So, I've started working my first "real job" last month, and it's pretty decent. Good benefits, decent pay, strong union despite being tech, and for reasonable hours (6 per day). The problem is that I took this job mainly so I can continue grad school. Currently I'm finishing up my master's, so I'm managing to conciliate doing both OK since I don't need to be in uni premises for anything anymore, but I'm unsure about being able to do a PhD later.
I figure once I work for a few months and get to work remote for most of the week I can do 6 hours of office work plus 6 hours of research work, or alternatively 6 + 4 and compensate by doing some research on the weekends. However I've heard conflicting feedback about this plan. One of my roommates says this is a horrible idea and that I'll become the Joker after a couple months, while one of my coworkers said I should wait a bit to see if this job won't demand too much of me (still in training currently), but that he thinks it's doable. Both are currently doing/have done a PhD at the same uni I want to enroll in. Also is 6ish hours per day even enough for a PhD?
Additional info: Public latam uni, so no tuition but the government grants are nothing to write home about (before getting the job it was barely enough to get by, and that was with help from my folks). The advisor I'm aiming for can be demanding at times but is also really nice and is new faculty. The PhD is in compsci (ML/NLP) and I plan to continue exploring a niche I'm already familiar with. Work schedule is fairly flexible, save for the fucking meetings (agile delenda est). A lot of credits can be done by getting good publications instead of doing uni courses.
Edit: Thanks everyone! I kind of feared "obviously no you moron" would be the general consensus. I probably got too optimistic about getting to keep doing research immediately. I'll wait for things to settle down and reevaluate my options. There's some mechanisms at the job that are supposedly designed so you can continue education, but my impression is that those are mostly reserved for MBA types, infrequently offered and also really contested, but I should ask around some more to be sure. I also know some better sources of funding are available once you enroll, but seeing my friend applying for those and failing repeatedly discourages me from betting on it. Worst comes to worst I'll save up some money, try doing this for a bit and quit if it proves unsustainable. Again, thanks for the input!
This reminds me of an older paper on how LLMs can't even do basic math when examples fall outside the training distribution (note that this was GPT-J and as far as I'm aware no such analysis is possible with GPT4, I wonder why), so this phenomena is not exclusive to multimodal stuff. It's one thing to pre-train a large capacity model on a general task that might benefit downstream tasks, but wanting these models to be general purpose is really, really silly.
I'm of the opinion that we're approaching a crisis in AI, we've hit a barrier on what current approaches are capable of achieving and no amount of data, labelers and tinkering with architectural minutiae or (god forbid) "prompt engineering" can fix that. My hopes are that with the bubble bursting the field will have to reckon with the need for algorithmic and architectural innovation, more robust standards for what constitutes a proper benchmark and reproducibility at the very least, and maybe, just maybe, extend its collective knowledge from other fields of study past 1960's neuroscience and explore the ethical and societal implications of your work more deeply than the oftentimes tiny obligatory ethics section of a paper. That is definetly a overgeneralization, so sorry for any researchers out here <3, I'm just disillusioned with the general state of the field.
You're correct about the C suites though , all they needed to see was one of those stupid graphs that showed line going up, with model capacity on the x axis and performance on the y axis, and their greed did the rest.