Update on life: why there's no fics · 2:49pm May 3rd, 2017
So I've been unemployed for a bunch of months now. You'd think this would give me lots of time to watch MLP and write fics.
You would be wrong. It gives me lots of time for job-hunting, and being incredibly depressed 
. At least the post-election panic attacks about Nazis have gone down, since they turned out to be incredibly incompetent 
.
The job-hunting, of course, takes up loads of time (fucking tech companies), and has yet to pay off basically at all. Turns out they all think I'm smart and friendly and not at all useful in their office if I don't know web-app development.
What? 
I thought it was a passing fad of the late 2000s! And moreover, I wanted to be an academic the whole time! I was supposed to get a PhD and work as a bloody researcher! Why would I spend valuable credit-hours on learning to shove things into and out of SQL and HTML?
It was my personal life that kept dragging me off the research track in the first place 
.
The other thing that dragged me off that track was fear, which is what I wanted to ask about here. My whole adulthood, people have told me: don't become an academic, it's a road to penury. And I listened: after undergrad, I tried working, and after my Master's, I tried working in industry again.
Notably, both those occasions have eventually led to me being unemployed and bitter. The first time because I was burned-out when I started and actually chose a bad fit for a job. The second time because after some budget cuts to projects, they laid me off almost two years in.
Needless to say: 
Especially since starting in 2013 I actually found a set of topics, or a broad paradigm, that I really want to research: cognitive science and artificial intelligence. And this past year I read some new books that are out, and it all came together in my head. Like, I can feel the designs sitting there in my brain, getting filled in a little more with every advance in the state of the art.
(I solemnly swear that I am up to no goodfriendship-and-ponies. I do not even slightly chuckle, let alone giggle maniacally, let alone brag that I will make the Emperor of Mankind look like the insignificant snit of grimdark he is on my nice universe full of sunlight and happiness. That would give away the surprise long before I can guarantee the surprise will even happen.)
Problem is, now I'm unemployed, living in that one city I deeply despise, with a Master's in the wrong subject, without a PhD or research prospects, trying to figure out what to do. I actually applied to the PhD programs I wanted this year, thinking that some time spent studying and volunteering on a related project would help, as well as a recommendation from my boss. It didn't.
So I know we have some academics on FimFiction. Can you guys tell me if it's really all as bad as everyone says? I mean, I kinda feel like killing myself anyway sometimes, so I might as well do it by working to death in research, right? Like, I don't want to really make money. I want to be able to keep working on the damn research until I can actually implement something. I presented a paper at a conference and actually enjoyed it a lot. I actually give a shit about the scientific method and standards of academic professionalism, when in any other field I just wanna get paid and sit home in sweatpants. See? I should totally be an academic.
I can kinda find labs and possible degree programs for myself. I really need advice on the way back in, and on how to stay on the research track forever once I'm actually in.
As a bonus, ALL HAIL THE PONE-EMPRESS OF PONEKIND!
Something in that much armor should not look that huggable.
4518557 Indeed.
What was the paper on?
4518870 The stuff my advisor had my doing back in 2014-2015ish. I should really write up the second chapter of my thesis from then as a paper.
Hold on tight to your dream.
I'm so sorry.
I agree with everything you said. Death sounds great and unemployment is the worst thing ever. I was unemployed for about three months after graduation in the beginning of the year and it freaked me out. Now I'm just underemployed, which is still a heckofa lot better than nothing.
Every demon you've ever had comes out of the woodwork when you're unemployed. I'm so sorry. And you applied to PhD programs and didn't get in? I'm so sorry, man.
Do not be distraught that you can't seem to create things, especially fanfic, while unemployed. That's normal, and nobody blames you.
I don't have any good advice. I'm so sorry, man.
I abandoned my PhD attempt because my advisor literally told me the problem with my paper was that it didn't have enough Greek letters in it, and later told me that the way to get a paper published was to rewrite the introduction rather than to do the research to come up with some novel content.
