Does anyone have experience working as an independent researcher?
Ive been working in engineering for a few years now. Ive gotten pretty good at my job, and Ive learned a lot. But it was never really my intention to work at a big corporation my whole life.
When I was a kid, on TV there were all these scientists and researchers who just had money to do research somehow. They didnt go to an office or go to meetings, they just had funding somehow to go do science stuff. There was often a big lab built right into their home so they could just wake up and tinker around with stuff. That was the dream for me growing up.
I could always just keep working where I am now, but I cant really do the kind of research I want within the normal structured environment that big companies want me to work in. I want to work on a difficult problem that I would expect to take years of experimentation before I would even hope of making any big breakthroughs.
Im wondering if anyone here has ever done any kind of work as an independent researcher. Like, living off grant money or something like that. Ive been looking at SBIR/STTR grants as a possible first step, but that would only get me 3 years, and after that Id need to find a continued income source.
This trope is based on the antiquated idea of the "gentleman scientist". When such scientists existed historically, they were independently wealthy and thus could fund their exploration into their passions themselves.
A lot of modern science can't easily be done by just one person, and a lot of those meetings you don't want to go to involve discussing your theories and perspectives with other working scientists and experts in their subspecialties who can offer their own advice and ideas when it comes to your work. Having peers to critique your methods before you spend a lot of time and money running experiments and writing a paper is incredibly valuable. Attending seminars and talks about new advancements in your field can also be hugely important and informative.
Even if I were independently wealthy enough to fund my own research career, I would try to affiliate myself in some way with a university lab and constantly attend/submit to conferences in order to benefit from contact with other experts in my field. And my dream research career is in theoretical linguistics, which is almost definitely a field in which it's a lot easier to actually do research and experiments alone than whatever your field of interest is.
Yeah, thats pretty consistent with what Ive seen searching about this question on reddit.
It just kind of bums me out, because I feel like Ive wasted my life. I always figured if I put in the time and worked really hard for years and years eventually I could just get to the point where I would have enough credibility to be able to pursue my own research.
But now I actually have professional experience and some credibility, and Im finding out that it just didnt matter. Its all locked behind a paywall that only universitys and corporations can get through.
So Im going to have to just keep working this job for like 35 more years and then I can retire and finally have the freedom to do what I want at age 65.
You can likely retire much earlier than 65 if that is your goal & you save a significant portion of your income:
These numbers don't make sense to me -- even ignoring the practical considerations when it comes to saving 65% of one's take-home pay, which simply isn't feasible for most people, I think there are fundamental issues with the math here.
Let's first make the simplifying assumption that your total retirement fund is equal to your contributions to it -- that is, your savings rate times your working years times your mean income over those years. In that case, your total retirement fund if you save 5% of your income over 66 working years is 3.3 times your average annual income, whereas if you save 65% of your income over 10 years it is 6.5 times your annual income. However, assuming you die at the same age in both scenarios, if you retire after 10 working years rather than 66, you have to spread out that retirement fund over more years, which means the money you have to live on per-year during your retirement actually goes down.
Now, in reality your retirement fund is an investment account, so the balance will increase due to investment returns not just your contributions. But even ignoring the fact that many retirement accounts don't let you withdraw from them until you reach a certain age, this only further disadvantages someone who retires early but contributes a larger share of their income because of the compounding returns cause time to be a huge factor in the ultimate value of the account. Someone saving 5% of their income in a retirement account for 66 years is going to have a MUCH larger retirement fund than someone saving 65% of their income for 10 years even at a pretty low interest rate, and they further have fewer years they need to spread that money out over.
Obviously it's not never possible to retire early. But I absolutely don't think it's possible with the years/rates you provide here, even if those savings rates did end up being feasible for a given individual (and I think they very very often are not).
Right--it's stupid to save money without also investing it. You'll actually do much worse than this due to inflation.
401k and IRAs all allow withdrawing early (if you're not working for the 401k sponsor) with a penalty, and a SEPP can be set up penalty-free withdrawals.
The table's assumption is that the funds will need to provide income forever. So in that regard, 50 years of retirement is the same as 20.
You're right, there's diminishing returns, and saving 65% of income is difficult for most people. But there's huge returns in going from saving 10% of income to saving 30%. You get 23 years of your life to spend however you please!
You've got this backwards a little bit, and this is what makes this math so powerful. Someone saving 5% of their income for 66 years needs to have a much bigger retirement account to support continuing to be able to spend 95% of their income. But if you're only spending 35% of your income, you need much less savings in order to support yourself.
