A simple formula, and why guided networks beat going it alone (and beat AI on the hard bits).
Most worthwhile problems are solved outside any single organisation and beyond its immediate network. A simple mathematical formula shows why the chance of success rises sharply when you search using Guided Network Intelligence through a curated community like the Amitypathers. This approach also reaches tacit knowledge – the practical know‑how that current AI cannot tap into.
The simple formula
Let:
-
p = the chance (as a percentage) that one well‑chosen person can give either the solution or a crucial piece.
-
n = the number of people you reach out to.
-
D = a diversity factor between 0 and 1. If your contacts think alike, D is low. If they are varied, D is high. It acts like a discount for overlap.
-
G = a guidance factor (≥ 1). If your search is curated towards people with solved analogues or component know‑how, G is higher.
Define:
-
n_eff = n × D (how many distinct attempts you really get)
-
p_eff = p × G (how useful each attempt is on average, capped at 100%)
Probability of solving:
P = 1 − (1 − p_eff) ^ n_eff
(read “one minus p_eff raised to the power n_eff”)
That’s it. If p_eff is small, you can read this as: more distinct tries (n_eff) and better-aimed tries (p_eff) push P up quickly.
Why single organisations underperform
Inside one organisation (and its close network):
-
n is small.
-
D is low (people share training, tools, and constraints).
-
G ≈ 1 (search is not strongly guided).
Result: modest P even after a lot of internal effort.
Worked numbers
Inside one organisation
-
n = 20 people, p = 2% each (one in fifty).
-
D = 0.4 → n_eff = 8.
-
G = 1 → p_eff = 2%.
-
P = 1 − (1 − 0.02)^8 ≈ 15%.
With a guided network (Amitypathers)
-
n = 150 people surfaced.
-
D = 0.8 → n_eff = 120 (less overlap).
-
G = 2.5 → p_eff = 5% (better fit: solved analogues/component holders).
-
P = 1 − (1 − 0.05)^120 ≈ 99%+.
The jump is not just “more people.” It is more distinct people (higher D) and better‑aimed outreach (higher G).
Why this beats AI on the hard problems
AI systems work best on explicit knowledge – things that are written down, recorded, or abundant in training data. Many important problems hinge on tacit knowledge and situational nuance:
-
Tacit know‑how: “the trick that makes it work” learned through experience but rarely documented.
-
Non‑public constraints: real procurement rules, safety margins, political realities.
-
Edge cases and recombination: stitching together partial answers A+B+C across fields.
-
Trust and permission: who can introduce whom; who will take a call; what is actually negotiable.
Guided networks access this human layer directly. In our formula that shows up as higher G (better fit to the real question) and higher D (exposure to different, experience‑based viewpoints). AI can assist – summarise, draft, compare – but for the decisive bits, people with lived context move the needle.
How we run this at Amitypath
-
Curated reach (↑n, ↑D): We mobilise Amitypathers across domains and geographies to minimise overlap.
-
Guided matching (↑G): We route the question towards people with solved analogues, component ownership, or privileged constraints.
-
Recombination: We treat partial answers as components to assemble—not as dead ends.
Practical use
-
Estimate a sensible p (e.g., 1–5%).
-
Make n large and diverse to lift D.
-
Use guidance to raise G – don’t send generic blasts; target analogues and component holders.
-
Calculate P. If it is still low, increase n, improve D, or sharpen G.
Closing
Most solutions sit beyond the four walls. A guided, diverse search turns low odds into high odds, and it reaches the tacit knowledge that AI alone cannot. That is the work of the Amitypathers – helping you reach further, learn faster, and pay less to get to a real answer.