A lot of websites flirt with career anxiety. Will Robots Take My Job grabs it by the collar and turns it into a number. Type in an occupation and it gives you a percentage risk that automation will take it over. The effect is immediate. Cashiers currently get an 88% calculated automation risk on the site. In the broader high-risk rankings, Financial Clerks, Packers and Packagers, Hand, and Refuse and Recyclable Material Collectors sit at 100%. On the other side, Anthropologists and Archeologists and Emergency Management Directors land at 0%. That bluntness is the hook, and it works.
Table of Contents
What makes the site worth more than a quick doom-scroll is that it is not just a gimmick calculator. Behind the dramatic name is a full occupation database with rankings, charts, downloadable CSV files, wage data, growth projections, public voting, and methodology notes that are clear enough for non-specialists to follow. The site says it started collecting audience views on automation risk in early 2019, then added a broader “job score” in 2021 using labor data, polling, and automation probabilities. It is also refreshingly personal: the About page says the project is run by Mark, who handles the research, updates, and moderation himself.
It turns career anxiety into a searchable database
The smartest thing about Will Robots Take My Job is that it takes a huge abstract argument — what AI and robotics will do to work — and makes it personal in seconds. Search your role, open the page, and you get a compact snapshot: automation risk, a visitor poll, projected growth, wages, and workforce volume. That is a much better web experience than reading a vague essay about “the future of work” and trying to guess whether it applies to you.
The site is strongest when it lets you browse sideways. You start with your own job, then you jump to adjacent roles, then to the low-risk and high-risk rankings, then to the CSV downloads, and before long you are not just asking whether your own role is safe. You are looking at how whole slices of the labor market are being sorted into pattern-heavy work and stubbornly human work. The site’s rankings are built around official labor series and its own model, and the downloadable data makes it feel closer to a public dataset than a novelty quiz.
The number is the bait and the rest keeps you there
The percentage is what gets shared in group chats, but the surrounding context is what gives the site staying power. A job page is not just “robots yes” or “robots no.” For cashiers, the site pairs that 88% calculated risk with an 86% user poll and a projected 9.9% decline in openings by 2034. That combination is what makes the page hit harder. You are not just being told that automation looks plausible. You are being shown a risk estimate next to a labor-market direction.
What the site is actually good at
| Open it for this | What you get | What to keep in mind |
|---|---|---|
| A blunt first check | A percentage-based automation estimate for a specific occupation | It is an occupation-level model, not a prediction about your exact employer or daily tasks |
| Browsing the labor market | High-risk and low-risk rankings that make patterns obvious fast | Rankings can look harsher than real life, where jobs often change before they disappear |
| Practical career context | Wages, projected growth, workforce size, and public voting alongside the score | The data is centered on U.S. occupational classifications and labor data |
| Deeper digging | Free CSV downloads with occupation scores and community votes | The spreadsheet view is useful, but it is still only as good as the underlying model |
That mix is the site’s real trick. It gives you enough data to feel grounded, while still keeping the experience sharp and readable. You can treat it like a curiosity machine, a student career-research tool, or the start of a bigger investigation into how structured a job really is.
The method is rough, serious, and more honest than the name suggests
The site wears its intellectual ancestry openly. Its About page points back to Carl Benedikt Frey and Michael Osborne’s 2013 Oxford Martin paper, the famous study that estimated about 47% of total U.S. employment was at risk of computerisation. That paper has been cited for years in automation debates, and Will Robots Take My Job was originally built around those predictions before moving to in-house calculations.
That part matters because the site does not pretend the numbers fell from the sky. Its current methodology page says the risk score is based on the abilities, knowledge, skills, and activities required for each job, with input data drawn from O*NET. The page even spells out the kinds of traits that make jobs harder to automate: originality, fluency of ideas, social perceptiveness, coordination, active learning, and complex problem solving. That is the difference between empty automation theater and something more useful. The site is trying to translate job design into risk, not just forecast headlines.
It is also careful, at least by internet standards, about its own limits. The calculations page says the score is not perfect and is best used for comparing occupations rather than treating one percentage as fate. It also explains why ratings change over time: updated occupational data, changes in classification choices, bigger datasets than the original Oxford work used, and feedback from more than 100,000 visitor votes. That makes the site feel more alive than a frozen academic artifact. It behaves like a living labor database with a memorable front door.