I know this math is uninutuitive, and that I'm not very good at explaining it. But the numbers are definitely correct. I've re-done the table myself, assuming an average of a 6% rate of return on investments and a 4% withdrawal rate in retirement. I've also chosen an income of $1,000/yr to help bring this ratio into easier to understand units.
...ah, so the assumption you're making is that someone who invests 60% of their income for retirement is going to continue living on 40% of their former annual income. So the options are "live incredibly frugally and never have respite until you die" or "make enough money that you don't need most of it and can spread it out over your entire life comfortably"?
Can you clarify these numbers? When it says "working years", that means stsrting drom age 18-20 or whenever you got your first full time job? So like, 20% means 37 years, but that actually means age 57ish?
Also, does savings rate mean before or after taxes?
There is a joke in academia that R.I.P. stands for research in peace, because of the tendency for researchers to pursue portfolios that advance their career vs their passion projects.
That said, depending on what you are researching, there may be ways to build a research program slowly around your area of interest. All you've described in the post is fundamental blue sky research in a nutshell. Lots of researchers spend years working up to a big breakthrough, but behind the scenes they are following a plan and often an organized progression of activities that support each other. They don't just fiddle until a big breakthrough happens.
What are your areas of interest, research wise?
I want to research nuclear fusion.
Ive been working at an R&D lab for a few years sharpening my embedded systems engineering skills, and in my off time Ive been learning how to do some general fabrication techniques like machining, welding, etc. Ive also been studying up on plasma physics from textbooks, and I had previously done a masters writing simulation code to give me some practice at developing physics engines.
I feel that I have done as much as I can to prepare, and at this point in order to make progress I need to start implementing some of my ideas and building prototypes to test my concepts out. Theres only so much I can do with a desktop computer and reading textbooks in my off hours. I would need to start buying hardware and raw materials, maybe build up a server rack to handle higher resolution simulations. Thats all going to cost money.
Add to that, I probably cant do this at home. If I was intending to not take it seriously, I could build a desktop fusor in my garage. But if I actually intend to really try, I need to have a place to set up where potential neutron radiation wont harm my neighbors. I figure Id need some kind of lab space thats not in a residential zone.
Have you worked as a PI or co-PI before? When I read all your different comments on this thread I don't get the sense that you are very familiar with current research processes.
While there are some competing philosophies regarding research, letting every flower bloom vs methodical pre-review of processes and statistical measures, etc, nuclear energy, shock physics, medicine, etc, are areas that favor the latter. It's not common to win grants in those areas without being able to support how your research will be translated to the next phase of discovery, e.g. mouse model to human, theoretical framework to experiment to prototype system, etc.
No one is going to fund a moonshot with a solo researcher when there are hundreds of qualified labs vying for the same dollars.
I've worked in a variety of research organizations, higher ed, government, etc, and work in an organization now that both receives and issues grants, so I'm happy to answer questions you might have. But it sounds like your first step is really understanding what modern research looks like today.
Edit: and even blue sky research is expected to produce generalizable knowledge, which requires methodical structuring of the experiments.
Ive worked a program as PI at my current job where we partnered with a university to develop a chip architecture from a professor there.
I just dont really want to follow that process on this new thing I want to work on, so Im looking for options to pursure this outside that path.
Years that you are actively employed & saving the percentage in the table.
Yup! Keep in mind though this table doesn't consider things like government benefits, which you'd be eligible for around 67. It also assumes you will be spending the same amount in perpetuity, which is unrealistic when for example, you no longer need an expensive commute or die after 20-30 more years.
After taxes, but if you save in a tax-advantaged account like a 401k, your taxes will go down so in reality it's a bit of a blend.
yeah trust me I feel your pain on this. I long to do research in my field of interest for the rest of my life, but the work-life balance and other sacrifices associated with actually working in academia aren't great even in the best-case scenario, and I'm simply not willing to go through the usual post-doc torture cycle in hopes that I someday get there. But that means I'm basically stuck working on stuff I don't care that much about in industry, which makes me really sad, because I fucking miss when I was doing linguistics research.
It depends on what you want to research.
If you taper your costs and expectations. You don’t have to be retired or rich to work on hobby projects in your spare time with your spare money.
I’m biased since I’m in Computer Science: even nowadays, there are a lot of popular open-source projects built by ordinary people. And some fields like chemistry, biology, and medicine, you can’t really do anything without government and/or corporate support. But the range of fields you can accomplish something cheap is very large. Hardware and robotics (whatever the popular replacement for Raspberry Pi is), history and literature (there’s a ton of online content not paywalled), food science, …
EDIT: Your idea of spending years working on something without even a guarantee of success reminds me of people like Andrew Wiles and Eric Barone, two very different people who spent years working on “moonshots” and succeeded. Granted, they are the exceptions and they both spent full time on their projects, but there are far less well-known projects that I know have been developed by people for years, and although they’ve never taken off, the people developing them seem to make a modest living and their projects have gotten some recognition.
