Medicine
Clinical reasoning
Diagnostic annotation, evaluation, and high-stakes judgment where being almost right is still wrong.
The Intellectual
Side Hustle
For a growing number of professionals, side income is no longer about escaping their expertise. It is about extending it into the age of AI.
Once upon a time, a side hustle meant work outside your vocation. What is emerging now is different: a market in which accountants, lawyers, clinicians, recruiters, strategists and a host of other vocations are paid to apply their judgement, knowledge, and professional discernment to help train and evaluate AI systems.
Continue02 — The shift
This is not
a gig economy story.
The dominant public story about AI is automation. But the deeper story is that as systems become more powerful, the value of verified, contextual human judgment rises with them. The next bottleneck is not only compute. It is expertise.
StructuralNot cyclical
PermanentCategory formation
03 — Definition
The Intellectual Side Hustle is the emerging opportunity for subject matter experts to use their professional knowledge to help train, evaluate, and improve AI systems.
The key word is intellectual. This is not built around generic hustle. It is built around applied expertise becoming economically portable beyond the boundaries of a person’s main job.
04 — Identity
The Human Data Expert
A Human Data Expert is a subject matter expert who uses deep domain knowledge to train artificial intelligence systems, often being paid more than they earn in their current vocation.
“The people AI needs most are often the people who never thought of themselves as part of the AI economy.”
Identity framing for accountants, lawyers, clinicians, teachers, recruiters, consultants, and other domain experts.
05 — Opportunity landscape
Where expertise
meets demand.
Law
Jurisdictional nuance
Interpretation, case analysis, and rule-sensitive reasoning that general models still mishandle.
Finance
Commercial context
Risk modelling, regulatory context, controls, and the judgment behind professional decisions.
Language & culture
Nuance at scale
Translation, local context, cultural interpretation, and the difference between fluency and precision.
06 — Live market signal
The roles
emerging now.
This is already a functioning market. The strongest proof is not hype. It is the fact that distinct role types are already appearing across the human data ecosystem.
01
Domain Annotator
Specialists who evaluate and label AI outputs in high-stakes fields. Credentialed. Verified. Compensated accordingly.
02
Red Team Expert
Authorities who probe model outputs for professional-grade failure modes that remain invisible to generalists.
03
Knowledge Architect
Practitioners who structure expertise into training-ready taxonomies, edge cases, and evaluative frameworks.
04
Evaluation Lead
Senior contributors who design rubrics and oversee quality at scale across specialist annotation programmes.
07 — Institutional validation
Even LinkedIn
wants a piece
of the pie.
When the professional internet starts testing an AI labour marketplace, this stops looking like a fringe startup story. It starts looking like a category.
Market signal
LinkedIn is testing the category in public.
According to the uploaded article, LinkedIn confirmed early testing of an “AI labor marketplace”, described AI training as one of the fastest-growing jobs in the US, and listed roles reaching up to $150 an hour across domains including finance, nursing, linguistics, red teaming, and software engineering.
$150/hrTop listed rate
Multi-domainFinance, nursing, coding, linguistics
The Wall Street Journal frames AI training jobs as part of the new labour picture.
Uber AI Solutions shows even familiar platform operators entering skills-based AI work.
09 — Intelligent scepticism
Fair questions.
Question 01
Am I training myself out of a job?
It is a fair concern. But this market exists because intelligent systems still need people who understand how real work is actually done. AI creates pressure, yes, but it also creates demand for human judgment in the places where being correct really matters.
Question 02
Is the money real?
Sometimes yes, sometimes overstated, and often dependent on the platform, role, and domain depth involved. The correct posture is discernment. There are visible compensation signals, and there is noisy marketing. The task is to separate them.
Question 03
Are these platforms legit?
Some are serious. Some are messy. Some are excellent for certain profiles and poor for others. A real market can still be uneven. That does not make it fictional. It makes navigation important.
“I hate side hustles. If you have side hustles, that means your main hustle isn’t working.”
10 — A serious objection
Prof Galloway is right
about most side hustles.
He is just not describing this one.
Professor Galloway's warning is a responsible one. Many side hustles are exactly what he is criticising: scattered effort, weak focus, and economic activity that looks productive without building real leverage.
But The Intellectual Side Hustle is not a random second hustle layered on top of a struggling primary career. It is the extension of real professional expertise into a new market created by AI.
A lawyer is still using legal judgment. A clinician is still using clinical reasoning. A recruiter is still using assessment skill. What changes is the commercial surface area of that expertise. It becomes portable, visible, and newly valuable inside the human data economy.
This is not distraction from mastery. It is mastery finding a second market.
11 — Next step
The next step is not hype.
It is orientation.
Once you understand The Intellectual Side Hustle, what you need next is a curated map: which platforms exist, which opportunities are credible, which routes suit your background, and how to tell signal from noise.
Explore the landscape