I think this a really really cool thing to do, but I want to nitpick about the difference between 0.1 and 0.001%. This doesn't matter very much for an individual but matters enormously socially - it's the difference between needing to fund 1000 and 100000 researchers to have a 50/50 shot at a breakthrough.
It's really great to see someone talk publicly about this. From 2016-2019, I was an engineer at Uber, but I knew that I wanted to found a startup in weird industries like biotech or construction.
How do you get real-world experience in a field that is not your day job?
How do you pivot to another career before you've built all your requisite skills?
I always liked the idea of internships and never understood why it was only for college graduates. In 2017, I went door to door to ask for internships at construction companies and local computational-biology labs. Almost everyone I pitched thought this was unusual, even though I offered to work for free.
Eventually one lab and one construction company said yes and I did internships. This led to a formative experience where I took the domain knowledge of engineering at Uber and problems I saw in construction to form a startup that eventually got into an accelerator (YC) and raised from VCs.
The world is filled with people that want to try different careers and actively explore their curiosities, yet our organizations impede this because it's almost impossible to get part-time/fractional work. What's even more sad is that fractional work is not celebrated, but having a work-life balance is.
For many, you need the stability of full-time job but the freedom of fractional work (for unconstrained exploration) to find your calling. So why is what you did the exception rather than the norm?
Hey Andy, I completely agree. I think a great way that UK universities could overcome their perpetual lack of research funding is by opening up formal part-time programmes to people who are just interested. And same goes for companies as well. And congratulations on your startup!
Hi CasualPhysicsEnjoyer. I love this. I had serious reservations about going into academia for a lot of the reasons you stated; I love chemistry and physics, but I fear losing that passion and motivation to the hierarchy and pressures of academic life. If you check out my Substack you'll notice I've found a way to keep it in my life as I pursue other tracks, but I never considered what you've done here. Part-time research is brilliant, and I'm going to look into it immediately. Do you have any advice about finding a lab to take you?
I have two explanations for the paucity of people doing this. I've been doing "basic ML research" in my free time and published a few papers as a result, even though my "real work" since finishing my PhD has been more "applied ML research". So these two reasons are really just me speaking from experience about the things that frustrate me. (Of course, the extremely positive things, which I won't get into, are why I still do it!)
First, doing research as a hobby creates a timeline problem for collaborations. Potential collaborators (who are likely in academia) will want to move at a faster pace, so they can get published, finish their PhDs, etc. This tends to limit one to working by oneself, and to projects that can be completed by oneself. I suppose the exception would be someone who advises a research project in their free time, but most people who want to do research as a hobby want to do the interesting, hands-on-stuff!
Second, the psychology of research makes it challenging to keep at long term. People don't mind doing intellectually challenging tasks as a hobby: people play difficult games, develop apps, and contribute to open-source tools. But research adds an additional psychological difficulty. In a computer game, there is frequent positive feedback and a decently predictable mapping from inputs to outputs. Neither are true of research, which has long feedback loops and frequently depends on serendipity (ie randomness). And when writings apps, software, and blogposts, you only need to by discovered by a few people on the vast internet who are inclined to like your work. You aren't forced to put your work forward to a small number of people, and hope that by chance, they happen to like it; and otherwise, they will tell you that your work is terrible. But the latter experience is exactly what one gets when one submits one's research to peer review.
I think this a really really cool thing to do, but I want to nitpick about the difference between 0.1 and 0.001%. This doesn't matter very much for an individual but matters enormously socially - it's the difference between needing to fund 1000 and 100000 researchers to have a 50/50 shot at a breakthrough.
It's really great to see someone talk publicly about this. From 2016-2019, I was an engineer at Uber, but I knew that I wanted to found a startup in weird industries like biotech or construction.
How do you get real-world experience in a field that is not your day job?
How do you pivot to another career before you've built all your requisite skills?
I always liked the idea of internships and never understood why it was only for college graduates. In 2017, I went door to door to ask for internships at construction companies and local computational-biology labs. Almost everyone I pitched thought this was unusual, even though I offered to work for free.
Eventually one lab and one construction company said yes and I did internships. This led to a formative experience where I took the domain knowledge of engineering at Uber and problems I saw in construction to form a startup that eventually got into an accelerator (YC) and raised from VCs.
I wrote about these experiences here
https://www.reddit.com/r/startups/comments/hlsj2t/how_i_got_into_yc_twice_with_2_different_ideas_in/
The world is filled with people that want to try different careers and actively explore their curiosities, yet our organizations impede this because it's almost impossible to get part-time/fractional work. What's even more sad is that fractional work is not celebrated, but having a work-life balance is.
For many, you need the stability of full-time job but the freedom of fractional work (for unconstrained exploration) to find your calling. So why is what you did the exception rather than the norm?
Hey Andy, I completely agree. I think a great way that UK universities could overcome their perpetual lack of research funding is by opening up formal part-time programmes to people who are just interested. And same goes for companies as well. And congratulations on your startup!
Hi CasualPhysicsEnjoyer. I love this. I had serious reservations about going into academia for a lot of the reasons you stated; I love chemistry and physics, but I fear losing that passion and motivation to the hierarchy and pressures of academic life. If you check out my Substack you'll notice I've found a way to keep it in my life as I pursue other tracks, but I never considered what you've done here. Part-time research is brilliant, and I'm going to look into it immediately. Do you have any advice about finding a lab to take you?
Hey, I honestly just messaged loads of labs until one took me on! Thanks for your kind words :)
I have two explanations for the paucity of people doing this. I've been doing "basic ML research" in my free time and published a few papers as a result, even though my "real work" since finishing my PhD has been more "applied ML research". So these two reasons are really just me speaking from experience about the things that frustrate me. (Of course, the extremely positive things, which I won't get into, are why I still do it!)
First, doing research as a hobby creates a timeline problem for collaborations. Potential collaborators (who are likely in academia) will want to move at a faster pace, so they can get published, finish their PhDs, etc. This tends to limit one to working by oneself, and to projects that can be completed by oneself. I suppose the exception would be someone who advises a research project in their free time, but most people who want to do research as a hobby want to do the interesting, hands-on-stuff!
Second, the psychology of research makes it challenging to keep at long term. People don't mind doing intellectually challenging tasks as a hobby: people play difficult games, develop apps, and contribute to open-source tools. But research adds an additional psychological difficulty. In a computer game, there is frequent positive feedback and a decently predictable mapping from inputs to outputs. Neither are true of research, which has long feedback loops and frequently depends on serendipity (ie randomness). And when writings apps, software, and blogposts, you only need to by discovered by a few people on the vast internet who are inclined to like your work. You aren't forced to put your work forward to a small number of people, and hope that by chance, they happen to like it; and otherwise, they will tell you that your work is terrible. But the latter experience is exactly what one gets when one submits one's research to peer review.