What you're seeing

That moving shape is a picture of how much AI affects a job.

Think of it like a weather radar — except instead of rain, it's showing where AI is hitting your career. Each point on the shape answers one question about your work. The bigger the shape stretches in a direction, the more AI is involved in that part of your job. It moves slowly because the data updates every three months as AI keeps changing.

α

Can AI do this job directly?

How many of the tasks in this job can AI handle right now, without any extra tools. High = AI can already do a lot of it.

β

Could AI tools do it soon?

Even if AI can't do a task today, how much of the job could a simple AI-powered app handle? This is always bigger than the first number.

↗

Are people already using AI for this?

Real data on how much workers in this field are actually using AI today — not predictions, actual usage. Updated every three months.

⇄

Is AI replacing workers or helping them?

There's a big difference between AI doing your job for you and AI making you better at your job. This measures which one is actually happening.

Type in your job above and you'll see your own shape — built from real data for your specific role. Every job looks different. That's the whole point.

When the score is the start, not the end

Go deeper than the diagnostic.

The diagnostic tells you where you stand. The interview tells you what to do right now — tuned to your specific situation.

The Career Diagnostic

$0

Your number, your projections, your tasks at risk, and a free 12-week upskilling email plan.

Score for any of 1,015+ common jobs
1-year, 3-year, 5-year projections
Three tasks-at-risk for your role
Three AI-resistant skills to build
Free 12-week upskilling email plan

Run my diagnostic

Most personal

The Pivot Interview

$29pay once · no subscription

A 12-question interview with the Pivot Engine — our proprietary Career Pivot Agent — tailored to your specific situation. At the end, get a refined score, three ranked pivots built around YOUR constraints, and a 90-day starter plan.

Real conversational interview, not a quiz
Refined score informed by your specific role + situation
Three ranked pivots scored against your constraints
Named programs by role (HIMSS CHDA, ANCC, AHRQ, etc.), named timelines, named next moves
PDF artifact emailed to you for keeping

Start the Pivot Interview →

30-day money back if it doesn't sharpen your thinking.

Why $29 is worth it

01

Grounded in published data

ChatGPT cannot tell you your Eloundou α exposure for your O*NET code. The Pivot Engine starts from a real score and real research, not vibes.

02

Structured 12-question interview

A scripted clinical-style intake produces sharper output than a one-shot prompt. The Pivot Engine pushes back when your reasoning has a hole.

03

A keepable artifact

A formatted PDF you can re-read in six months. Plus the refund — if it doesn't sharpen your thinking, money back.

Diagnostic Pro · launching Q3 2026

Quarterly recompute, weekly digest, private community. $9/mo · cancel anytime.

Why this exists

Every other "Will AI take my job?" tool is broken.

The category is dominated by one site running a 13-year-old academic paper that pre-dates GPT entirely. The newer ones are quizzes built for clicks, not careers. None of them are run by anyone who actually reads the research.

01

The data is current

Re-computed every quarter against the latest Anthropic Economic Index release. Real frontier-LLM usage data on real occupational tasks — not 2013 estimates.

02

The methodology is published

The exact formula, weighting, and source citations live on a single methodology page. Read it. Argue with it. That transparency is the product.

03

Formulated by a physician

Built by Taylor Gardner, DO — board-certified physician with a credential, a name, and a stake in getting the diagnosis right the first time.

How the diagnostic is computed

One number. Four signals. Full transparency.

Your Career Diagnostic blends published occupational data into a single 0–100 number. Higher = more durable. Lower = more exposed.

Eloundou α exposure

The fraction of an occupation's tasks an LLM can directly perform — from the original 2023 OpenAI/Stanford rubric.

Eloundou β exposure

The fraction of tasks reachable via LLM-powered software — the augmentation surface a developer could plausibly build.

Anthropic Index intensity

How heavily real users actually deploy AI on this kind of work today. The freshest signal in the dataset, refreshed quarterly.

Augmentation vs automation

Of that real usage, how much is humans being helped versus humans being replaced. The two paths look identical until you separate them.

Read the full methodology →

Browse the full list

1,015 common jobs and counting.

Each one has its own dedicated page with a deeper breakdown, recommended pivots, and a personalized email plan. New occupations added weekly.

FAQ

Honest answers.

Why is a physician running this for non-physicians?

Because the work that built this — reading research, parsing methodology, separating evidence from hype — is the same work whether the audience is physicians, lawyers, or accountants. The dataset covers 1,015 common jobs across 14 verticals. The methodology is universal. The physician credential is a signal that the analysis is serious, not that the audience is medical.

Is this just another "will AI take my job?" quiz?

No. This is a published score backed by three citable sources, with the formula and weights on a public methodology page. Quizzes generate clicks. This is meant to inform a real career decision.

How accurate is the diagnostic?

As accurate as the underlying public datasets. The model combines them transparently. We update it quarterly as Anthropic refreshes the Economic Index. We publish every change.

Why pay $29 when I could just ask ChatGPT for free?

Three reasons. First, the score is grounded in published data, not vibes — ChatGPT cannot tell you your Eloundou α exposure for your O*NET code. Second, the structured 12-question interview produces sharper output than a one-shot prompt. Third, the deliverable is a PDF artifact you keep — and a refund policy if it doesn't sharpen your thinking.

Will my email be sold?

No. The free 12-week plan is delivered to you, and you'll get the quarterly recompute when your role's score moves. That's the full extent of the contact. Unsubscribe in one click.

What if my exact role isn't listed?

Pick the closest one. The dataset covers 1,015 common jobs and grows weekly. Most physician specialties, most allied-health roles, the major tech and finance roles, and core trades are already in.

Is the score career advice?

It is informational, not personal advice. It is one input into a career decision, not the decision. Talk to a mentor, a career coach, or a colleague before any major move.

Who built this

Real credentials. Real name. Real stake.

Taylor Gardner

DO · Board-certified physician Verified

I build tools for professionals navigating the AI transition. I read every Anthropic Economic Index release the day it drops, every Eloundou follow-up paper, every BLS occupational projection. I've spent a career studying for high-stakes exams and reading peer-reviewed research; I know what real data looks like and what hype looks like. The Career Diagnostic is what I would have wanted when I started thinking about my own future. Read more →