TSIS is designed to enable lots of people to work together to solve big problems. As such it needs to optimally employ the knowledge and opinions of lots of people, who are not experts.
For a good summary of what works best, and demonstrating that this is an excellent approach, see the “Wisdom of Crowds” by Jame Surowiecki:
• Similar writing to Malcolm Gladwell and Michael Lewis
• Anecdotal explanations of solid scientific findings
• Should be called the “Wisdom of Wisely Constructed Groups”.
fail us when it comes to making decisions for groups of people:
• Chasing experts is a costly mistake (p XV).
• Uniformity and consensus works against smart decisions (many examples).
Gurus fail often because it is very hard to identify true gurus. With enough experts, some will simply have lucky track records, and self styled experts consistently overestimate how often they are right (p35). Meanwhile there are lots of stories in the book of how groups of experts go wrong with consensus and groupthink, or herding (socially safe but suboptimal choices), p49.
works far better is (1) diversity, (2) independence, and (3) aggregation of
1. Diversity = individuals with different backgrounds and information
2. Independence = each contributor to work and act as independently as possible (pXX), without being nullified by consensus or groupthink (p40-onward).
3. Aggregation = combine opinions after formulating them, rather than seeking consensus, either incrementally along the way (e.g. SARS) or finally at the end (most cases).
The Google search “pagerank algorithm” works on this basis, treating the diverse universe of independent web pages as votes for the other web pages that they reference, after these are themselves weighted. (p16) Google quickly became the top search engine for this reason.
“With most things, the average is mediocrity. With decision making, it’s often excellence. … You could say it’s almost as if we’ve been programmed to be collectively smart.” (p11) However there are necessary ingredients. For instance, it is critical to consider lots of “loser” ideas before making the wisest pick. “Sometimes the messiest approach is the wisest.” (p29)
Diversity is critical to adding new information. Education can often make experts think alike, leading to “too much exploiting and not enough exploring” different alternatives (p31). Even if gurus exist, it is frequently improbable to identify them (p35). The argument is for diversity. “Smart people will not lead you astray, but finding the smartest person will” (p36).
Independence does not mean isolation, but relative freedom to pursue one’s own opinion (p41).
Aggregation systems are critical to combine independent opinions (p74-5). One thing to avoid is sequential aggregation, where people may think that those before them must have more knowledge so they just follow along. This can create an information cascade, or “fad”, but fortunately people naturally learn to avoid this for their most important decisions (p63).
Wisdom of Crowds in the TSIS Project
The TSIS design encompasses all these considerations, such as the subsystems for information sharing, delegation, and activism. They are core assumptions, embedded in design choices.
Chapters 5 offers solid evidence that people can naturally organize themselves to solve problems without top down control. We only need to provide systems to enable this to happen. There are several warnings, but the gist is very positive.
Chapter 7 emphasizes the importance of accountability and trust, especially p115-123, which is addressed by the accountability of identity validation, and reputation communities.
TSIS is designed to accommodate norms and conventions arising to make things work more smoothly over time (p97-100), especially with the accountability that TSIS envisions. In the worst case, that norms can be gently imposed with small tweaks to the design down the road.
TSIS might beneficially add “decision markets” (p19-21), for aggregating opinions and votes. “The most mystifying thing … is how little interest corporate America shows for them.” (p21). What’s not obvious is how to motivate people to participate, while minimizing bluffing to sway outcomes.
The “Wisdom of Crowds” is a resounding case for designs like TSIS to make the most of large numbers of people to solve big problems together!