References

Adversarial collaboration

Why do some people parse a space well and others get snowed?

• Energy landscapes
• Explore/exploit
• Global optimization
• Shape of space vs strategy for optimizing in it

What are situations where you'll need to integrate a large data burden?

• Either data that you don't understand the structure of, or just too much of it.
• Quantum mechanics: different representations mastery can not translate
• Recognition of form
• Domain shear in programming architecture: components that haven't been designed with each other in mind. Domain Driven Design book. Rationality community's double cruxing.
• Umbrella topics: not thinking of the whole as the sum of its parts, e.g. Obamacare

At what point do you need to do work in the space as opposed to just reading about it?

• Build a thing, visit a place, have an experience?
• Podcasting on a thing: what bar is needed for sufficient understanding to talk about it?
• Journalism: you have to actually touch primary source, not just other journalistic products.
• How? Go places, read textbooks, compute. Run numbers (as done often on Slate Star Codex).
• Observation vs Randomized Controlled Trial: build the thing to know what biases are built in.

Potential topic areas

• Moral philosophy
• AI risk
  • Visited OpenAI
• Education
  • My dad
  • Jesse
• Archery
• Science fiction and fantasy
• Physics simulation
  • Should do this!
• Drug development

Preliminary research: How do you do it?

• What to do after first-depth internet search hits limits.
• Make an actual research plan.
• Decide if e.g. you need experts, technical documents etc.
• Start drawing yourself a map.
• Written out decision tree process.

Avi's map: Designing a House

• My goal is to learn what I would need to know to get a house that I would be happy with built.
• Re-read "A Pattern Language" carefully
• Find out the typical procedure for working with an architect - communication style, cost, time.
• Learn what major classes of housing site/lot there are (e.g. terrain type, climate, access, zoning).
• Find a set of example houses that have at least one thing I like a lot to draw parts from.
• Ideally, do Habitat for Humanity or in some other way get involved in the process of building a house.
• Make a list of activities and functions that are important to me for living in a house.
• Locate and speak with a professional architect and ask them what I should learn before embarking on designing a house.

Bugsby's map: Space

• My goal is to learn about the past and present of space exploration enough to present an opinion about the future.
• Skim my old Astronomy textbook.
• Read a history of space exploration.
• Re-read a popular technical analysis of the good and bad things about the shuttle program.
• Find and read a book about the present state of space exploration.
• Find some opposing viewpoint articles about private space companies.
• Learn enough about SpaceX and Blue Origin to say what's different about each of them.
• Read James Michener's "Space" OR re-read Kim Stanley Robinson's "Red Mars" OR read the sequel "Green Mars".

Avi and Bugsby discuss Bugsby's data science process and brainstorm a design for a tool to improve it. There is a great deal of programming jargon and philosophy.

References

Command-line interface (e.g. bash)

Bugsby's workflow

• Gather data
• Process data with commands/scripts
• Feed processed data to model building
• Produce output which may get piped to something else

Python

Visual programming

XKCD - is time spent automating a task worth it 

Semantic web

GUI ("gooey")

"repl" (read-evaluate-print loop) command line

"typed"/"typing"

Caching

Pipes

Macros

Using Dynalist - store instructions in the form of a work log, be able to "tag" sections of it as "programs"

Unit testing

Users/groups/permissions

Transaction log (in the sense of database or file system)

Software liberal

We explore what a morality based on "good enough" means, what it would like, and whether we already follow it.  Discussed: genies, strong AI, and tech startups.

References

Effective altruism

Rationalist community

Satisficing utilitarianism

Innumeracy

Consequentialism and Utilitarianism

Pascal's Wager

Roko's Basilisk

Repugnant Conclusion

Peter Singer

We explore AI Risk, the notion that computer intelligence will cross the human level, start compounding exponentially, and, if not trammelled, mulch everything we hold dear into paperclips.  Why aren't Avi and Bugsby more worried about this?