I went back into industry so that I would be able to work on something with actual substance. It just shows that you can end up completely wasting your time on either side of the fence. The grass isn't always greener on the other side. However, if you are passionate about one side or the other, even knowing the drawbacks, then I would suggest that is the side you should persue.
Wherever you go, there you are. You may need to improve your technical and/or academic and/or people skills. What is your Masters in, if it's the wrong field? Perhaps you should get one in the right field? I suspect a PhD program would be more likely to accept you if you were continuing out of their Masters program. Plus, that way you'd know the professors and have had time to work them and impress them.
4519376
That's kinda how I feel about research, yeah. Even when considering the downsides, I consider research a valuable job that needs doing within society. When I go to industry, I interview at most tech companies these days and end up thinking, wait a minute, does the world even need this?
My old employer isn't like that, but they just didn't have projects for me anymore.
That is, basically, the plan. I've been waiting for a month now for a professor to decide whether he'll take me on as a research assistant in his cognitive science lab. I've gotten into a Master's program for AI/AGI (that exists, in one place in the world), but had heard very mixed things about the experience. Now I've asked more people and feel less worried about it. I think I can get into a Master's at one of the world's best cognitive science programs.
I think that either of these Master's would help me move forward to a PhD with a good advisor at a good lab. They both do cost money, though, and I'm still torn on whether cognitive science or AI studies offer the best path forward to the goals I want to accomplish. I'm in the universal learning machine camp for my preference in paradigms for studying thought: this means that on the one side, the "more psychological" types disagree with my assertion that we ought to have a general, principled theory of cognition, and on the other side, the "more machine learner" types believe that training task-specific, difficult-to-explain models and wiring them together is how you build a mind.
So on the upside, I have a bunch of psychological, cognitive-scientific, and AI/AGI evidence to cite for what I want to do. On the other hand, any lab I would want to do it in has competitors in its field who will criticize the work for being "fundamentalist" about the belief in universal principles of cognition.
My previous degree was CS with no focus on cognition or ML/AI.
4519172
SO TRUE. You just get depressed, lose all enthusiasm for things, and want to withdraw from the world. You end up feeling like, ok, I'll be fine if I can just find a way to subsist while doing hobby stuff.
Which is mostly a fine point-of-view on life in general, IMHO, but loses out when some of your hobby stuff is difficult large-scale work-y things. When you enjoy building an airplane the hard way, withdrawal from the world can feel relieving while still being depressing.
First off, consider moving somewhere else if you hate the place you're in. Lots of tech companies will hire you from afar. If you have a Masters in anything and know how to code, you shouldn't have difficulty finding employment. Seriously. Any decently-sized city should have a couple dozen firms begging for you, allowing you to test them one by one until you find the perfect fit.
Second, research can be fun but often isn't. Are you sure you want to work on publicly-funded research? Even if you get a Doctorate, academia is hard to break into, and your options there are somewhat restricted. Wanting to do research isn't bad, but it isn't the easiest road to trot.
If what you want to research can be done either as a side hobby or through private funding (i.e. an actual job performing research), those are still valid options. A lot of the AI research today isn't done at universities. It's done at places like Google and Facebook.
Regardless as to the path you choose, it would make sense to build a nest egg by working in industry for a few years. That can free you up from some potential stress, and if you do go into academia, you'll have no end of stress. It will also build a narrative you can use if you do decide to apply to another PhD program.
4520361
This paragraph opens two more cans of worms. The first one is that "where I live" is currently severely constrained by "with whom I live", and that issue has been an issue since about 2010. The person I live with has told me, plain and simple, that they're not ready to move away from where they grew up, may never be so ready, and even thinking about it hurts them a lot.
This doesn't just bug me because I have wanderlust. It bugs me because this isn't where I grew up, it's not where I have any family at all, a whole set of my friends are in another place, and that other place (where I was posting from when I first made my FimFiction account) feels far more like home to me than this place.
But also some wanderlust, fair cop.
