What does One-Person Firms, Avatar Livestreams and Task Exposure conclude, in numbers, about “Agent Chains and the Founder's Queue”?

Answer: Chapter 2 shows that errors compound along agent chains (10 steps right 0.95 of the time deliver 0.599), and that the founder answering escalations is a queue that saturates.

Scope of the numbers: Revenues, hours, costs, conversion and exit rates, refund rates and job-loss hazards are illustrative or synthetic; all data are generated by its programs; the book is not financial, legal, tax or career advice, and promises no income. Chapter 3 writes a looping run-of-show as a Markov chain: with illustrative exit rates the best cycle is about 4.0 minutes, with 32.0 orders per 1,000 viewers.

中文版

Data table

The numbers below are printed verbatim by the chapter's Python program and can be reproduced by running it. printed values that match the answer: 0.599, 0.95, 10

pkno checkend checkevery step
0.9550.774 ( 5.0)0.931 ( 6.3)0.974 ( 6.3)
0.95100.599 (10.0)0.815 (13.9)0.949 (12.6)
0.95200.358 (20.0)0.565 (31.9)0.900 (25.1)
0.9850.904 ( 5.0)0.982 ( 5.6)0.990 ( 6.1)
0.98100.817 (10.0)0.952 (11.9)0.980 (12.2)
0.98200.668 (20.0)0.867 (26.2)0.960 (24.4)
0.9950.951 ( 5.0)0.993 ( 5.4)0.995 ( 6.1)
0.99100.904 (10.0)0.982 (11.1)0.990 (12.1)
0.99200.818 (20.0)0.952 (23.5)0.980 (24.2)
(share of jobs delivered right; expected steps of work,
checks included)
Show the full program
# Chapter 2. p03: Reliability of a chain of AI agent steps. A job passes
# through k steps; each step is right with probability p and errors are
# independent. Without checks the job is right with probability p**k.
# A check after a step catches a wrong result with probability 0.9 and
# sends the step back to be redone (costing one more step); a check costs
# 0.2 of a step. Compared: no checks, one check at the end (catching any
# error with probability 0.9, then the whole job is redone at most once),
# and a check after every step. Exact expectations, no simulation.

def no_check(p, k):
    return p ** k, float(k)

def end_check(p, k, d=0.9):
    ok = p ** k
    caught = (1 - ok) * d
    right = ok + caught * ok           # one full redo after a catch
    cost = k + 0.2 + caught * (k + 0.2)
    return right, cost

def step_checks(p, k, d=0.9):
    # per step: geometric redo until right or an error slips through
    q = 1 - (1 - p) * d                # step ends (right or slipped)
    right_step = p / q
    tries = 1 / q
    return right_step ** k, k * tries * 1.2

if __name__ == "__main__":
    print("    p     k   no check        end check       every step")
    for p in (0.95, 0.98, 0.99):
        for k in (5, 10, 20):
            a, b, c = no_check(p, k), end_check(p, k), step_checks(p, k)
            print("   {:.2f}  {:2d}   {:.3f} ({:4.1f})   {:.3f} ({:4.1f})"
                  "   {:.3f} ({:4.1f})".format(p, k, a[0], a[1], b[0],
                                               b[1], c[0], c[1]))
    print("   (share of jobs delivered right; expected steps of work,")
    print("    checks included)")

Single source and citation

Fan, Y. (2026). One-Person Enterprise Automation, Avatar Livestream Funnels and Task-Exposure Arbitrage (Version 1.0.0), Chapter 2: Agent Chains and the Founder's Queue. 5D Library. https://doi.org/10.5281/zenodo.24090235

DOI: 10.5281/zenodo.24090235 · Read the monograph · 5DL-B-090 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026090,
  author    = {Fan, Yafu},
  title     = {One-Person Enterprise Automation, Avatar Livestream Funnels and Task-Exposure Arbitrage},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  doi       = {10.5281/zenodo.24090235},
  url       = {https://doi.org/10.5281/zenodo.24090235}
}

Other chapter answers in this book