- Eng at Meta
- Prioritization is the main group-decision making system that is hard
- Pushing for UI vs reducing code complexity
- 5ish people with different thoughts on what to do next
- Manager doesn’t decide; fairly independent atm
- Austin: Explained futarchy via conditional prediction markets
- Q: How to weigh high chance of low payout vs low chance of high payout in decision markers?
- People will be drawn to the high chance markets
[Austin] Food for thought
- How does Manifold do feature prioritization?
- Mostly based on “this is what I think is important and I am interested in”, with some amount of “somebody needs to do this”; “this will conflict with another initiative”;
- Internal (Manifold devs) vs external (users)
- Internal devs understand difficulty of work
- External users understand what they want
- What’s the minimum number of engaged traders for a market to be useful?
- 1: Just tracking down numbers is helpful for clarifying thought
- Even one person can bet against themselves over time
- They can also share the market publicly as a reporting metric
- Austin has been doing this e.g. with
- 2: Two people can place bets against each other
- 3: Need an impartial arbitrator plus two people with a disagreement
- 5: Communication starts becoming a problem
- 10: More classic market; size of a large single software team before getting broken up; two pizzas
- 20: Entire Series A startup, can still know everyone on the team; feels like one company
- 50: Start breaking up into smaller subteams
- 200: