jstx1

12 ms·

91.
by jstx1 3y ago
My problem with the "just grind leetcode" advice is that I'm not getting interviews in the first place. But if you're able to get interviews, yes.
92.
by jstx1 3y ago
Always disliked the design of iterators in Python - you use 'def' but it's not a function, you call it like a function but it preserved its state - there should be a less confusing way to express these semantics than yield.
93.
by jstx1 3y ago
> but I've heard (mind you, a few years ago) of people squeezing in Google Deepmind just by knowing the topic they're supposed to work on What position? No coding test for a research scientist is more plausible than no coding t
94.
by jstx1 3y ago
> Do I need a &mut Option<Thing>, or an Option<&mut Thing>? Follow up to this example: let's say that I don't know the answer, what do I search for if I want to find out?
95.
by jstx1 3y ago
Programming simple things in Rust is less readable than programming the same things in other languages - and you can directly point at the language design as the root cause for this. I'm here doing things like creating an object, itera
96.
by jstx1 3y ago
It's not modern languages, it's Rust specifically - they didn't optimize for readability and ended up with a language that's hard to read.
97.
by jstx1 3y ago
I think yes. I'm some % more productive because of Copilot and especially because of ChatGPT(with GPT-4). No idea what the number is but it feels significant since these tools often helps me get unstuck, generate new ideas, explain cod
98.
by jstx1 3y ago
VCs have no idea what to do with their money, on the other hand potential founders gravitate towards dev tools because it's a space they know well. It's not necessarily that they're seeing value (not saying that there isn&#x2
99.
by jstx1 3y ago
It's a choice to treat all candidates equally. And intentionally excluding candidates is the whole point of designing an interview process, it's lazy to lift your arms up and declare that they're all biased.
100.
by jstx1 3y ago
It's easy to fake projects and commits. And when they aren't faked, you can't have an expectation of every candidate having these because it's so much work. And if you decide to bypass some of your usual interview proces
101.
by jstx1 3y ago
It's not one or ther other; many interviews assess both because they're both meaningful signals about a candidate. If you're coming from the position that one is "understand and assess and write and communicate a deep se
102.
by jstx1 3y ago
I want to add a technical factor - the thing that slows me down the most is lack of knowledge about peripheral operational details - what all the configuration files do, how my repository is templating something from a different repository,
103.
by jstx1 3y ago
Normalized dot product is equivalent to cosine similarity.
104.
by jstx1 3y ago
Definitely. It was okay for a couple of years but it’s not how I want to spend the rest of my career.
105.
by jstx1 3y ago
> an employer to say “yes”… This is a "just draw the owl" kind of comment. It's hard to get hired with just a certificate for plug and play tools and it isn't the kind of job OP wants anyway.
106.
by jstx1 3y ago
> AI Engineering is basically Data Engineering focused on AI. I work in machine learning and this isn't how I see it at all. Data engineering specifically evolved as a term to differentiate the people who work with data but don&#x27
107.
by jstx1 3y ago
> matrix multiplications in Rust, 1. not exotic 2. LLMs can do this easily 3. it's already abstracted away in efficient libraries so largely an obsolete skill even before LLMs (maybe it's good for educational purposes) All of t
108.
by jstx1 3y ago
> I'd benefit from honing in my skills while working on my own project. You can do these things while at university, I wouldn't take a year off.
109.
by jstx1 3y ago
You're glossing over the fact that your choice of WFH or in-office affects your coworkers too. WFH when you're the only one doing it and everyone has in-person communication all the time is much more difficult. Or if most of your
110.
by jstx1 3y ago
It looks slimy and gives off the impression that you don't respect your users.
111.
by jstx1 3y ago
There's lots of moving parts, some systems are very complex and it's a space where experience is worth a lot - it's hard to get started with ops 101 just by yourself like you can do with frontend web development for example.
112.
by jstx1 3y ago
No, you don't need to transition to ML, you don't need to work on ML, you don't need to learn ML (unless you want to of course). There's no ticking clock, you don't need any of this in order to have a good career. &
113.
by jstx1 3y ago
> if it exhibited autonomy and value judgments. Who wants this from ML systems? I want them to be useful, not to have autonomy and value judgments.
114.
by jstx1 3y ago
Is the salary higher? Do you want to move to Switzerland? Do you want to do the Engineer II job? If all answers are yes, take it.
115.
by jstx1 3y ago
When it comes to tech HN tends to be bearish on some useful things just because they also happen to be hyped up (Kubernetes, machine learning), and overvalue tech which is rarely used in the real world (obscure programming languages). >
116.
by jstx1 3y ago
> It would be nice to have an auto reply script that makes them fill out a Google form with relevant information for you to review later. No, that's weird. Ask them questions in the chat instead of redirecting them to a form. It&#x2
117.
by jstx1 3y ago
> I do think “AI” is a hype-bubble, but machine learning and specifically the sort of LLMs we’re seeing now are definitively not a hype-bubble. There isn't any kind of distinction - when people say AI in 2023 they mean ML, LLMs, im
118.
by jstx1 3y ago
> Trying to champion one or the other feels like pointless virtue signalling really. No, I mean something very specific - I can do things at work with LLMs that I couldn't do a year ago without them. I can't say the same about
119.
by jstx1 3y ago
It isn't. There's a fundamental difference between hype around something useful and hype around something that isn't useful.
120.
by jstx1 3y ago
GPT-4 is considerably better than the free version. And I find it hard to believe that whichever model you used has been wrong 100% of the time.

More ›