Theres a guy on youtube who is working on his own silicon fabrication workflow. Its pretty crazy to imagine a guy being able to make ICs, even low resolution ones, in his own home.
Well, you do need money, and grants are capricious. What I've always done is be a small business owner, and try to shoehorn my research interests into my work -- was it strictly necessary or efficient to solve so many problems at work with deep RL or CNNs or other technology? No, but it did get the job done eventually, and I managed to fulfill my main interest in learning how to use it, while still paying the bills.
It would have been much harder to justify a lot of decisions I've made over time to an actual employer, but in the end when you come up with a great solution to something, you can sometimes turn that into enough money to do your own projects full time, which is something you also generally can't do as an employee.
I just finished submitting an NIH SBIR for a healthcare startup and have submitted any DoD ones in the past, so I can offer a few thoughts in addition to what @krellor said.
The administrative costs of running the SBIR can be part of the grant proposal, so while there is a learning curve, you can budget for accountants, etc. They also want to see how you will spend the whole grant over the period of performance, including paying salaries that are justifiable. They have published salary ranges for different roles which you can still deviate from, but the more deviation, the better the justification needs to be. So you can't just sock the grant away for a rainy day, you need to be able to spend it as part of a going concern.
That said, there are companies that make a going concern by writing and winning SBIR grants. This involves successfully converting a certain percentage of Phase I grants to Phase II awards, and eventually a certain percentage of Phase II to Phase III. You could look at joining one of these organizations as a way to learn about that aspect of the work. One way to find them is to look at past SBIR awards in the areas you are interested in. That data should all be public.
I think the STTR grants have fewer restrictions on them because they are geared toward basic research, but you pretty much have to have a university affiliation. However, that doesn't have to be you. If you make friends with people in a university setting, you can basically split the grant with them. Often faculty are also looking for collaborators, so anybody coming to them with a funding idea should be worth at least an initial hearing, assuming they have not already planned to submit a proposal as part of another group.
While this all sounds pretty daunting, the cost to give it a try is relatively low, except for your time.
Another avenue to look at is jobs in corporate research labs. While this is still not "citizen scientist with lab in basement", it's going to be more applied than university research, but still early phase. The pressures are going to be similar to academia in the sense of needing to write proposals, win funding (you will have internal business sources in addition to SBIR/STTR).
As far as I have seen, the administrative / grant writing part is a "cost of doing business" everywhere. The caveat to that is that eventually, you might have enough cachet/name recognition that people are coming to you looking to collaborate and be willing to do the scut work.
A final thought: in all these scenarios, the people you approach as collaborators or funders will be asking the question, "why is this person the right person to do this research? Do they have the actual skills needed and do they look credible on paper?" So you could be asking yourself these questions and making adjustments to career, certifications/credentials, and experience in the mean time.
The thing is, I already work for a big corporate R&D lab owned by one of the biggest companies in the industry. I just dont feel like I can do the type of work in the way that I want there.
I find my job frustrating sometimes, because there are these other stakeholders who want to do things a certain way, and I have to design my work around that. And sometimes these external decisions put me in a tough spot where Im struggling to work through an issue.
Sometimes Ill encounter a bug with a strange error message I cant parse. And Ill look it up and find some StackExchange thread that perfectly describes my problem. And the answers offered will be someone saying "well, have you considered completely changing how youre doing this and using this other method?". But I cant use that suggestion, because it wasnt my call to do things the way we are doing things.
Like, these higher up people than me want me to work in this structured way where I come up with an exact plan for how Im going to implement something, and then once the plan is there I cant deviate from it unless I have a really good reason.
Thats just not really how I like to work. I like probing at a problem from a few different angles at once and digging down into the weeds of whats going on under the hood. Which doesn't mesh well with the kind of corporate top down organizational structure we have.
I feel like whether I work for a big corporation and have managers and senior employees, or I work for a University and have advisors and senior postdocs, Im going to be struggling to be able to pursue the leads that I want while collaborating with others.
Maybe you should look at early to mid-stage startups. Like around series A or Series B funding, ~ 30-50 employees. There's enough money for some of this exploration, but they haven't grown a full corporate structure yet. They are going to be focused more on the method than the results, and as the org grows, you might be able to carve a research role with some independence out of the calcifying corporate structure.
