Skill-tree for lifelong learning

[nileshtrivedi]

See this Twitter thread as well as #14 about related topics.

The idea is to gamilfy lifelong learning via defined skill levels, healthy social competition, and build a demonstrable portfolio of skills and accomplishments over a lifetime.

Some questions:

• Should skills be organized like a university curriculum (a tree/hierarchy) or a graph (directed acyclic, like the one shown in the tweet above)? LearnAwesome's topic nodes can currently have up to two parent topics but the higher-order topics are broad and more general (like, "mathematics") and not sorted sequentially (like, "learn to count").
• How many levels should each node have? Should it be very granular (eg: "Count from 1 to 10")?
• Are the levels one-time challenges or recurring (i.e. habits) ?
  • Habitica is open-source. Could that be taken as a starting point?
• Who defines the challenge at each level? If user-generated, how do we ensure quality?
• How do we evaluate the progress on each level (socially? anonymously? via in-house experts?)
• Should we optimize for time spent with the app (say, for virality)? Can we aim for healthy usage (eg: Wordle)?
• Should the sequence of levels be modeled as a Syllabus in LearnAwesome or a Course? If the latter, this is probably a UI layer on top of existing Course and Topic models.
• How does progress on a skill aggregated into high-level points (such as: health, wealth etc)?
• How do we make this inclusive (language, wealth, gender etc), wholesome and healthy?

[oxalorg]

Amazing questions @nileshtrivedi !!! Here are my biased opinionated thoughts!

Should skills be organized like a university curriculum (a tree/hierarchy) or a graph (directed acyclic, like the one shown in the tweet above)? LearnAwesome's topic nodes can currently have up to two parent topics but the higher-order topics are broad and more general (like, "mathematics") and not sorted sequentially (like, "learn to count").

I prefer a more rigid tree / hierarchy rather than free flowing graphs because it's easier to understand and follow. But bounding reality into trees may not be so simple. If this is hidden complexity then definitely model it as graphs, but expose as subgraphs/trees to players.

If tree paths are already being created by community then adding links between graphs may get complicated? But also fun and unique at the same time. Easy to explore different paths similar to your existing one.

How many levels should each node have? Should it be very granular (eg: "Count from 1 to 10")?

Yes levels should definitely be granular but have a cap (say level 99/100). There is where the competition and motivation to "level up" comes from. Both percentage and percentile based levels wouldn't make much sense. (is there any other way to model levels?)

Are the levels one-time challenges or recurring (i.e. habits) ?

Can be both, maybe one-time challenges for starters

Habitica is open-source. Could that be taken as a starting point?
Who defines the challenge at each level? If user-generated, how do we ensure quality?

Ah right HabitRPG renamed to Habitica! This has been very interesting but here the focus is to "build your own world" or rather build your own trees. We want something more "social" and something more "opinionated" than that. Paths to follow must be common among guilds or multiple people, we don't want every player to make their own skill tree. This needs more thinking for a better solutions.

For starters let people vote. Make guilds have ranks, popular guilds will attract more folks. Popular guilds can create their own skill trees. Guilds should most likely be for like minded people. Eg: Vegan guild is likely to have a few paths I want to follow and level up in. Guilds can share their skill trees with other guilds. Other guilds can also possibly fork the trees if they want to diverge a little (eg: Whole food plant based diet can fork the Vegan trees or vice versa). Guilds can either be open or private.

How do we evaluate the progress on each level (socially? anonymously? via in-house experts?)

I don't like the concept of in-house experts. There must be privacy for users playing the game. Your idea about social evaluation makes most sense to me. So it can be verified either by other fellow adventures or the guilds you are a part of.

Should we optimize for time spent with the app (say, for virality)? Can we aim for healthy usage (eg: Wordle)?

Not sure but the point of the app is to get people to play the game of life, not play the app 🙈

Should the sequence of levels be modeled as a Syllabus in LearnAwesome or a Course? If the latter, this is probably a UI layer on top of existing Course and Topic models.

Hmmm interesting thought. I would say the levels should be modeled to very concrete yet generic goals rather than "specific" goals. But again who is creating these levels? They should have the freedom to choose how to model these

How does progress on a skill aggregated into high-level points (such as: health, wealth etc)?

I don't think we define any high level points by ourself. Winning the game should be different for everyone. Folks interested in maximizing wealth will probably chose different skill trees to excel in. Others might maximize health, some might want to optimize for fun, some for altruism, some for learning, some for wildlife, some for language learning, some for video games, some for cooking, some for ______ (any number of reasons which motivates people / answer to eudaimonia)

How do we make this inclusive (language, wealth, gender etc), wholesome and healthy?

This is also a tough question, but having user profiles + guild chats + friends chat + public profiles are really critical. Making it friendly for everyone probably won't be easy (kids will also be a major target audience).