How Path works, and where it falls short
Path is a planning engine built by Feynman Labs. It works out your odds for every route and simulates thousands of futures to see when you could get there, builds your plan in order against your real calendar, grounds it in public data, and runs a safety system on every conversation. A large language model from our AI provider writes the words. This page explains each part, what the odds mean, the limits you should know about, and how to tell us when an answer is wrong.
The short version
- The Path engine, its data and its safety system are built by us. The words come from a large language model from our AI provider. It is an AI, not a person, and it can be wrong.
- The odds are a simulation run on your device. They are estimates to steer by, not predictions or promises.
- Facts come with a source you can open. Numbers without a source are marked "estimate".
- Check anything important, like requirements, deadlines and costs, with the school or official body.
- If something looks wrong or unsafe, tap Report this answer.
1. The Path engine
Path is built by Feynman Labs. The engine, its data and its safety system are Path's own; the words come from a large language model from our AI provider:
- The simulator. For every route, your device runs thousands of simulated futures (between 4,000 and 48,000) to find when people get there, and works out your odds from each stage's pass rate and tries. It runs again every time you move a slider. See What the odds mean.
- The planner. Path builds your plan in sections, in order: it checks your goal, lays out routes, finds your edge, maps schools and milestones, builds your daily plan and gives a verdict. Each section builds on what the earlier ones found.
- The calendar. The engine works out your real calendar: your age and school year today and, in each country Path is open in, that country's school year and, where we have it, its application cycle. Every section is written against that calendar and the same consistency rules, so ages, dates and deadlines line up across your plan.
- The data. 23 public data sources, from government wage and job-outlook statistics to occupation classifications, universities, programmes and school calendars, kept current by a daily refresh job, a fact-checked facts pack for 22 popular goals, and Path's own checked list of scholarships, grants and student aid. See Path's data.
- The safety system. Crisis detection on every chat message, each country's verified help lines, age rules for each country, and a report button on every answer. See The safety system.
- The writer. A large language model from our AI provider, named on Service providers, writes the text: routes, the edge, the daily plan, the verdict and chat replies. Path's server gives it the instructions and the data to work from. The app only sends your details, never the instructions, so every request carries Path's safety rules. Only the Edge and Map sections may search the web for current information (a few searches per request).
We don't train AI models, and we don't use your plans or chats to train any AI model.
2. What happens in each part of the app
| Part | What happens | Where its facts come from |
|---|---|---|
| Goal check | Sums up your goal and sets targets (grades, study, sleep, exercise, experience). | Facts pack |
| Routes | Lays out 3 or 4 real routes to the goal, each as stages with pass rates. | Facts pack |
| Edge | Finds moves, open doors, time windows and risks. | Facts pack, Path's checked scholarships, grants and student aid, and web search |
| Map | Suggests schools and milestones. | Colleges near you (College Scorecard in the US, ROR elsewhere), programmes with graduates, pay and debt (College Scorecard field of study data, US), facts pack and web search |
| Daily plan | Builds a weekday and weekend routine and habits to check in on. | Your plan |
| Verdict | Picks a route, gives the strongest argument against it, and says what would make you switch. | Your simulated odds from the Path engine |
| Discover | Turns your experience into skills and roles that fit. | O*NET job data (including related jobs and interest profiles) with BLS wages, outlook and AI exposure categories and fields of study (O*NET Education crosswalk) in the US, or your country's occupation and pay data |
| Pitch | Drafts a résumé summary and an outreach message. | Your plan |
| Reading photos | Reads subjects and grades from a transcript, or text from a résumé. | Your photo |
| Chat (your "twin") | Answers questions about your plan and helps when you get stuck. | Your plan and the AI's general knowledge |
3. The safety system
Path's safety system works on every request: rules the AI must follow, crisis detection on every chat message, each country's verified help lines, age rules for each country, and a report button on every answer.
It's an AI, not a person
Nobody at Path writes your answers or reads your chats as they happen. To make that clear:
- The app reminds you that you're talking with an AI when you start, and again every hour if you're 13–17 (with a suggestion to take a short break), or every 3 hours if you're 18 or older.
