world--you can be very confident in some beliefs, but you should still be willing to change your mind about them based on evidence. If I go outside and see that the sky is purple, I might change my mind about believing the sky is blue. If I read a study that says many GMOs have been linked to cancer, I might change my mind about GMOs being safe.

Many people think their beliefs are infallible. They spend most of their time trying to persuade other people of their beliefs and very little of their time seeing if other people's beliefs persuade them to change their own opinions. But most people also fiercely disagree with each other about many of their core beliefs (X political party is best, Y religion is true, Z is the most important issue facing humanity right now). So it stands to reason that many of them are simply wrong and unwilling to even entertain this possibility. Uncommon Sense is for the people who think they could be wrong about their beliefs, people who seek out contrasting viewpoints to test their beliefs, people who see changing their beliefs as an opportunity to learn rather than as a personal attack.

Who should sign up/is this stuff I already know?

If you are not already familiar with and interested in most of the following concepts, Uncommon Sense might be for you:

  • Bayesian reasoning
  • Applied game theory
  • Collaborative truth-seeking
  • Moral philosophy
  • Epistemology

Here are a few fun little puzzles that should give you a taste of the flavor of questions we focus on:

  • You are in a jury trying to decide whether or not to condemn someone for kidnapping and ransoming an innocent man. Police are certain the perpetrator was one of 30 people staying in a certain motel who match the basic demographics of the kidnapper and have weak alibis. The businessperson then picked the accused out of a lineup. You know a perpetrator gets correctly identified in a lineup about 60% of the time, and any given innocent person gets identified about 10% of the time. Based on this evidence, what are the odds that the accused is guilty? (hint)
  • You are an economics professor who forgot to go to the ATM before leaving for work, and who has only $20 in your pocket. You have a lunch meeting at a very expensive French restaurant, but you're stuck teaching classes until lunchtime and have no way to get money. Can you trick your students into giving you enough money for lunch in exchange for your $20, without lying to them in any way? (source and answer)
  • According to the paradox of voting , voting is never a rational choice for a self-interested individual to make because your vote has a miniscule chance of influencing the election but there is a significant cost to yourself to go out and vote. Yet, it’s possible that even a selfish person should go out and vote. Why might this be? (hint)

Why are you qualified to teach this?

We are a team of Stanford students with experience in computer science, economics, machine learning, and public policy. We have combined our own knowledge and experiences with select ideas from cognitive science, economics, psychology, philosophy, computer science, and forecasting, game theory, statistics, and many other fields. (These various disciplines may all seem disparate, but they all provide useful tools that help instill what it means to think critically.) We regularly run programs to teach these concepts to Stanford students, and we developed our curriculum with advice from people who have run workshops, summer camps, and seminars aimed at teaching critical thinking techniques to professionals, high schoolers, and students at top universities.

But at the end of the day, we don’t claim to be more rational or correct than anyone else. We think we have some ideas people might find useful, ideas we’ve thought about and some ideas we’ve adapted from organizations like CFAR (though we are not affiliated with them). We regularly try new teaching methods and are constantly and iterating, learning, and improving. It seems like students in our programs agree, as does preliminary data we’ve been collecting about the efficacy of our program (though our sample size is small and we can’t implement a RCT). We encourage you to be skeptical of everything we say and call us out if you think we’re wrong about something.

We’d love for you to apply and test the program out! If it’s not useful, you’re free to leave (and maybe give us some feedback!)

Why was I rejected?

We receive many wonderful applications, but unfortunately we can only accommodate a small number of students to maintain our low student-to-teacher ratio. You are encouraged to apply again next year. An application decision reflects only our guess about how good of a fit you would be for this particular program at this time, nothing more.

What does a typical meeting look like?

Students start each day with a five-minute Fermi problem warm-up. Then, we dive into the day’s concept. Each week focuses on adding a different tool to students’ repertoire, and weeks are grouped into monthly themes like “statistics” and “ethical frameworks.” Often, the week’s focus is determined by what students want to learn more about. Examples of weekly topics include: Game Theory as a Dark Art, Cognitive Biases Speedround, Intro to Forecasting, The Economics of Your Career Choice, and How to Read a Scientific Study. Meetings typically include at least ten minutes of discussions in small groups as well about twenty minutes dedicated to activities that explore a concept in-depth. Some sample discussion questions: How Do you think equality is a terminal or instrumental goal?

