Chew the frog: is the investment between the investor and the pod, or the investor and the learner

[jeffreywescott]

Notes from meeting that spawned this issue.

Building Blocks

• Roles
• Contracts
• Processes
• Statistics
• Game Objects

  Around Game Roles

• investor
• game maker
• game host

  In Game Roles

• learner
• mentor
• practitioner
• facilitator

  Principles razor

In order for an issue to be marked as passing, the proposed solution must be aligned with our principles. Check off the principles that this solution is consistent with.

• [ ] Cultivate a learning community.
• [ ] Prioritize scale over quality
• [ ] Align incentives
• [ ] Build a team game
• [ ] Trust people's intentions and potential
• [ ] Enable self-organization & self-determination
• [ ] Map feedback loops
• [ ] Optimize for agility
• [ ] Simulation not preparation
• [ ] Sustain the tension between individual and collective gain

[tannerwelsh]

Principles razor

In order for an issue to be marked as passing, the proposed solution must be aligned with our principles. Check off the principles that this solution is consistent with.

• [ ] Cultivate a learning community.
• [ ] Prioritize scale over quality
• [ ] Align incentives
• [ ] Build a team game
• [ ] Trust people's intentions and potential
• [ ] Enable self-organization & self-determination
• [ ] Map feedback loops
• [ ] Optimize for agility
• [ ] Simulation not preparation
• [ ] Sustain the tension between individual and collective gain

[tannerwelsh]

Fleshing out each of the options; would love people to add/edit with clarifications so that we can refine the choices:

Option 1: Investors invest in individual learners

• Investors invest in individual learners by paying for their stipend ($20k).
• Individual learners are responsible for repaying their stipend + 10% fixed interest once they earn > $50k.

External to this is the tuition cost, which is not an investment per se but a credit issued to individual learners for which they are responsible (plus inflation) after meeting salary threshold ($50k).

Unknowns:

• Can learners have their investment revoked? If so, how?
• What is the guarantee on the investment, if any? What happens if learners don't make more than $50k?

Option 2: Investors invest in teams of learners

• Investors invest in teams of learners (pods) by paying for their collective stipend fund ($20k x no. of learners in pod).
• Pods are responsible for repaying their stipend fund + 10% fixed interest once aggregate pod salary is > ($50k x no. of learners in pod).
• By default, individual learners are liable for their share of stipend fund (e.g. 1/12 of total fund for pod of 12), but pods can redistribute the liability based on own preferences and circumstances.

External to this is the tuition cost, which is not an investment per se but a credit issued to individual learners for which they are responsible (plus inflation) after meeting salary threshold ($50k).

Unknowns:

• Can pods have their investment revoked? If so, how?
• Can pods have their investment downgraded (i.e. someone drops out, so they don't need the rest of the stipend fund)? If so, how?
• What is the guarantee on the investment, if any? What happens if pods don't make more than $50k per learner?

[jeffreywescott]

So, I took a crack at the Investor-Learner Contract in a branch in contracts.md.

I believe it does a good job of maintaining our core principles while at the same time mitigating risk for investors.

Assumptions

• Assessment / interviewing are skills that will be acquired and honed during a Learners time in the Guild.
• At 2 months into the program, these skills will not be developed to the degree necessary to commit to financial co-entanglement with other Learners.
• The investor needs a way to mitigate risk.
• Learners further along in the program will have sufficient skills to interview and assess earlier Learners.

[jeffreywescott]

Side-note, @tannerwelsh, I believe we moved from "community" to "collective", but the first principle in the issue template seems to still say "community".

[shereefb]

If the contract is voided at any time during the trial period, any grant money received by the Learner must be refunded to the Investor.

@jeffreywescott not sure I understand this. So, a month into the trial, a player has already recieved $1000. The contract gets voided, and now the player needs to pay the investor $1000 back?

[shereefb]

@jeffreywescott I think this is a great start. Something feels "right" about this in my gut.

A few comments:

1. I think we should move far away from legaleze. These contracts should read like straight and simple talk. Legaleze does not inspire me to play. http://paulgraham.com/talk.html
2. I'm weary of part of the assessment being objective in game stats. It makes sense, but it sets up a relationship between the player and their stats that is judgmental and evaluative. Rather than discerning, useful feedback. Something to consider.
3. I don't we should specify WHO will interview the learner. It's a scaleable idea to have it be "older" learners, but makes me nervous.
4. I like this criteria:

At least 75% of the Learners who have worked with the Learner being evaluated must be willing to join a Team with the Learner (to be determined by a survey of all Learners who have worked with the Learner being evaluated).

Going with Will's comments from yesterday though, what incentivizes me as a learner to say that I'm not willing for someone to be on my team. In other words, I'm a kind person, I don't want anybody kicked out, so I'll say yes to everyone if there's not cost to saying no.

[shereefb]

Riffed off of @jeffreywescott work here: https://github.com/LearnersGuild/learning-os/pull/186

[jeffreywescott]

Yes -- the refunded money is to prevent people from dropping out halfway through the trial. If they stick it out and try, the money is a grant. If not, it's a loan.

[jeffreywescott]

Fair point on legalese. Perhaps "contracts" should be called "agreements" to inspire more straight-forward writing?

[jeffreywescott]

The survey data collected on team willingness could be an anonymized stat during the trial period. If we also collect "confidence", we could make a formula and the 75% could be 10 yeses at 75% confidence or 8 yeses at 100%. Something to play with.

[shereefb]

Resolved by #188

[WillGrant]

This system works over all, but I have two concerns.

1.) The $1000 stipend for the trial period. We are asking people to show up for 2 months, 40 hours a week for $1000 a month. I don't think people can survive on that. Especially not anyone with kids or family they are responsible for. And people with responsibilities are a part of what we are looking for.

Consider that a person has to leave their job to take the risk that they will not pass the trial.

2.) I don't think this solves for the question of why a learner should vote someone down. They don't have an incentive to. It does not cost them anything to let someone slide through. Our metrics -- like relative contribution data -- create an incentive for honesty because their delta impacts their score. What is that mechanism here?

Could we tie the learner to the evaluator -- even lightly? I don't think we have to chain them together, but a light tie would be enough to make people evaluate clearly. Its close to the real world -- if I recommend that my company hires somebody and they can't do their job, I don't have to pay their salary. But it does impact my reputation and prospects for advancement. How do we replicate that?

[tannerwelsh]

:+1: to @WillGrant re: the $1k stipend. I think we should offer the full $2k, give people a real chance.

[shereefb]

Ok, getting rid of trial period, and culling, and 1k stipend! Simplifying. https://github.com/LearnersGuild/learning-os/pull/228

cc: @WillGrant and @tannerwelsh