Yeah, there are a number of startups in the research space I want to work in, but I really hate the marketing pressures involved in courting investors.
I want to research nuclear fusion, but companies like Helion are out here saying theyre going to put power on the grid by 2030. I just dont really think I buy that, and personally I really hate overpromising and not being able to deliver.
I want to be realistic with people that this will probably be something I spend the rest of my life working on, and even then theres a very good chance it wont ever pan out. But theres no market potential in that, is there?
To work in that kind of environment in a technical capacity, I think you have to be able to divorce yourself from the public claims, and recognize that those promises are marketing for the purpose of fundraising. I did it for ten years at a startup, and the framework I had for myself was that I would stay in it as long as I believed the work I was doing was work toward the long term goal. Once that was no longer the case, I was out.
The truth about startups is that they will form and dissolve as the money ebbs and flows, and the people with expertise will flow from one to the next as they dissolve and reform. At least, this was the case in the autonomous vehicle space. The companies changed, but it was all the same people in different configurations under the hood.
If you were going to make a move to Helion (or whoever), I would do it with this framework in mind – not thinking this company is going to be the be-all end-all, but that you were going to establish yourself in the space you care about, understanding that will open up opportunities and networks later on that might help you advance your personal goals.
I think the citizen scientist picture you painted in your OP is a kind of wish fulfillment. And it's okay to pursue that if the exact form of it matters to you that much. But I think the comments here have given you a range of more realistic avenues to pursue an adjacent goal, if you could transform your dream slightly. So it might be worth some soul searching about what parts of it are most important to you.
Honestly the exact form matters quite a bit to me. Having complete freedom in making design choices is important to me. If Im not going to get to that point in the long run, Im not sure if theres much benefit to moving to a startup.
From the comments Ive received, its sounds like the only way I could realistically expect to work truly independently would be to self fund out of my own pocket. If I was going to get external funding through grants I would think itd be a good idea to publish results and establish a reputation for myself. But if thats not going to be an option and I am just going to be paying for equipment from my own salary anyway, then it wouldnt really matter if I get my name out there.
It might be better for me to just double down on my current corporate job and try to move up the ladder to a point where Im making a lot of money, and then fund my research off that.
For SBIR/STTR they are really looking for something that has a feasible path to commercialization to fund, so they would only work if your interests fit that mold. Additionally, the funding agency gives great consideration to whether the organization they fund has the elements needed for success. They are likely funding small businesses with an existing workforce, sales, etc, with promising initiatives they need help funding. Not a lone individual.
There are also all the challenges of operating a small business. Incorporating is easy, but sustaining an actual legal entity is work. Likewise, grant writing, progress reports, final reports, plus the actual research are significant efforts with timelines.
Which isn't intended to discourage you, bit you would really need to partner with other like-minded people, preferably with complementary backgrounds, to field a team that would look competitive on a grant application.
Alternatively, look for positions in academia. Even if it's not a faculty position, many positions that support research give you the ability to touch a lot of different areas and occasionally field your own project.
I work in a university lab as an undergrad researcher, so my knowledge here isn't complete by any stretch. This comment will apply mostly for physics, though it could be applicable elsewhere.
As other people commented, working independently really depends on what field you're going into—theoretical/computational research is generally free as long as you have a computer and something to do calculations with, while experimental work requires you to spend a lot of money. There's always shed science, like the kind that Explosions and Fire and Styropyro do, but that comes with its own set of caveats—namely that your own safety practices will determine how long you live, or at the very least live without a debilitating injury—and getting payed to do that is a completely different story.
That said, you could get a job as an independent scientific consultant. One of the developers for a package Mathematica package I (unfortunately) had to use a year or so ago, Simon Rochester, has been an independent researcher for a while, doing both academic work and scientific consulting—getting the best of both worlds as it were. You would still need to work for other people, attend meetings, and do all kinds of stuff like that—you know, a bit like your current corporate job—but you would be able to do things at your discretion. Of course, you would probably need some academic collaborators and the like—hypothetically, you could reach out to some old professors and see what they're up to, and then work from there. (I'm considering working as a consultant and faculty researcher somewhere, but I'm not totally sure how I'd do it.)
Working as a theorist independently is a bit iffy. You could do it, sure, but there are some caveats—reaching out to academics is really iffy if you're doing theoretical research, mostly because a lot of so-called "indepdent theoretical physicists" are just crackpots. All I can say to that end is to just not be a crackpot: don't claim to solve all of physics, or disprove x well substantiated theory, or anything like that.
That's the best I can offer at the moment.