- Chat coaches like this may not be suitable for some people under 18. If you're 13–17, we suggest going through your plan with a parent or guardian.
- The chat doesn't remember past conversations. It only sees your plan and the last few messages of the conversation you're in.
- The AI follows Path's safety rules on every request. It must not recommend anything unsafe, illegal or unhealthy, such as crash diets, skipping sleep, performance-enhancing drugs, dangerous training or faking documents.
- If a chat message suggests you might be in crisis, the app shows help lines (in the United States, such as 988) before the AI replies (see below).
- For people aged 13–17, the AI keeps everything age-appropriate, stays away from dating, sexual content, drugs, alcohol, vaping and gambling, and suggests involving a parent or guardian for decisions about money, travel, meeting people or physical safety.
If a message sounds like a crisis
- What is checked. Every chat message, before the AI answers, in two stages: Path's own rules on our server, then a second check by the language model, which reads the message in context. We look for signs of suicide, self-harm, abuse or immediate danger.
- What you see. A short, caring message and help lines you can call or text with one tap. In the United States: the 988 Suicide & Crisis Lifeline, the Crisis Text Line, the Childhelp National Child Abuse Hotline and 911; if your plan is in Oregon, YouthLine too; if you're 13–17, also Teen Line and The Trevor Project. In the United Kingdom, Ireland, Canada, Australia, New Zealand and the United Arab Emirates: that country's own lines, verified against each service's own site or a government page (in the United Arab Emirates, 800-SAKINA's nationwide service also by a call to the line), with their pages checked automatically every day, or by hand every month where a site blocks automatic checks. The AI itself is told to put the plan aside and point you to the same help.
- What we keep. Only a daily count of how often help lines were shown, split by age band and US state, or, outside the United States, by age band, country and region. It has no link to you and none of your words.
- What we don't do. Nobody watches chats as they happen, and we don't contact anyone. Path is not an emergency service. In the United States, if you are in danger, call 911. Outside the United States, if you are in danger right now, call your local emergency number.
Age rules in each country
Each country Path is open in has its own minimum age, and the age screen applies the one for where you live: 13 in the United States, the United Kingdom and Canada (14 in Quebec), 15 in Australia, 16 in Ireland, and 18 in New Zealand and the United Arab Emirates. People aged 13–17 get the stricter rules above, time limits with breaks and lower daily limits.
4. What the odds mean
How they are made
- The engine has the AI lay out each route as a series of stages, for example "degree", then "application", then "selection". Each stage gets a pass rate, a number of years and a number of tries.
- Where a published figure exists (for example "10 picked from more than 8,000 applicants"), the pass rate is based on it and links to the source. Otherwise it is marked as an estimate.
- The Path engine on your device then adjusts each stage's pass rate for your current grades, habits and experience, and works out the chance of getting through every stage, counting the tries each stage allows. That chance is your odds.
- It also runs each route thousands of times (between 4,000 and 48,000 simulated tries) to see when people get there. The ages at which they reach the goal give a typical age and a range (the middle 80% of results).
This runs on your device, so when you move a habit slider the odds update straight away.
How to read them
- Use them to compare. They are best for comparing routes and seeing which habits move the number most.
- They are not about you as a person. They don't know about luck, health, money, family, discrimination, rule changes, or the many things you haven't entered.
- They are rough. Many pass rates are estimates, and small differences (say 4% against 5%) don't mean much.
- They haven't been checked against real outcomes yet. We plan to measure how well they match what really happens and publish the results.
- A low number is not a "no". Hard goals have low odds for almost everyone. The plan is there to show what would raise them.
5. Path's data and how facts are checked
The facts pack
Path keeps a "facts pack" of sourced facts for 22 popular goals, such as astronaut, nurse, chef, lawyer and game developer. Each fact is a short claim in our own words, with the publisher, the year and a link to the page that states it.
