# Evolution of Promise Design: Literature Classification Table ![[promise-evolution-diagram.svg]] ## Classification of Literature by Success Probability and Promise Level Functions | Stage | Function Type | P(s) vs φ Relationship | Core Assumptions | Key Literature | Theoretical Contribution | |-------|--------------|------------------------|------------------|----------------|-------------------------| | **M-1** | Zero-order | P(s) = P₀ (constant) | Promises irrelevant to success | • Crawford & Sobel (1982) - Cheap talk<br>• Farrell & Rabin (1996) - Costless signaling | Treats promises as mere signals with no real impact | | **M** | First-order | P(s) = φ | Bigger promises → more success | • Kamenica & Gentzkow (2011) - Bayesian persuasion<br>• Rayo & Segal (2010) - Optimal information disclosure<br>• Grossman (1981), Milgrom (1981) - Disclosure theory | Information design can manipulate beliefs and outcomes | | **M+1** | Second-order | P(s) = φ(1-φ)ⁿ | Balance selling vs delivering | • Kremer (1993) - O-Ring theory<br>• Barlow & Proschan (1965) - Reliability theory<br>• Rivkin (2000) - Complexity strategies<br>• Ethiraj & Levinthal (2004) - Modularity | Operational complexity creates interior optimum<br>φ* = 1/(n+1) | | **M+2** | Bayesian | Beta(μτ, (1-μ)τ) | Precision as design variable | • Raiffa & Schlaifer (1961) - EVSI<br>• Dixit & Pindyck (1994) - Real options<br>• Trigeorgis (1996) - Staged investment<br>• Jordan & Graves (1995) - Process flexibility | Design uncertainty structures<br>(μ*, τ*) joint optimization | ## Evolution of Entrepreneurial Agency ### Agency Evolution Across Stages - **M-1**: Exogenous constraints only (promises don't affect outcomes) - **M**: Subjective choice enabled (choose promise level) - **M+1**: Bayesian learner (update from market signals) - **M+2**: Evolutionary designer (architect of uncertainty structures) ### Core Theoretical Insights #### 1. Traditional Economics (M-1) **"Promises are mere cheap talk"** - No commitment mechanism - Babbling equilibrium - Examples: Internal ventures with fixed budgets #### 2. Behavioral Economics (M) **"Bold promises as self-fulfilling prophecies"** - Promise mobilizes resources - Linear relationship creates maximum promise incentive - Risk: Fraud temptation (Theranos, Nikola) #### 3. Operations Management (M+1) **"Complexity constrains promises"** - Multiplicative reliability: φ(1-φ)ⁿ - Optimal promise decreases with complexity - φ* = 1/(n+1) balances sell vs deliver - Failure: Better Place ignored n>15 complexity #### 4. Our Theory (M+2) **"Design both content and fidelity of uncertainty"** - Aspiration μ: Promise level mean - Precision τ: Distribution concentration - Joint optimization: (μ*, τ*) = (1/(n+1), V·n/[c(n+1)²] - 1) - Success: Tesla's graduated precision (τ: 5→12→30→60) ## Practical Implementation Framework ### Promise Design by Complexity Level | Complexity | n value | Optimal φ* | Optimal τ* | Example Domain | Promise Strategy | |------------|---------|------------|------------|----------------|------------------| | **Low** | n ≈ 1 | 50% | High if V/c large | Software, Apps | Bold promises OK | | **Medium** | n ≈ 5 | 17% | Moderate | Hardware, EVs | Conservative promises | | **High** | n ≈ 10 | 9% | Low initially | Deep tech, Biotech | Minimal promises | | **Extreme** | n > 15 | <6% | Very low | Infrastructure | Incremental only | ### Failure Modes and Success Patterns #### Common Failure Modes 1. **M-1 Trap**: Treating promises as irrelevant (government contractors) 2. **M Trap**: Maximum promises without delivery (WeWork, Theranos) 3. **M+1 Trap**: Ignoring complexity constraints (Better Place) 4. **M+2a Trap**: Fixed high precision → learning trap (can't pivot) #### Success Pattern (Tesla Example) 1. **Start vague**: Low precision (τ ≈ 5) with "200+ miles" 2. **Pay for precision**: Increase τ only after verification 3. **Gradual evolution**: τ progression: 5→12→30→60 4. **Maintain learning**: Keep μ(1-μ)/(τ+1) > 0.01 ### Mathematical Synthesis **Expected Utility Maximization:** ``` E[U] = V_sd × ∫φ(1-φ)ⁿ · Beta(φ; μτ, (1-μ)τ) dφ - c·ln(τ+1) ``` Where: - **Likelihood**: Market reality (sellability × deliverability) - **Prior**: Designed belief structure Beta(μτ, (1-μ)τ) - **Cost**: Information/verification cost c·ln(τ+1) ### Key Propositions 1. **Financial Tightrope (M3.1)**: φ* = (V_sd - V_ns)/2(V_sd - V_snd) 2. **Complexity Ceiling (M3.2)**: φ* = 1/(n+1) 3. **Learning Trap (M4)**: Trap when μ(1-μ) < ε(τ+1) 4. **Optimal Architecture (M5)**: (μ*, τ*) = (1/(n+1), max{0, V·n/[c(n+1)²] - 1}) ## Conclusion: From Object to Subject to Body The evolution from M-1 to M+2 represents a fundamental transformation: - **Object** (M-1): Passive, constrained by external forces - **Subject** (M): Active choice-maker, but naive - **Body** (M+1): Embodies operational constraints - **Evolutionary Subject** (M+2): Designs uncertainty structures > "We endogenize both the *content* (μ) and the *fidelity* (τ) of entrepreneurial claims, unifying persuasion, reliability, and staged experimentation in a single optimization problem." ### The Master vs Slave of Promise - **Slave** (M+2a): Fixed τ, trapped by initial precision - **Master** (M+2b): Choose both μ and τ, design uncertainty strategically This framework transforms promise-making from art to science, providing mathematical foundations for entrepreneurial success.