5 ms·
I work as an AI research engineer, so I’m having a totally different experience than the others I’m seeing here. Job market is very very strong for senior AI ro
by fnbr 4y ago
I work as an AI research engineer, so I’m having a totally different experience than the others I’m seeing here. Job market is very very strong for senior AI roles. Multiple offers that were >2x what I was making at a BigTech research lab (depending on how you value the equity). Interview process was much easier than I was expecting- mostly situational interviews (although I was referred in for those).
Having said that, many places refused to interview me because I didn’t have experience with LLMs. I was surprised how strict some places were with experience requirements.
- jfrbfbreudh 4y agoAre we talking 7 figure offers? I was under the impression that $500k packages were not uncommon for AI researchers at Google.
- lettergram 4y agoLLMs have been interesting - I need to update my resume to focus on it. I have experience working on models like that since 2018 and are frankly less interesting than some alternatives (to me). My guess is a lot of companies / hiring managers are searching for key terms. Screening is likely not from people who know what they’re talking about
- ChatGTP 4y agoSame, feels like it's time to get the magic acronym on the resume too.
- lettergram 4y agoLol name checks out
- ChatGTP 4y ago:)
- fnbr 4y agoYeah, it was dumb. One company gave me an offer to lead engineering for their LLM team and another wouldn’t even interview me because I didn’t have any LLM experience. Clearly one of them was wrong.
- bigiain 4y agoObviously there’s a profitable side hustle here. Train yourself an LLM on resumes that land interviews and offers, and sell access to it… :-)
- pedrosorio 4y agoSounds like the differentiator here was the company not the resume
- jakeinspace 4y agoI’m assuming that "AI research engineer" implies a PhD, but I’m curious what your progression was. How long were you in academia before moving to industry research?
- fnbr 4y ago4 year BSc + 1 year MSc. Experience matters much more than having a PhD. Lots of people without phds at the big ai research labs as engineers. I’d say it’s uncommon for engineers to have phds (although almost all have masters).
- e-_pusher 4y agoDo folks still expect you to keep up with research papers as an ML engineer?
- fnbr 4y agoYeah, you should be familiar with the state of the art in your field. You don’t need to be as strong with theory though as Research Scientists.
- misterhtmlcss 4y agoI’m a web dev and I’m wondering if there is really any practical way of levelling up to become an ML Engineer. Seems there are ‘schools’ out there but based on what I read on Hacker News these places won’t remotely prepare me to make that transition. Not complaining just would appreciate some of your thoughts on what are good options to consider to make that leap.
- bunnyhop 4y agoTBH the code behind generative stuff is "fairly" simple, I made a GAN for my Masters degree back in 2019 that produced 16x16 tile assets for x-bit styled videogames. The rough part is CREATING the math that supports the learning models, its very easy to implement it after the fact. If you have a strong data structure foundation, and are decent with math then it shouldn't be a hard transition to pick up ML. It's the flavor of the year and everywhere you look schools are hustling for your money, can't say I've tried any of them/have an opinion of them so GL with that.
- bigiain 4y ago> depending on how you value the equity Zero. Value the equity at zero. Or at most value it like a lottery ticket - something that lets you amuse yourself dreaming about what you’ll do when you “win”. But make all your ongoing plans assuming they equity will be worth zero. Either because the startup fails, or because you get screwed out of any payout by the founders or board, or by future investors insisting on deals where early employee equity gets diluted and preferenced down to zero. Or even occasionally because they intentionally never have a liquidity event and stay private with no available way to monetise your shares. Those, or exits where employees make at most “new car” money not “fuck you” money, are by far the most common. Do a startup if you want, they can be amazing ways to gain an incredible amount of experience to add to your resume in a short time, they can be a great fun work style at least for a while, they can be great at finding out who amongst your co workers are genuinely super talented and productive people and building networks with them. But don’t fall for the “yeah, we’re paying well below market, but look at how generous our option grants are!” snake oil recruitment pitches…
- necovek 4y ago(I guess you missed the context) I don't think the same can be said of FAANG equity, which is the context this was in: while it is still a good lesson to treat it as worth zero, you can regularly exchange it for cash at market prices.
- fnbr 4y agoSalary was still 2x FAANG comp (and roughly equal to my FAANG TC). So I’m fine taking the equity bet under those circumstances. I agree with you though. One company valued their equity at 3x (!!) what it was worth in October.
- jstx1 4y agoAre your offers mostly from startups? I want to move away from ML but this might be a bad time for it.
- fnbr 4y agoA mix of startups, smallish research labs (that is, smaller than Google Brain/FAIR/DeepMind), small, profitable AI companies. I focused on smaller companies as that’s what I wanted. Bunch of larger companies were hiring though.
- jstx1 4y agoJust saw on your profile that you were a senior research engineer at DeepMind. That might be why you’re getting showered with offers and not necessarily the state of the job market for ML.
- hinata08 4y ago> I’m having a totally different experience than the others I’m seeing here I also feel like you have a different perspective. It's not just about the actual experience. Some complain about how hard it was to get a job, when they just needed a few weeks or months to get one. They were also surprised to have been turn down by some companies, which is healthy. You can't match everywhere. You're not qualified on everything. Job ads are minimalist to ambiguous, so tech teams have to do a lot of filtering to make up with what HR published. You may also want to apply outside of your comfort zone, because you don't risk anything,... Also, some seem surprised to have to interview. While 8 is definitely too many, it makes sense to do some, when you apply for positions with experience. You're no longer graduate, they want to know about you. I was also surprised by users who reported to have a lower wage now, which is probably still much higher than the median income in the USA Such views don't picture a weak market, but a more realistic market. We're mostly among well paid tech industry workers here, so we can complain about the processes. But out of respect for most workers who earn up to the median wage, it's not decent to call these times a difficult job market. I hadn't fully realise what kind of bubble the tech jobs used to be over there, before reading this thread. Anyway, gg for getting a nice job.