The best part is where the crowd argues with the model
This is where the site gets very internet in the best way. It does not just publish a model score and leave. It also asks visitors what they think the risk is, then turns those answers into poll results and sentiment charts. The homepage and chart pages even show weighted and unweighted public views over time, including a weighted average that gives more influence to larger professions and a six-month rolling average to smooth the mood swings.
That means the site can show two very different kinds of truth at once. One truth is statistical and structural: what the job appears to require. The other is cultural: what people think is happening to work right now. Those are not the same thing. People often overestimate how fast flashy AI reaches the real workplace, and they sometimes underestimate how quickly routine admin or logistics work gets chipped away by software, self-service systems, and workflow automation. The site makes that tension visible instead of hiding it.
That is also why the percentages are so clickable. They let readers borrow certainty for a minute. Work is messy. Jobs evolve. Tasks split. New tools arrive unevenly. But a clean number scratches a deep itch. The site understands that urge and packages it better than almost anyone else on the web.
It is strongest as a thinking tool, not a prophecy
The biggest risk with a site like this is obvious: people can treat it as destiny. That would be a mistake. The site is built around U.S. occupation data, U.S. labor statistics, and occupational categories that can only approximate the messy reality of a specific company, team, or person. A payroll clerk at one firm may be doing highly routinized work. A payroll lead somewhere else may spend half the day fixing exceptions, negotiating, training, and handling edge cases that the job title hides. The site cannot see that.
That limitation is not a flaw so much as the price of the format. The site works at the level where the web works best: broad comparison, quick pattern recognition, and curiosity-driven exploration. It is less useful as a personal oracle and much better as a prompt. Why is one role high-risk? Which parts of the work look automatable? Which skills keep showing up in safer jobs? Why do some public votes look calmer or more panicked than the model? Those are good questions, and the site pushes you toward them.
There is also a broader point here about the web itself. A site like this feels memorable because it does something the internet still does better than polished platforms: it turns a big, abstract fear into a weirdly addictive niche experience. It is specific. It is a little unsettling. It rewards wandering. It has enough seriousness under the hood to justify the click. That is classic web-gem energy.
FAQ
It depends on the occupation you search. The site’s pages assign different probabilities to different roles. Cashiers are listed at 88% calculated risk, while the high-risk rankings include some occupations at 100%. On the low-risk side, some occupations in the rankings sit at 0%.
The site says its current score is calculated in-house using the abilities, knowledge, skills, and work activities required for each occupation, with data from O*NET and labor-market information added around it. It no longer just mirrors the original Oxford Martin estimates.
No. The site frames the issue as automation risk from robots or artificial intelligence. Its occupation pages ask visitors about replacement by “robots or artificial intelligence,” which fits the way work is actually changing now through software, machine learning, and workflow systems, not just humanoid machines.
No. The site itself says the score is not perfect. It is more useful as a comparison tool across occupations than as a hard prediction for one person’s future. Your exact tasks, industry, employer, and ability to shift into harder-to-automate work still matter a lot.
Jobs that depend more on originality, social perceptiveness, coordination, creative thinking, active learning, and complex problem solving tend to look safer in the site’s model. The low-risk rankings reflect that pattern.
Yes, but with caution. The site’s own About page makes clear that the underlying model and labor data are tied to U.S. occupational sources. That still makes it useful for spotting patterns in work, but the wage, growth, and classification details are U.S.-specific.
Author:
Jan Bielik
CEO & Founder of Webiano Digital & Marketing Agency
This article is an original analysis supported by the sources cited below
Will Robots Take My Job
Official homepage of the project, including public sentiment charts and the site’s core navigation into rankings, charts, and occupation pages.
About
Official project background page covering the site’s origin, creator, polling history, and data sources.
About our calculations
Official methodology page explaining how the automation-risk scores are produced and which job attributes matter most.
Occupation Data CSV
Official data download page describing the site’s downloadable occupation dataset, including risk scores and community votes.
Jobs Most Likely to Be Taken by Robots
Official ranking page showing the occupations with the highest automation-risk scores.
Jobs Least Likely to Be Automated
Official ranking page showing the occupations with the lowest automation-risk scores.
Will Cashiers be replaced by AI & Robots
Official occupation page used as a concrete example of how the site combines calculated risk, polling, and growth data for a single job.
The Future of Employment
Oxford Martin School page for the Frey and Osborne study that shaped the early public conversation around computerisation risk.