hakimx

METHOD / SELECTED WORKS / DASHBOARDS

The X Lab

Where the questions get worked through: the method behind them, selected works and essays, and interactive dashboards built from real data.

01METHOD / HOW I SOLVE FOR X
GIVENa, b, cfacts · context · constraintsFINDxthe question that matters
Clarity before calculation.

Before solving, define the equation.

A useful answer begins by separating known variables, unknowns, constraints and relationships—then choosing the right method for this particular problem.

  1. 01

    Understand

    What do we already know? Facts, context, constraints, evidence.

  2. 02

    Define X

    What exactly are we trying to solve? Clarity before research.

  3. 03

    Research

    Which evidence, models, frameworks or precedents actually apply?

  4. 04

    Connect

    Combine disciplines. Look for relationships the obvious lens misses.

  5. 05

    Test

    Challenge assumptions against reality. Measure, learn, iterate.

  6. 06

    Decide

    Turn analysis into clarity, direction and action.

02THE X INDEX: SELECTED WORKS

Questions worth
working through.

PROJECT NOTES, ESSAYS AND OPEN QUESTIONS / OUTCOMES TO BE DOCUMENTED

AI ADOPTION

X.002 / AI + HUMAN JUDGMENT

Adoption is only the first question.

Research into generative AI raises questions beyond whether people use a tool. What do they trust it with? How do they evaluate its output? What changes in their own judgment?

GIVEN
A research interest in AI adoption, usefulness, trust and creativity.
UNKNOWN
Which tasks benefit from assistance, and which need stronger human evaluation?
APPROACH
Separate the task, the user and the working context. Organize the literature around concepts and conflicting findings rather than a list of tools.
TO TEST
Compare perceived usefulness with output quality, verification effort and the ability to explain a decision.

Research direction. This page does not present a completed empirical study.

AI / ENTRY-LEVEL WORK / LEARNING

How will AI change entry-level professional work?

The useful question is more specific than whether a job disappears. Which tasks change, how are decisions checked, and what happens to the opportunities through which beginners develop judgment?

What remains unknown

  • Which tasks are delegated to a tool, and which responsibilities remain with a beginner?
  • Does the new workflow provide useful feedback and opportunities to learn?
  • How do supervisors assess work when the process of producing it changes?

A proposed research path

  1. Define the role and work setting before comparing experiences.
  2. Map tasks, supervision and feedback in the existing workflow.
  3. Review relevant primary research and record conflicting findings.
  4. Compare tool-assisted and unassisted task pathways.
  5. Document uncertainty, limitations and what would challenge the initial interpretation.

Evidence to look for

Task diaries, examples of reviewed work, accounts from beginners and supervisors, and evidence of how mistakes are detected and corrected.

Current status

A research brief, with no completed findings or source collection published on this page. The next milestone is a defined role, a bounded question and a documented evidence trail.

HUMAN JUDGMENT

01 / ESSAY DRAFT

What happens when answers become cheap?

An answer has value only in relation to a question. Faster production does not settle whether the question is useful, whether the evidence fits or whether the recommendation can survive contact with reality.

That makes framing a problem part of the work. What is given? What is assumed? What would change the decision? A persuasive paragraph can hide uncertainty just as easily as it can explain it.

A question to carry forward

If producing an answer takes less effort, where should the saved effort go? One possibility: into better questions, better verification and a clearer account of what remains unknown.

03 / OPEN QUESTION

What should a first job teach?

Early work is more than a list of tasks. It can teach someone how to recognize a mistake, assess an ambiguous request and explain a choice to another person.

If tools change the tasks beginners perform, how should that learning happen? Which responsibilities provide practice in judgment, and what kind of feedback makes the practice useful?

This is a question for investigation, not a forecast. The Lab sets out how it could be studied.

Read the research brief

ESSAYS

02 / WORKING NOTE

What if the marketing problem starts somewhere else?

More enquiries are not always the next useful improvement. A delayed estimate, an unclear handoff or a missed follow-up can make additional demand harder to serve.

Before prescribing a campaign, trace one customer's path. What did they ask? Who answered? What happened next? Where did the process depend on somebody remembering?

The working hypothesis is simple: marketing decisions become stronger when they account for the system receiving the demand. The next step is to test where that system loses momentum.

RESEARCH QUESTIONS

X.006 / ENTREPRENEURSHIP ECOSYSTEM / LEBANON PRIME MINISTER'S OFFICE / 2019–2020

Lebanon's entrepreneurship ecosystem.

What would it take for Lebanon to produce ambitious, high-growth ventures consistently? This study applies Daniel Isenberg's entrepreneurship ecosystem strategy to Lebanon, domain by domain.

