Adam Zachary Wasserman
Author and researcher: philosophy of science, artificial intelligence, and the metaphysics of complex emergent structure.
Kansas City, MO
Current Research Program
An interlocking two-volume program on the structure and limits of intelligence, grounded in pre-registered controlled experiments on cross-linguistic large language model training dynamics. Volume 1 establishes the epistemological discipline; Volume 2 applies it to the question of where complex emergent structure comes from.
The Robot in the Dark (Volume 1, manuscript in preparation)
Every scientific instrument has an edge where the light runs out. The most consequential moments in science are not what happens inside the illuminated region but what happens at the boundary. The book traces the pattern through evolutionary biology, physics, mathematics, large language models, and finally consciousness itself, arguing that instruments which name their edges become more authoritative, not less. Subtitle: Instruments, Illusions, and the Limits of Knowledge.
The Replicators (Volume 2, in preparation)
Complex emergent structure (biological evolution, software systems, language itself) is more parsimoniously explained by many collaborative intelligences operating across substrates and time than by either single-creator or no-creator accounts. The argument is grounded in the pre-registered controlled experiments listed below and in a structural analogy with working software systems, which are uncontroversially produced by hundreds or thousands of programmers collaborating across decades. Subtitle: Multiple Collaborative Intelligences as the Substrate of Complex Structure.
Layered structure
The two volumes mirror the layered structure of the foundational Substack series Weinstein vs. Dawkins (May 2022, three parts): groundwork first, then the meta-claim that the groundwork makes rigorous. The convergence-of-signals method, the discount-bias diagnosis (in which our use of consciousness as the measuring instrument leads us to discount any agency that does not look like human consciousness), and the silicon-vs-DNA parsimony argument are load-bearing for the Replicators thesis and depend on Robot's epistemology.
Books
The Robot in the Dark (in preparation)
Manuscript draft, 2026. Zenodo deposit forthcoming.
The Replicators (in preparation)
Outline complete; sample chapters drawn from the Weinstein vs. Dawkins three-part Substack series (2022).
Honest Code (commercially published)
Sixteen principles for eliminating hidden state in software. Side-by-side comparisons in twelve programming languages with real nanosecond execution traces.
The Chaos Factory (2017)
The inside story of corporate IT failure. Favorably reviewed by Bloor Research.
Pre-registered empirical research
A pre-registered program of controlled experiments on cross-linguistic LLM training dynamics, with all four lead deposits in place and the child-scale extension accepted for peer-reviewed presentation. Collaborators on the BabyLM 2026 paper and the EMNLP main-conference extension: David Beauchemin (Laval) and Hafedh Mili (UQAM).
The Scaling Hypothesis Is Language-Contingent
Pre-registered controlled ablation training identical 125M-parameter transformers on matched English and French C4 corpora. French achieves grammatical competence at 197M tokens; English remains at chance through 4+ billion tokens. Pre-registration: OSF SJ48B. Training logs: github.com/adamzwasserman/fractal-language.
Right Tool, Right Job: Why Training Language Matters More Than Training Data (BabyLM 2026 / EMNLP 2026)
Child-scale extension. MÉTRON-FR (125M GPT-2-small, 92.5M words of French) reaches 85.97% accuracy on QFrBLiMP at epoch 3. Total project compute cost: $65 USD. Formalizes the Language-Only Hypothesis. Accepted for presentation at EMNLP 2026, Budapest, October 2026 (with David Beauchemin, Laval). The 12-language MÉTRON family extension is in preparation.
English Considered Harmful
Companion preprint to the language-contingent scaling line.
The 70% Rule: When Axiomatic Prompting Helps, and When It Hurts
Pre-registered work on the conditions under which explicit logical axioms in prompts produce qualitatively different outputs on reasoning benchmarks. OSF pre-registrations: 7Z49A, MZF79.
Process Discipline
Methodological preprint on pre-registered empirical practice for AI-assisted enterprise software development.
Writing
- Essential Musings Substack archive, 40+ essays since 2022 working through the philosophical and empirical questions the books bring together. Includes the foundational Weinstein vs. Dawkins three-part series (May 2022) that introduces the Replicators thesis.
- Hackernoon Technology articles
- Welcome to the Jungle Enterprise technology and careers
- Hacker News Historical and conceptual commentary
Open source
DATAOS ecosystem
DATAOS (DOM As The Authority On State) is the operational philosophy underlying the Honest Code book. Its premise is that the DOM, treated as the single authoritative source of application state, eliminates the need for state-management libraries entirely. The multicardz application (~140K lines of production code) is a working operational demonstration of distributed-collaborative-intelligence design principles in a 4E-cognition-compatible system.
| Project | Description |
|---|---|
| DATAOS | Architecture: DOM as the single source of state |
| domx | DOM state observer, under 1KB, zero dependencies |
| genX | Declarative behavior via HTML attributes, under 1KB |
| stateless | React hook that reads state from DOM instead of useState |
| multicardz | ~140K lines, 17 subsystems; spatial tag-based data organization; embodied-embedded-enacted-extended cognition-compatible |
Open Honest
Open Honest is the empirical and operational research program built around the Honest Code principles: a correct-by-construction development framework whose limits are bounded by architecture rather than by hope, and a peer-reviewable measurement instrument for enterprise software quality whose own thresholds are explicit and falsifiable. Eight pre-registered studies (Papers A through H) sit alongside the working code.
| Project | Description |
|---|---|
| Honest framework | Correct-by-construction development framework with multi-language analyzer, four-language tree-sitter detector, and a 276-scenario behavioural specification |
| Honest audit | Four-layer measurement instrument for enterprise software quality; explicit indicator thresholds; instrument edges named |
Other professional work
Buckler AI
AI compliance for regulated markets. Developed the ClearSignal methodology for compliance attestation. Fractional CTO.
Sizzl
Fractional CTO consultancy for enterprise technology.
Background
Forty years of operational practice in enterprise software in regulated industries. CTO of the International Air Transport Association (IATA, the multinational NGO that operates global air-travel standards and the industry's settlement systems). Director of Consulting at CGI, one of approximately 600 directors carrying full profit-and-loss accountability across CGI's global consulting practice, leading the telecoms and health sectors. Senior technical-leadership roles at Bombardier and Yellow Pages.
The 2017 publication of The Chaos Factory began a deliberate pivot from operational leadership into contributing to the body of knowledge. The unifying through-line across the writing since then: distributed multi-agent collaboration across substrates and time is the better explanation for complex emergent structure than either reductionist randomness or singular design.
Self-taught. No formal degree. Unix practitioner since 1986. Built one of the first production web applications in 1996 for the Employee Relocation Council. The work is the credential.