I regret quitting astrophysics

In 2013 I decided to quit my career in astrophysics, move back “home” and become a data scientist. The blog post I wrote about my decision was probably my best read publication as a professional astronomer and it was moving to read all the reactions from people who were struggling with similar decisions. I meant every word in that blog post and I still agree with most of what I said. Now, 7 years after the fact, it is time to confess: I deeply regret quitting.

This post is meant to give my point of view. Many people who left academia are very happy that they did. Here I present some arguments why one might not want to leave, which I hope will be of help for people facing decisions like these.

I miss being motivated. In the first few years after jumping ship many people asked me why I would ever wanted to not be a professional astronomer. I have always said that my day-to-day work wasn’t too different, except that what I did with data was about financial services or some other business I was in, rather than about galaxies and the Universe, but that the “core activities” of work were quite similar. That is kind of true. On an hour by hour basis, often I’m just writing (Python) code to figure things out or build a useful software product. The motivation to do what you do, though, is very very different. The duty cycle and technical depth of projects are short and shallow and the emphasis of projects is much more on getting working products than on understanding. I am doing quite well (in my own humble opinion), but it is hard to get satisfaction out of my current job.

I miss academic research. The seeds of astronomy were planted at very young age (8, if I remember correctly). The fascination for the wonders of the cosmos has changed somewhat in nature while growing up but hasn’t faded. Being at the forefront of figuring things out about the workings of the Universe is amazing, and unparalleled in any business setting. Having the freedom to pick up new techniques that may be useful for your research is something that happened to me only sporadically after the academic years. The freedom to learn and explore are valuable for creative and investigative minds and it doesn’t fit as well in most business settings that I have seen.

I miss working at academic institutions. The vibe of being at a large research institute, surrounded by people who are intrinsically motivated to do what they do was of great value to me. Having visitors over from around the globe with interesting, perhaps related work was a big motivator. That journal clubs, coffee discussions, lunch talks, colloquiums etc. are all “part of the job” is something that even most scientists don’t always seem to fully appreciate. Teaching, at the depth of university level classes, as a part of the job is greatly rewarding (I do teach nowadays!).

I miss passion and being proud of what I do. The internet says I have ”the sexiest job of the 21st century”, but I think my previous job was more enjoyable to brag about at birthday parties. I can do astro as a hobby, but that simply doesn’t give you enough time to do something substantial enough.

I don’t miss … Indeed, the academic career also had its downsides. There is strong competition and people typically experience quite some pressure to achieve. The culture wasn’t always very healthy and diversity and equality are in bad shape in academia. Success criteria of your projects and of you as a person are typically better motivated in business. The obligatory nomadic lifestyle that you are bound to have as an early career scientist were a very enjoyable and educational experience, but it can easily become a burden on your personal life. The drawbacks and benefits of any career path will balance out differently for everybody. If you get to such a point, don’t take the decision lightly.

The people who questioned my decision to become an extronomer were right. I was wrong. It seems too late to get back in. I think I have gained skills and experience that can be very valuable to the astronomical community, but I know that that is simply not what candidates for academic positions are selected on. On top of that, being geographically bound doesn’t help. At least I will try to stay close to the field and who knows what might once cross my path.

31 thoughts on “I regret quitting astrophysics”

  1. Dear Marcel,
    I left academia and went into data science in 2017 and, word for word, this is exactly the way I feel too.
    Maybe I tend to remember only the good parts of doing research, and I miss them very much…
    But the people who know me tell me that I look happier now, and I think the reason is that the job is much less stressful, at least it’s been so for me. The stakes seem lower: the job is less personal, is not so identity-defining, and even if everything goes badly, there will be another job.
    I know that when I decided to leave academia, it was from a desire to improve other parts of my life. I didn’t feel like my job had to dominate my life, and I feel the same now. I take comfort in this thought.

    1. Hi Chiara,
      Nice to hear from you! I hope you are doing well! You are very right in your comments about the lower stakes. I guess for me my job now feels like a job, whereas back in my astro days it didn’t (and I now realize that that is valuable, to me). Do you still do any research? I have found that doing some, even if it’s a meaningless bit, is fun as a hobby (although having hobbies that don’t involve a screen is great, too). Thanks for your perspective, it helps 🙂

      1. Ciao Marcel!
        I’m doing well thanks! It’s great to hear from you too 🙂
        You’re right, astro didn’t feel like a job, and that is one of the reasons I loved it (but maybe that does change when one becomes a professor). I did some bits and pieces of research after, but not at the moment. Shame because my computing skills have grown A LOT 🙂 I try to do non-technical stuff as hobbies. I got into film making and screenwriting, which is a crazy world and a lot of fun!