GIVEN
Isenberg's six-domain model (policy, finance, culture, support, human capital, markets) and Lebanon's ecosystem in 2019.
UNKNOWN
Which domains hold back innovation-driven ventures, and what policy should prioritize.
APPROACH
Desk research across World Bank, WEF, GEM, BDL and Arabnet sources; a six-domain diagnosis; Isenberg's one-venture-per-50,000–150,000-people target applied to Lebanon.
OUTPUT
A research article (December 2019), the study design, and six recommendations.
Read the full study↗︎

X.001 / BUSINESS SYSTEMS

From individual tasks to a connected business.

A local flooring business is a useful place to study a larger question: what happens when marketing, customer service, scheduling and technology are treated as one system?

GIVEN
Customer enquiries, site measurements, estimates, installation scheduling and follow-up all depend on accurate information moving between people.
UNKNOWN
Where does that information get lost, repeated or delayed—and which change would remove the most friction?
APPROACH
Map the path from first enquiry to completed installation. Identify handoffs, clarify responsibilities and connect the estimate, schedule and customer record.
TO TEST
Estimate turnaround, missed follow-ups, duplicate entry and the customer's experience of the handoff.

This note describes the problem and proposed measures. Verified before-and-after results have not yet been published here.

X.003 / MARKET INTELLIGENCE

Understand the market before entering it.

A promising opportunity becomes more useful when its assumptions are visible. Who needs the offer, how do they buy, and what would make them switch?

GIVEN
An offer, an intended customer and limited time or capital.
UNKNOWN
Whether the demand, access to buyers and economics support the decision.
APPROACH
Map alternatives, speak to potential buyers, identify purchasing constraints and separate observed evidence from expectations.
TO TEST
Willingness to pay, buyer access, delivery cost and the assumptions that would change the recommendation.

A working approach to market research; a client-specific case and its results will be documented separately.

READING NOTES

No reading notes published yet.

03F(Y): INTERACTIVE DASHBOARDS

Projects as
functions.

f(Y000)

X.005 / WORLD BANK ECONOMIC DATA / LIVE API + SQL + POWER QUERY

What can income tell us about digital adoption?

  • Who connected, and when?
  • Does income predict access?
  • Who closed the gap?
SOURCE
WORLD BANK API / LIVE
ECONOMIES
US · DE · JP · HU
YEARS
2000–2024
ARCHIVE
24 SEP 2026 SNAPSHOT

Three investigations computed in the browser from live World Bank data, with SQLite SQL and Power Query M that reproduce an archived snapshot. Exploratory case studies, not peer-reviewed research; the four-country sample is illustrative, not representative. No Power BI Desktop report (.pbix) is claimed.

f(Y001)

REGIONAL ECONOMICS / PYTHON PIPELINE / STATIC DATASET

Regional Industry & Cluster Intelligence.

Three-metro industry and cluster analysis using employment specialization, shift-share decomposition, establishment data and regional economic indicators.

  • What makes Detroit's industry mix distinctive?
  • Which strengths does it share with Chicago and Cleveland?
  • Is growth driven by mix or by regional performance?
SOURCES
BLS QCEW · CBP · BEA · ACS · QWI · CPI-U
METROS
DETROIT · CHICAGO · CLEVELAND
CONTEXT
MI · IL · OH
DATA
QCEW 2014–2025 · DATASET 2026.10.22+1083a0d

Built from official BLS, Census and BEA files retrieved on 21 October 2026: QCEW annual averages 2014–2025, CBP 2023, ACS 2024, QWI 2017–2024 (Detroit to 2020), BEA GDP to 2024. Every figure traces back to dataset 2026.10.22+1083a0d, which passed all validation checks and is checksummed. Private employment only; findings come from documented rules and are descriptive, not causal.

f(Y002)

MARKET INSIGHTS / PYTHON PIPELINE / PUBLISHED RESULTS

Electric cars across 50 states, 2025 to mid-2026.

How electric car buying changed in every state and DC when the federal tax credit ended, what explains the differences between states, and what owners complain about compared with similar gas cars.

  • How hard did each state fall when the credit ended?
  • Which state traits go with more electric buying?
  • Do people who search for electric cars buy them?
  • Do electric owners complain about different things?
SOURCES
AUTO INNOVATORS EV REPORTS · CENSUS · FHWA · GOOGLE TRENDS · NHTSA
STATES
ALL 50 + DC
PERIOD
2025 Q1 – 2026 Q2
UNIT
STATE × QUARTER · 306 ROWS

Every figure is the published output of the research pipeline; the page calculates nothing. Missing values are left missing, joins fail loudly, and complaint data is published only as totals and percentages. Describes patterns: not causal, not a sales forecast.