References

Eliezer Yudkowsky

Public figures who've expressed concern: Elon Musk, Bill Gates, Hillary Clinton

Machine Intelligence Research Institute

OpenAI

Nick Bostrom - Superintelligence

Value alignment problem

Hippocratic Oath in software/civil engineering

Iron Law of Oligarchy

Superintelligence: the idea that eats smart people

Principal-agent problem

SSC AI risk persuasion experiment

EY AI box escape experiment

References

Master of the Senate

Episode 3 - Legacy: the size of a science slice of the pie is shrinking

Rationalists coordinating people on a project article

Technical debt

Kula Sushi LA, Silicon Valley

Influence - Cialdini

Things to Hang on your Mental Mug Tree

Chan Zuckerberg Initiative

Microsoft research announcement around the same time (note from Avi: Chan Zuckerberg was actually by far the more reasonable of these, the opposite of what I said in the episode.  Apologies!)

Endeavors

Moon landing Tennessee Valley Authority Manhattan Project Hoover Dam The Pyramids Human Genome Project Skyscrapers Civil Rights Act Affordable Care Act Standard Oil, Amazon, Google SpaceX Blue Origin Dragon Going to Mars Sea farming Growing a city (Bellevue) Starting a new country Starting a religion Starting a movement Occupy Wall Street/Tea Party Making old age not suck/fighting degeneration

References

xkcd - ranking people

Paul Graham cities essay

Article - The urban archipelago

Iron law of institutions

Iron law of oligarchy

Article about the significance of "again" in "Make America Great Again"

Places

The Barbican

Seguin, Texas

Sisters, Oregon

References

Bret Stephens climate change article

Arguments as soldiers

California wine growers buying land in British Columbia

States and cities offering to step in and honor US climate deals

KSR - global warming as huge engineering project

Reinsurance companies

Bloomberg climate change NGO - covering US portion of UN budget

References

Elasticity of demand

System I and II

Achieving flow

Failure mode of computer scientists in building tools: You Aren't Gonna Need It, Premature Optimization,

Dominion (again) / Isotropic

Overdamped/underdamped control.  For the short version, see this Quora answer.  For the math, see this.

Efficiency practices

Making (and checking) checklists

Series of many small efficiencies: bumping drawers closed with your hip, staying in touch with best balance contact point

Pipelining

Merging steps (in general, streamlining processes)

Retrospection - which behaviors over last week could have been improved?

When successfully correcting an error: what practices would have avoided that error in the first place? What would those practices have cost me, and would it be worth it?

Scheduling - both making good use of time, and avoiding schedule failures that require costly correction

Recognizing inefficient behavior patterns

Main takeaway: the balance between optimizing for a value function and checking whether you need to update the value function is the key to efficiency.

References

Goodreads

Anathem

Pythagorean

Foundation - Asimov

Doberman

Neal Stephenson Clang video

Gentlemen of the Road

The Yiddish Policeman's Union

Clown on Fire

History of vanilla

Lois McMaster Bujold

• Vorkosigan Saga
• A Civil Campaign

Ursula K Le Guin

• The Word for World is Forest Wizard of Earthsea
• Hornblower

Gaudy Night

Lord Peter Wimsey

Among Others

High Fidelity

Neil Gaiman and Amanda Palmer

Very Short Introductions

• Physical Chemistry
• Minerals (crystal habits)
• Psychology
• Existentialism
• Complexity
• Russell

Einstein's Relativity

Chaos - Gleick

Anarchy, State, and Utopia

Rawls A Theory of Justice

C.S. Lewis

• Screwtape Letters
• Narnia
• Space Trilogy

The Magicians - Lev Grossman

Kingkiller Chronicle

Jo Walton's close reading of the Kingkiller Chronicle

Robert Jordan

References

K-Blocks

rm -rf

Hardcore Linux

(fencing accident videos are out there but we're not linking them)

Pressure of an arrow tip - calculations, interesting comparison vs pistol

Chuck/lathe

Top roping

Hang gliding

New Republic

Plato's Republic

Electoral college

SSC reverse advice article

Manic Pixie Dream Girl

Soviet seduction department