Activities include: a cold war nuclear simulation, stag hunts, a walk through a map of moral philosophies, a lottery with varying odds, and a chance to forecast when at least 10 million Americans will receive a COVID-19 vaccine.

Why should I sign up for this?

If you love to poke, prod, and question this complicated thing called “reality” and figure out how it works, or if you have something you want to achieve (be that a better world, happiness, or something else), and if you want to join a community of people who strive to become sharper and stronger and better informed, then this is the place for you.

What if I have more questions?

We would love to hear from you! Email us at info@bitbybitcoding.org.

What’s the difference between this program and the school-year program you offer?

The summer seminar is a much more intense, interactive, and rigorous program than our extracurricular offering. The summer program demands a much deeper level of intellectual engagement that includes projects, write-ups, and readings. We feel these aspects are necessary to prepare students to fully understand, utilize, and build off of the tools and concepts we teach. The summer program also fosters a community of talented students who see eachother every day for discussions, games, and collaborative projects. This kind of environment isn’t possible with a school-year program that only meets for an hour a week.

How does the schedule work?

We expect students to carve out plenty of time for or program during the weeks where it runs. The full schedule will be announced soon, but in general, there will be several meetings lasting 30-120 minutes throughout the day. The hope is that this model will minimize “Zoom fatigue.” Some activities will be mandatory, while others will be optional. There will also be asynchronous activities like meetings with groups of other students to work on projects, independent readings, time to write and think independently, and other challenges. We expect there to be about 1 hour of required real-time activities, 2 hours of optional real-time activities, and 2 hours of asynchronous work each weekday.

About nominations

What if I have no one to nominate me?

If you do not know anyone suitable to nominate you (ie if you are homeschooled), please email info@bitbybitcoding.org explaining your situation.

What if I want to nominate more than 10 students?

We recognize some instructors may know an unusual number of good candidates for our program (for example if they run a class on critical thinking). If you feel confident that more than 10 of your students would be an excellent fit for Uncommon Sense, email info@bitbybitcoding.org briefly explaining your circumstances.

What kind of students should I nominate?

Things that might make students a good fit for Uncommon Sense include:

  • Having agency: when faced with a challenge, do they come up with and pursue their own solutions without being prompted or told to do so? Do they work on ambitious projects on their own?
  • Carefully thinking through their values and actions: when they make important decisions, do they know precisely why they are making them?
  • A tendency to optimize for important goals: when they want to do something, do they plan out their actions and consider many options? When shown a better way to do a difficult task, do they adopt the new method? When they decide how to use their time or resources, do they think about the alternative ways they could use these resources?
  • An affinity for truth: do they seek out ideas that challenge their opinions? If they find evidence that contradicts their opinions, do they eagerly change their mind? Do they enjoy learning end exploration?
  • Interest in rationality and cognition: do they want to improve their thinking? Do they enjoy talking about cognitive biases, or reading about psychology, economics, AI, cognitive science, philosophy, or a related subject?
  • A desire to make the world better: do they have a drive to achieve some lofty goal to improve the world? Do they care about the details of this plan and are they taking steps to achieve it?

A good applicant probably possesses at least three of these traits, though there are sure to be plenty of exceptions. These are merely guidelines.

Why do you require teacher nominations to apply?

Our program is not suitable for all students. The pace is fast, the questions are tough, and the work is demanding. We need students who thrive in tackling difficult questions, who value science and knowledge, and who actually want to learn the material we offer rather than to collect a name for their resume.

Unfortunately, trying to figure out which students will thrive in our program is a difficult task.

We could require students to write long essays and wax poetical about bayesian reasoning, but high school students are already inundated with application essays to write, and the best essays tend to only be a sign of good writing ability (and access to good education, familiarity with Standard American English rather than African American Vernacular English, etc). Not of the holistic kind of student we seek.

We could look at test scores, but these measure privilege and test-taking ability as much as they measure academic prowess or intellect. Moreover, many students haven’t taken these exams, and they don’t capture the traits we’re looking for very well.

We could look at grades, but these can vary widely between schools, and again might reflect access to tutors and other resources more than inherent ability. Plus, grades also don’t encompass the traits we are looking for too well.

Instructors already have a holistic view of their students, and we trust their judgement much more than any other metric. Applying this simple filter prevents students who aren’t a good fit from needlessly filling out the application, and might encourage students who are a good fit but feel insecure about their talents to apply. Obviously instructor nominations are not a perfect metric, but they’re the best we have at the moment.