  2. Hi Marcel, I made much the same move in 2014 and I really think it’s so difficult to know whether or not to regret the move. There are so many availability biases at work – things often seem nicer at a distance and harder close up.

    I think Chiara really has a good viewpoint too, and some of the benefits of leaving behind the fairly brutal competition and job insecurity (at least early career) can be really hard to estimate if you don’t have access to the counterfactual.

    Finally, being able to roughly guess your age from some of the info in your post, please don’t worry too much – many, many non ex-Astronomers in our stage of life are also looking back on past decisions and wondering whether we can or should regret them. Welcome to the human race, brother, and thank you for the thought-provoking post!

  3. The first black home image was created using AI & Data Science techniques…

    https://techcrunch.com/2019/04/10/the-creation-of-the-algorithm-that-made-the-first-black-hole-image-possible-was-led-by-mit-grad-student-katie-bouman/

    maybe you want to reevaluate your position? Maybe you went to learn more about data science to be a more proficient astronomer? Having a wider understanding of the application of a technique and science can lend very surprising insights.

  4. Thanks for sharing your thoughts! I left 6 months ago after completing two postdoctoral research positions and could not be happier. I have been fortunate and landed a position in a company where I am given almost unfettered freedom to pursue my own goals as a data scientist, which has made the transition feel less jarring. I feel as though I have more time for pure research now, as I have colleagues who take care of the administrative tasks that used to bog me down as an astronomer (someone else presents my results; another person gives me design specs for plots; and a project manager keeps me on task).

    The main thing that I have experienced over these months is a realization that the way in which astrophysics and being an astrophysicist defined my identity was extremely toxic. I now see myself as a much more well-rounded individual: I run, I volunteer, I socialize (online, of course); I play videogames and tabletop RPGs; and I am able to afford nice things and live in a vibrant, cosmopolitan city where I do not feel like an immigrant.

    Perhaps my view will also change in a few years but right now this is easily the best decision I have ever made.

  5. Thank you for writing this. I left neutrino physics to do the same thing in 2013 and this resonated with my own experience. I wish I had advice to offer.

    I try to tell myself the increased financial security and stability are better for me, but the alienation from my work takes its toll on my soul.

  6. Hi Marcel,

    You are not alone (though my background is in atmospheric sciences).

    I’m happy with my (private-sector) job but I’m subscribed to several academic mail list related with open positions because I’m some kind of masochistic (and dreamer). It is almost impossible to find something remote friendly. And there are a lot of offers which think I could be very valuable. It is quite depressing.

    But, as others have said, it is also true that low paid and unstable positions are far from perfect to have a family and sane life balance.

    There are pros and cons and sometimes you are not fully aware until you have lived both situations.

  7. Hi Marcel, I just came across your blog and it was quite a surprise! I’m also a data scientist and ‘extronomer’, and literally just yesterday I published a post on a very similar topic (https://pablorosado.com/posts/is-it-possible-to-go-back-to-academia/).

    I think I agree on most of your points (and those by Chiara), but for me the conclusion is not that I regret quitting academia. I think it’s always hard to find a job that suits your personal and professional expectations. But in industry at least you have a lot more opportunities to find one than in academia. Also, and even though I do love astrophysics, I think there are plenty of ways to do impactful work in industry where you can be motivated, knowing that you are working towards a meaningful goal that you can be proud of.

    1. Hi Pablo,
      Thanks for the link to your post, you hit the nail right on its head! I, too, have done unpublished work (because why do the effort if you move into industry?), that is outdated by now. Going back in a research (and teaching) oriented role is probably diificult (especially when geographically bound). My issue is not so muc about the impact I make with my work (I think I’m doing fine on that front), but rather with an intrinsic interest in the work I’m doing. I’m looking forward to your future posts!