The second can of worms is job-hunting. There's a reasonable chance something might come through at this point, but suffice to say, "if you can code, there are companies who will hire you" is just not true. Or at least, it assumes that "you can code" means "you can code for the web". I actually trained as a coder before that was a thing, so my skillset is obsolete. Any new position I get involves some training now.
I did it once. It wasn't easy by any means. It's hard. It's also, in turns, fascinating, wondrous, and valuable to the world. Besides which, half the actual frustration and difficulty comes from your superiors just being mysterious and inscrutable at you: they don't see how you can possibly need so much hand-holding to perform basic tasks and understand research methodology, and you don't see how the hell they expect a three-word email to specify what needs doing. Hilarity ensues.
I'd prefer academia, but would take an industrial lab, national lab, or NGO job if it meant I still got to publish.
That was my life since 2015, actually. Might be this next year, too.
4520420
There's a specific theoretical framework in neuroscience that I think can come a lot closer than we've ever been to both creating actual AI and to prescribing actual goals to it. Since that's a neuroscientific theory first and foremost, if we make an "inverse" form of it, the way they've done for inverse and cooperative-inverse reinforcement learning, we could start to have a serious theory of FAI.
I've actually composed and given a presentation on this. It was confusing and counterintuitive as hell. I can send it to you if you like, though.
The academic programs better suited to what I'm considering either turned me down this year or, in one case, involve dealing with a weird PI. He came up with that theoretical framework I want to abuse, but he doesn't really believe that "value" as such exists at the neurological level at all and he mostly doesn't want to do AI or applied work. He's a theoretician and a biologist before he's any kind of roboticist.
I could try to explain to him that ponies are profoundly important and pony-outputs will skyrocket as my project progresses, but I don't think he'll get it.
Working full-time plus working part-time on research generated two papers in two years. That's a pretty good rate for an amateur, but the second paper was in fact very amateurish and got thoroughly rejected by a journal. I legitimately think I need training and advice to move forward on the things I want to accomplish: I don't think I can just sit in a garage long enough and tinker this into existence.
Actually, getting mad is just my way of dealing with other painful emotions, so it's actually really unhealthy even when the underlying problems or concerns are legitimate.
4523767
.
Hey LW!Fluttershy, thanks for talking this through with me
Both? My big target PhD program was in this area, but, well, the world's top scientific institution is really hard to get into. I didn't even get the interview.
My partner and I have been having a lot of talks and arguments over this. We had a bunch of rough patches when we were younger (like, college kids younger) that made it difficult at the time for either of us to move around too much and remain sure we would stay together. So I ended up moving sometimes, and she ended up moving sometimes, and it was a major pain point.
Now we've gotten more supportive. I'm helping her through a career transition this year and next, and she says if I need to move for reasons I can.
So yeah, I'm gonna try to apply to programs based on what will actually help build up the work I want to do.
I attended AI and cogsci talks at said local university for a while. I need to get on the email list for that stuff myself, now that I'm not receiving the mails automatically at work.
I've actually asked questions at talks by the guy who I wanted to work under, and he said they were the right questions. SEMPAI HAD NOTICED!
I also almost got an appointment with someone who works on Beneficial AI in our area, but he's a pretty prolific academic, so he's damned hard to get a hold of.
Yeah, that email thread happened
.
Part of my current plan is to get through this nice cog-sci textbook that just came out this year, which largely focuses around goals and action (rather than perception, as is typical), note down which subfields are important to the stuff I want to work on (motivation, emotion, goals, action, and representations of the above), and try to find labs to talk to which work on that stuff specifically.
There's sort of a narrow bridge to cross here. If you go "too far left" into cognitive psychology, psychologists want to "let the data speak for itself" by dropping principled models in favor of more ad-hoc ones. If you go "too far right" into machine learning and AI, technologists want to accomplish fixed tasks by learning functions that map a fixed input space to a fixed output space -- dropping the principled modeling of the mind itself.