Ways to earn
Article by
Mindrift Team
The most common jobs for PhDs outside academia are in industry research and development, data science, consulting, science writing, and policy. AI training sits alongside these as flexible side income rather than a replacement path. It can be useful during a PhD or postdoc because it pays for skills you already use daily: deep domain knowledge, rigorous checking, and clear written reasoning.
Leaving academia is rarely an easy decision. For many PhD holders, it starts as a question during a postdoc contract, a cancelled faculty search, or a third round of applications for a handful of faculty posts. The question is usually less "should I quit?" and more "what else is my training good for?"
This guide gives an overview of the main alt-ac and industry paths, what transfers from a doctorate to each one, and how long each takes to result in income. It also covers where AI training fits – including as a side activity during a PhD or postdoc – and the rules you should check before starting any outside income while you are still in a program.
The numbers behind the academic squeeze
Academic careers remain a real option, but they are no longer the default destination for new doctorates. According to the U.S. National Center for Science and Engineering Statistics, 40% of 2024 doctorate recipients with definite non-postdoc employment commitments in the United States expected their principal job to be in academia, down from 56% in 2004. The share was lowest in physical sciences (18%) and engineering (14%), as reported in the NCSES Doctorate Recipients from U.S. Universities: 2024 report.
In other words, most PhDs who secure non-postdoc employment at graduation already go somewhere other than a university. Exploring other paths is a mainstream outcome rather than a fallback plan, and treating it that way simplifies the search.
Six paths PhDs take outside academia
Each path rewards a different part of doctoral training. The table compares them on what matters most when you are deciding where to look first.
Path | What transfers from a PhD | Typical entry route | Can you start during a PhD? | Main friction |
|---|---|---|---|---|
Industry R&D | Research design, domain depth, technical methods | Industry postdoc, scientist role, referrals | Internships, collaborations | Competitive, location-dependent |
Data science and ML | Statistics, programming, modeling | Portfolio, technical interviews | Side projects, Kaggle, open source | Need production-level coding skills |
Consulting | Structured problem solving, fast learning | Case interviews, PhD recruitment programs | Case clubs, pro bono projects | Different pace and communication style |
Science writing and communication | Explaining complex topics accurately | Freelance pitches, editorial roles | Freelance articles, blogging | Irregular income at first |
Policy and government | Evidence synthesis, subject expertise | Fellowships, think tanks, agencies | Policy fellowships, advisory work | Long application cycles |
AI training | Domain expertise, rigorous verification, clear reasoning | Platform application and skills assessment | Yes, often project-based and flexible | Project availability varies |
Industry R&D
Industry research suits PhDs who want to keep doing research with a product or applied goal. Pharmaceutical, semiconductor, energy, and materials companies hire doctorates for roles that look familiar but run on shorter timelines and stricter priorities. The hardest adjustment is usually trading publication as the measure of success for impact on a product.
Data science and ML
Data science draws heavily on quantitative PhDs, especially from physics, statistics, computational biology, and economics. The gap is rarely the math – it’s software engineering practice, such as version control, testing, and code that others can maintain. A small public portfolio closes that gap quickly.
Consulting
Management and technical consulting firms value PhDs for their ability to learn an unfamiliar field quickly and structure ambiguous problems. Several large firms run dedicated programs for advanced-degree candidates. The trade-off is pace: projects last weeks, not years, and communication favors short conclusions over complete arguments.
Science writing and communication
Science writing suits PhDs who enjoy explaining their field more than extending it. Freelance writing for magazines, research institutes, and companies is a common entry point, and it can start during a PhD. Our quick guide to freelance science writing covers how researchers get started.
Policy and government
Policy roles need people who can read a body of evidence and summarize what it does and does not support. Science policy fellowships, think tanks, and government agencies all recruit PhDs. Application cycles are long, so starting the search early matters more here than for most paths.
Where AI training fits for PhDs
AI training fits best as a parallel activity. AI companies need doctoral-level experts to write research-grade problems, check model reasoning step-by-step, and explain precisely where an answer went wrong. That’s the same skill set a PhD exercises when refereeing a paper or marking a qualifying exam.
On Mindrift, STEM projects ask contributors to design challenging problems in their field, build reasoning chains, and verify AI outputs – often using Python to check calculations. The specifics differ by discipline.
Check out our domain-specific guides:
Three features make AI training compatible with doctoral life:
Self-scheduled tasks: You choose when to contribute, so it fits around experiments, teaching, and writing deadlines.
No AI background required: Your domain expertise is the qualification. The platform handles the AI training infrastructure.
Low switching cost: There is no need to leave your program or position. Many contributors use it to test whether applied, non-academic tasks suit them.
The main limitation is availability. AI training is project-based – projects open when AI companies commission them and close when enough data is collected, so a given domain can be busy one month and quiet the next. Treat it as supplementary income and a window into how AI is evaluated, not as a replacement for a stipend or regular income.
Rules to check before you start during a PhD or postdoc
Outside income during a doctoral program or postdoc comes with rules that general freelancing advice sometimes doesn’t account for. Checking them first protects your position and your funding.
Review these four areas before applying to any freelance project:
Funding and contract terms: Some fellowships, scholarships, and postdoc contracts limit or require approval for outside paid activity. Your graduate school or HR office can confirm what applies.
Visa conditions: Student and researcher visas in many countries restrict freelance or self-employed income. Ask your international office before you accept any task – this is the most common and most serious oversight.
Research confidentiality: Never use unpublished data, results, or methods from your lab when writing prompts or evaluating AI output. Draw on published knowledge and your general expertise only.
Platform eligibility: AI training platforms set their own location and payment requirements. Our eligibility and geographic restrictions list where contributors can currently join.
Once those boxes are checked, a few hours a week is a realistic starting point that leaves your research timeline intact.
Get inspired!
Check out stories from contributors with academic backgrounds in the Mindrift community.
Article by
Mindrift Team
