- Economic abstraction: blockchain architecture can be modified to remove the existence of a single native blockchain token; potentially any number could be used
- Proof of stake:
- Economic abstraction makes calculating economic security more difficult, and increases attack surface on the calculation
- Transaction fees:
- Validators would have to calculate which fees offered are adequate
- Optimize for iteration speed
- Barriers to deploying code and features is super frustrating!
- Leaders need to "commit, explode, recover": commit to a plan, execute, and react to results
- Automate
- Automation must be driven by data and monitoring
- Build the right abstractions
- Build the right underlying framework and use the crap out of it
- Do code reviews right. Period.
- Foster a respectful environment
- People should be comfortable challenging each other
- Shared ownership of code
- Don't build walls between teams; incentivize going across those walls
- 20% time
- You can also use it to incentivize people to learn about other parts of the tech stack, by working on it or documenting it
- There a still a lot of hard problems with significant impact left. That being said, success is not assured; are you ok with a substantial amount of risk and ambiguity?
- Responsibilities are serious. People are motivated; if you set a high bar, someone will try to push it higher.
- Give high performers the room to work on the most interesting and high-impact problems
- Seek truth, rather than follow. Follow the facts, from first principles, and allow your mind to be changed.
- The "no asshole" rule is far too low. Overtrust your colleagues.
- Recognize, award, and admire people who contribute to everyone's success (global vs. local); ownership over the whole vs. your share
- Opinionated on what to build, neutral in who gets access to the results.
- Micro pessimists but macro optimists: problems are everywhere, but things will be better in the future, and we will make positive impact
- Engineers own their impact: everyone is responsible for creating as much value for users and the company as possible
- Hire problem solvers, and leave decision making up to the individuals (but also help them get the most and best quality information to make decisions with via tools and processes)
- Engineers define and prioritize impactful work with the rest of their teams
- Default to information sharing
- Helping others takes priority
- Help your colleagues find leveraged problems to solve
- Goals
- Create a numeric target, to measure effectiveness against
- Teams define their own subgoals and projects on a quarterly basis, guided by overall company strategy
- "We believe in shaping good judgment in individuals instead of imposing rules across the team"
- Empower the individual, align them with your overall strategy
- On new processes or tools:
- Facilitate awareness of the idea or tool, and then let it stand on its own merit. If it .doesn't catch on, it'll organically die
- Values:
- Leave it better than you found it
- Managers are facilitators; get obstacles out of the way (career obstacles, prioritizations, technical help, etc)
- "You just have to speak up when things aren't right, evaluate ideas on their merits, and build things that fix what's broken"
- "We hire people to have an opinion and be creative"
- "Inventing the future requires detaching yourself from the past"; celebrate your work getting replaced
- Objectives: write out what you hope to accomplish such that later you can easily tell if you have reached, or have a clear path to reaching, that objective
- Key results: numerically-based expressions of success towards an objective
- Its up to individual contributors to figure out how to get to these metrics
- Make these flexible enough to allow for a number of creative solutions
- OKRs are a communication framework!
- Meetings and managers are the greatest causes of work not getting done at an office
- Don't default to meetings to communicate; save it for when you absolutely need them!
- "Too many meetings destroy morale and motivation"
- Hill charts are genius