THEORY
The Proxy Seduction Mechanism
Standalone explanation of the PSF mechanism with embedded figure. Covers the precondition
(transformative experience), dual constitution, three-level evaluative erosion, and
diagnostic implications. Written for thesis reviewers, not journal submission.
Mechanism figure (SVG)
Theoretical attributions
Reviewer-facing
Open mechanism page →
EVIDENCE ANALYSIS
The Gap Has Structure
Evidence report with cross-referenced entries. Maps the empirical constellation
against PSF's theoretical claims, showing where perception-reality gaps emerge
across studies, practitioner accounts, and corporate data. Linked version with
navigable citations.
Cross-referenced
Paper-facing
Falsification mapping
Open evidence report →
EVIDENCE TRACKING
Living Evidence Constellation
Interactive hub for tracking the accumulating evidence base. Swim-lane timeline
visualization by category (empirical studies, practitioner accounts, corporate claims,
supporting theory, foils). 101+ entries as of March 2026. Filterable, searchable,
with JSON export/import for portability.
101+ entries
Timeline + cards
5 evidence categories
JSON export/import
Open evidence constellation →
DIAGNOSTICS
Discourse Traps of Proxy Seduction
Diagnostic for identifying how practitioner discourse naturalizes evaluative
capacity erosion. 21 traps across four epistemological categories: 8 inductively
observed, 6 deductively predicted (all confirmed), 3 analytically emergent, 4
cross-domain predicted from AI diagnostics. Each trap includes cross-domain
portability prediction (mechanism-level, domain-bound, or uncertain).
Co-occurrence matrix tracks clustering patterns. Also on the Research Hub for fieldwork use.
21 traps (4 categories)
Co-occurrence matrix
Field coding
Open discourse diagnostic →
PAPER PIPELINE
Paper: Discourse Traps of Proxy Seduction
Concept stub for a practitioner-oriented paper. Genesis: traps discovered inductively
from practitioner discourse, predicted deductively from PSF mechanism and confirmed
independently, extended analytically, then predicted cross-domain from AI diagnostics.
Four contributions: epistemological progression as method, co-occurrence clustering as
compositional grammar, falsifiability architecture with testable predictions, cross-domain
portability analysis distinguishing mechanism-level from domain-bound traps.
Sequences as Paper 2 after AMR.
Concept stub
Paper 2 in pathway
Post-AMR
Open paper concept →
PARALLEL PAPER — MIT SLOAN MANAGEMENT REVIEW
The AI Alibi
Co-authored with Neeti Gupta and Florian Urmetzer. Argues that corporations deploy
AI language as rhetorical cover for workforce decisions driven by financial pressure.
Cases: Block/Dorsey, Amazon, Baker McKenzie, Accenture, Klarna, Commonwealth Bank.
Evidence integrates Cai et al. 2026 MIT activity ontology (activity-level AI footprint),
Massenkoff & McCrory 2026 Anthropic labor market study (occupational-level exposure gap),
and O'Connor/Burn-Murdoch FT analysis (sector-level salary divergence and pipeline hollowing).
Three-pattern structure: partial validity, zero-presence overreach, differential burden
with undisclosed pipeline consequence. Submitted to MIT SMR portal, March 2026.
6 cases
3 evidence layers
~2,500 words
Submitted MIT SMR
With Gupta + Urmetzer
Open AI Alibi draft →
ADJACENT DOMAIN — RESEARCH INTEREST, NOT COMMITTED
AI Diagnostics in India
Research landscape review across radiology, pathology, and clinical laboratory AI in Indian
diagnostic centers. Tracks seven PSF-relevant open questions: outcome measurement gaps (proxy
substitution through sincere belief), deskilling and automation bias (evaluative capacity erosion),
feedback loop contamination (self-concealing degradation), dataset bias as invisible accuracy gap,
and regulatory scaffolding gaps. Potential cross-domain validation of PSF in a regulated,
life-critical context structurally different from the software development mainline.
Potential collaboration with Neeti Gupta.
7 PSF-mapped questions
3 core PSF instances
"Most different" case
Post-PhD pipeline
Open diagnostics tracker →
EMPIRICAL PHASE
PSF Research Hub
Interview protocol (2x2), coding rubrics, candidate tracker, fieldwork tools, and
analysis instruments. Placeholder cards for data repository, pattern tables, dumbbell
visualizations, and empirical findings report. Growing as the research progresses.
11 tools
3 active, 8 planned
~50 interviews
Open research hub →
PHD DELIVERABLE — IS05
Module Research Essay (MRE)
4-page essay linking the ISO5 Qualitative Research Methods module to PSF research.
Centers on the sociomateriality critique (Essay 1, Runde) and its direct connection
to the form/function distinction in the PSF. Covers Cambridge Market fieldwork
(Essays 2 and 3, Leavell) as methodological practice, and reflects on choosing
methods suited to theoretically driven research. Evaluated as part of the First
Year Assessment process.
~1,000 words
IS05 Michaelmas 2025
Included in FYR
Open MRE →
PHD DELIVERABLE — FORMAL ASSESSMENT
First Year Report (FYR)
10,000 to 15,000 words. Systematic literature review, research objectives,
methods, progress to date, and programme of further research with realistic
timing. Includes the MRE. Submission deadline August 28, 2026. Determines
registration for PhD degree.
10K–15K words
Due Aug 28
Includes MRE
Placeholder
ADVISOR MEETING
Progress Update: Barrett (April 2026)
15-slide HTML deck for thesis advisor meeting. Covers theoretical progress
(literature gap, mechanism, evidence constellation), publication strategy
(OS Perspectives + IJMR), and empirical plans (two populations, discourse traps,
coding rubric, pilot design). Speaker notes panel and slide sorter built in.
Reusable 12-column grid presentation system.
15 slides
Speaker notes
Slide sorter
Grid system
Open presentation →
CONFERENCE
IfM PhD Conference Poster (May 2026)
A1 academic poster. Proxy Metrics, Evaluative Capacity, and the Hidden Costs
of AI Engagement. Full theoretical architecture, evidence constellation dashboard,
and research stage overview. Framework unnamed in public materials.
A1 poster
IfM May 2026
Academic