  8. Hello, appreciate your honest and revealing blog post. However, I would encourage you to question your assumptions. You have much to add as an academic with your experience as a Data Scientist. Obviously it won’t be easy, just as it wasn’t easy to become what you have become. But have you really made a decision that has had the impact the way you have interpreted it?

    “It seems too late to get back in. I think I have gained skills and experience that can be very valuable to the astronomical community, but I know that that is simply not what candidates for academic positions are selected on. ” – Are you so certain?

    “On top of that, being geographically bound doesn’t help. ” – COVID-19 impact on this?

    1. Hi! Thanks for your comments. First off, there is no covid impact on me being bound to my current location. Moving up along your questions: if you look at academic job openings, the selection criteria are often related to the length of your publication list, the research funds you have obtained etc., all of which has been standing still for 7 years. So, for positions in actual reserach, that would also need to be fairly close to home, I think I’m pretty sure.

  9. Interesting post, Marcel (and comments, others); and sad to hear that you feel like you might be on the wrong side of the decision to leave astro. In my case I left astro for another part of academic research (epidemiology/statistics) which has been very rewarding in terms of some of those things you miss (namely, time to explore new ideas & enthusiasm for the research). But I also have had incredible amounts of stress surround my continued lack of job security (age 39 and still no chance of a contract >3 years) and isolation from friends and family (due to moving around all the time). If I didn’t come from a middle-to-upper class background there would be no chance I’d be on the property ladder by time. And if I do manage to stay in academia past 40 it seems I must inevitably give up 90% of my research time to spend doing admin, writing grants, managing budgets, etc etc. For example, what proportion of senior professors do you see doing real science versus just pissing their days up against a wall? It’s like a 1-to-5 ratio at least; presumably not because they’re all lazy but because their working lives just get taken over by the non-science side of things and they give up.

    1. Hi Ewan,
      Good to hear from you and thanks for the input! I agree that many professors don’t get to do a lot of science themselves. I do think they are in a scientific environement and touch upon a lot of science (while writing grant proposals, supervising students or sitting in on a lunch talk, for instance). I don’t think I would mind “only” supervising science, and not doing much myself, if that is in a stimulating academic environment.
      I can relate to your comments about private life, security etc. I left in part because of my family and it is also that aspect that makes me not apply to data-science-in-astro jobs all over the world. I think the job would be awesome, but I am not even remotely willing to leave my partner and children for it.

  10. Not too late to get back in to astronomy/astrophysics. One of my colleagues took a ten year break before returning to the world of radio astronomy from working at the Met Office.

  11. This post really resonated with me. I left particle physics in my 20s. I couldn’t see a permanent post on the horizon, and other parts of my life – the prospect of marriage and children – took precedence. I became a teacher. I have regretted not trying harder ever since, although I have enjoyed my job. Now in my 60s, I have just retired, and have found part time work in both physics and education departments in university. It feels great to be back in academia, even as a part-time retiree.

  12. I did exactly the same at the end of my PhD in 1971! I, too, went into computer-related work and spent most of my career in university teaching and researching in the field of computing.

    Life’s quite different, today of course. Permanent jobs for science PhDs are few and far between. I would never advise anybody to follow a career in academic science.

    I kept up my interest in astronomy and after retiring from teaching I was lucky enough to find a job running a public observatory for five years.

    Now, in my late 70s I still study astronomy and am currently analysing data from the Gaia space observatory (using Python!) and have a paper in press at the moment.

    My advice would be to look forward with hope, rather than backwards with regret.

    Good luck.

  13. This post feels very similar to what I lived.

    I got a Masters of Engineering, then worked for a couple of years at a superb physics lab and finally decided to go for a PhD and then became a data scientist.

    As you and others say, the nomadic life was not so appealing on the personal side. Two body problem and such. Besides I had the opportunity to work with some “nomadic” people and I didn’t want to go down that road. I felt that it was a good move to leave academia at 30 instead of at 35 or 40.

    I miss physics and engineering but I also like data science. What I don’t like about my current job is that it’s not challenging at all. It feels like the difference between running and walking and as a consequence I don’t care about my job. I went from designing a particle accelerator to selling parts for machines for a huge corporation. As a consequence, I’m learning a lot of new things and I also have a lot of peace of mind. After all I don’t give a damn about selling more or less.

    There are other things I don’t like, like the corporate/bullshit language but I see it as learning a new language.