When the facts pack is refreshed:
- An AI researcher re-checks each goal against live sources, preferring official ones: government agencies, regulators, official statistics, the governing body of the sport or profession, and peer-reviewed research.
- A separate AI fact-checker opens every cited page and marks each fact as supported, corrected or unsupported.
- Unsupported or unchecked facts are dropped. If a goal ends up with fewer than 10 facts, it keeps its previous version.
- A new version goes live only after automatic checks pass: every goal present, at least 10 facts each, and every link secure.
Public data
Path draws on 23 public data sources. Job skills, education levels, related jobs and interest profiles come from O*NET, wages, job outlook and AI exposure categories from the U.S. Bureau of Labor Statistics (for U.S. occupations only; an AI exposure category compares occupations and is not a forecast of job loss), and college admission rates, prices and graduation rates from the U.S. Department of Education's College Scorecard. In the United States, the fields of study that prepare people for a job come from the NCES Classification of Instructional Programs and the O*NET Education crosswalk, and each programme's graduates, pay and debt from College Scorecard's field of study data. For people outside the United States, Path adds occupation names from ESCO, pay and job figures from the International Labour Organization, national statistics offices and Eurostat, universities from ROR and school years from the UNESCO Institute for Statistics. We load them from the official files, and a daily refresh job keeps them current: each source is reloaded on its own schedule, from weekly to every three months. See Data sources.
Scholarships, grants and student aid come from Path's own checked list: Path checked each item on the provider's own page, and most pages are read again every night for key facts, hiding an item whose key facts or page are gone; a few pages that need JavaScript, and flyers, are re-checked by hand about once a month, and an item that cannot be confirmed for 35 days is hidden. On an item from Path's list, the dates and link the app shows come from the list. The AI is told not to restate amounts or dates in its own words, but it can still make mistakes, so check the item and the provider's page. See Scholarships, grants and student aid.
How to tell them apart in the app
- A source tag links to the page a fact comes from.
- "estimate" marks a number the AI worked out without a direct source.
- College records from College Scorecard or ROR are marked as data.
- A scholarship, grant or student aid item from Path's checked list links to the provider's page with Checked on and the date it was last confirmed.
Pages found by web search (in Edge and Map) are linked, but they are not checked the way the facts pack is. When no fact covers something, the AI uses its general knowledge, which may be out of date.
6. Limits
- The AI can be wrong, out of date, or sound more certain than it should.
- It can misread a photo. Check the grades or text it reads before you rely on them.
- It was built first for the United States. In the United Kingdom, Ireland, Canada, Australia, New Zealand and the United Arab Emirates it uses each country's school year, help lines and data, and its application cycle where we have it, but rules there may be less complete. It works in English.
- It only knows what you tell it, and it can't check it.
- It may reflect biases in the data it learned from and in its sources. If you spot this, please report it.
- It is not a doctor, counselor, lawyer or financial adviser, and it is not an emergency service.
- There are daily limits on AI use to keep Path free, so it may pause until tomorrow.
- Path's list of scholarships, grants and student aid is a short list it has checked, not every one there is. So far the items it lists for the United Arab Emirates are for UAE nationals only, and course outcomes are shown only for US programmes.
7. How to report an answer
- Tap Report this answer next to the answer.
- Pick a reason: wrong, harmful, unsafe, offensive or other.
- Add a note if you like (up to 500 characters), then send.
Each item on Path's list of scholarships, grants and student aid has its own flag, Report this listing. It sends only which item it is and the reason you pick, and nothing goes to the provider.
A person on our team reviews reports every week and uses them to fix prompts, facts and safety rules. What we keep with a report is explained in the privacy policy. You can also write to us at legal@feynmanlabs.com.
Reports are not watched in real time. In the United States, if you or someone else is in danger, call 911, and for suicide or self-harm, call or text 988. Outside the United States, if you or someone else is in danger right now, call your local emergency number.
8. Your data and the AI
Your plan details, chat messages and any photo you add are sent to our AI provider, named on Service providers, only to get the answer you asked for. Your account number is not sent. We don't save chat messages or photos. For the details, see the privacy policy and service providers.