    On the other hand, it feels nice that I have the possibility of doing things my way, enjoy free time and get more money. After one year this is good, although I’m already looking for a more interesting job. Let’s see in two or three years.

  14. I think you should try to get back into it. I don’t have an extensive background in academia, but there are a lot of product development sensibilities missing in academia. Bring that back with you, and share the knowledge.

  15. Hi Marcel et al —

    Others — here, and on Hacker News — have referred to the possibility of software or data science jobs “next to” academic astronomy. These are often associated with the upcoming astronomy “megaprojects”: dealing with the likes of the SKA or the Rubin Observatory takes some fairly serious engineering. It’s possible to build a career out of this: labs like SLAC or NOIRLab (or STScI, as I’m sure you know!) in the USA, or SRON and ASTRON in the Netherlands, offer long term positions for people interested in this type of work.

    Whether they’re appropriate for you really depend on what you’re interested in: if it’s pure blue-skies academic freedom and the ability to set your own research agenda, or if its the prestige and job-for-life security of a tenured professorship, they might not be for you. But if it’s the opportunity to stretch your engineering and analytic skills at the cutting-edge of making astronomy happen, these roles are well worth investigating.

    I’ve just taken up a management position at ASTRON, the NL institute for radio astronomy, and also have recent experience with the situation in the USA. I’m not currently hiring, but I’d be happy to chat about prospects in this area with anybody who feels like dropping me a line.

    1. Hi John,

      Thanks for reaching out. In fact, institutes like ASTRON (even though it’s a tad bit far away from home for a daily commute) and SRON are certainly on my list of potential employers where I check the vacancies sites every once in a while. As some have pointed out, maybe I should dig a bit more into job titles that at first glance don’t seem to fit me well. Thanks for your input and enjoy your time at ASTRON! I have visited fairly often and really like the atmosphere there!
      Cheers, Marcel

  16. Hey Marcel, nice post. I’m a data scientist too, and a physicist by studies, I had decided to leave academia after my PhD – it wasn’t due to the academic job market being tight, it was genuinely a preference choice. I understand your concerns and as our common friend Javier (I saw this from a post he put on LinkedIn!) says “there is not much science in data science”. In my case, I’m very happy of my choice and feel like I’ve been growing a lot, I work at a startup though and mine has always been a research-like role with lots of freedom to explore things, and actually set up the data directions of the company. 5 years and counting and I’m very happy. So I kinda found my niche, but I do reckon mine is a lucky situation – this is something I think about a lot hence I was interested in your post. Without the intent to sound dull or patronising (sorry if this does!), but have you considered switching to a smaller business, with better freedom and maybe a research-like role? It won’t be like doing academic research but can give you lots of satisfaction – you mention working in financial services, I don’t have experience of that but I’d imagine it’s a particular industry with lots of constraints.

    1. Hi Martina,
      Thanks for your thoughts and perspective! I have found that research-like positions are hard to find. For most companies, core business is not “finding something out” and to an (understandable) extent, you will have to earn back your money. Good to hear that you have found a lot of freedom in your role! At one previous company, a healthcare insurance provider, I had reasonable freedom, but that still didn’t mean my work environment felt like the academic environments I had been in before. Thanks again for reaching out!

  17. Hi Marcel!

    Such an interesting post and interesting replies. I switched from academic physics to data science in industry 5 years ago. I strongly recognise the feeling of missing intrinsic motivation (but I also enjoy less stress). Those feelings just made me decide to switch: from “data science” in industry to “data science” (or whatever we would like to call it) at RIVM (Dutch National Institute for Public Health and the Environment). This position is partially academical and boosts my motivation. Perhaps this employer could be another institute to put on your list.

    Cheers, Jurriaan

  18. Hi Marcel,
    thank you so much for this post!
    I was just thinking along the same lines this week, although my conclusion is that I probably would have left anyway, sooner or later.
    Motivation, people and the academic environment in general are also what I miss the most. Even if I still work with space and the ISS, I can’t stop thinking that mine is never the less a second rate job compared to what I used to do, even if most people seem to think the opposite.
    I’m glad you mentioned the journal clubs: I use to dread them at some level, but now I would very much love to have them back!
    All the best to you and your family.

Leave a Reply to Mehmet Alpaslan Cancel reply

Your email address will not be published. Required fields are marked *