career-advice

[icy berry]

i do the work myself

[sonic pilot]

Google and docs mostly or?

[icy berry]

sometimes, but google dont work anymore, so i have stopped using it

[sonic pilot]

How

[icy berry]

by thinking

[sonic pilot]

What do U use docs then

[near ocean]

People solved all kinds of problems before LLMs, they used google and documentation and forums like stackoverflow and general chatting to their coworkers

[sonic pilot]

What id it requires a Library to solve the problem

[icy berry]

you seem very qurious about thins, why do you ask?

[icy berry]

normally not no.

hard problems are not solved with code, thats just an implementation of the solution

and for that you look up details that you dont remember, programming is not about memorization

[sonic pilot]

Because l kinda of avoid docs

[icy berry]

you should not

you learn from understanding them

[glacial heart]

Yo eivl can I ask one help from u #ot0-psvm’s-eternal-disapproval

[icy berry]

sure

[sonic pilot]

Yeah l realised that like pro know the code off head because they have probably used them a lot of time so it becomes muscle memory but at fist they had to look up the code on how to do it

[icy berry]

yes, you have to learn programming first. that is true

and if you use AI or youtube, you will not learn iit

[sonic pilot]

Yeah true docs is the way because you write the code and you have to understand something your writing/coding

[icy berry]

yes, programming is about understanding concepts

and you use that understanding to solve problmes when you work professionally

the code is just how you tell the computer what to do. at that poiint the problem is often solved already

[sonic pilot]

Like let's say l have a problem l have found a solution online I can either watch a tutorial or use the official docs what should l do they both solve the problem remember

[icy berry]

if you fiind yourself in a possition like that, and one thing is easy, you will not learn much by doing it easy

[sonic pilot]

Frl

[viscid chasm]

AI/ML or/and cyber or/and web?

Im focusing on mainly learning ai rn

[open ivy]

What interests you more?

1. Looking at the weights and biases, trying to figure out how it all works. "brain damaging" it and seeing what happens. Experimenting with different architectures. Etc
2. Building an app with AI. Making something people enjoy. Finding a new use-case for a given LLM. Etc.

[fierce sail]

got any advice for data engineering?

[mortal wedge]

Yeah, I agree and as hard as it's gonna be I'm gonna go for it anyway.

Yeah, hopefully this is hybrid in a 1-3 times a week in office and not hybrid as in you get one day at home and 4 in office, lol

Most companies are not set up to create the sort of workflow in house to do this. Most likely it would be a business that comes up with a way to automate a lot of this stuff and then sell their workflow/software to another business. B2B sales, basicallyh

There's going to be a lot of conflicting information. A lot of it depends on who is going to be viewing it and who you're optimizing for.

Hiring managers may prefer skills at the bottom, but ATS and HR prefer it at the top. As ATS and HR are more predictable, I choose to optimize for them by placing them at the top. Doesn't matter how good it looks to a hiring manager if it never gets to them.

[halcyon kraken]

I can't figure out what kind of jobs I can do with Python, what I can and can't do. I'm not very good at math, so I can't do those kinds of calculations. I don't know, if I knew Python, what fields could I find work in? Let's say I found a job, what exactly would I do? I can't start without seeing the future. I don't know if I've explained what I mean. I'm 25 years old and I want to make it my profession, but I doubt I'll reach a sufficient level. I hope you understand what I mean.
What I want from you is a small roadmap, something like, "If you improve in this area, it's possible, you'll find a job, you can work in these areas." Honestly 😄

[cunning saffron]

How proficient at python do i need to be to pursue data science

[mortal wedge]

Fairly but not overly proficient. It's more important to be proficient in methods of statistical analysis, but make sure you know how to do those analyses in Python.

[pine sleet]

Let's say I found a job, what exactly would I do?
you would do what the job description says you would do. what is your educational background? do you have a degree? is it tech related?

[halcyon kraken]

business management, I recently graduated from university.

[manic trail]

helo

[open ivy]

Isn't it frustrating to go to a tech meetup event but there are no discussions about technical geeky stuff? Like going to chess club but they forgot to bring the pieces. Hard to make a case for technical employability if no one is interested...

I asked AI about this issue, for what it's worth. It said to chase after "scary sounding words" like "GPU pipeline optimization". Not sure if it's accurate? It would be nice to have a better place to discuss my portfolio.

[manic trail]

@open ivy where's your portfolio? on github? if so, can i see it?

or don't reply.

[open ivy]

My GitHub is a mess. Thankfully I am getting very close to the point where I can build a website, as I have been "tightening up" my portfolio. Remind me in two weeks and I can show you (if allowed by self-doxxing rules).

[manic trail]

@open ivy and what about AI

turns out these days you need to be a VERY VERY qualified dev to get a job.

[peak halo]

Yep

[manic trail]

i can tell @peak halo is qualified by the thickness of his glasses

there is a direct relationship between the thickness of glasses frames / lenses to IQ

[peak halo]

While I'd usually disagree, I am the glasses, and the body is a meat puppet.

[manic trail]

also, he's not afraid to share his github. HIRED!

[peak halo]

Anyway, the market will hopefully self-correct. But that means that CS enrollment needs to go down for a while

[manic trail]

@peak halo my prediction is that AI will take away nearly all dev jobs

[peak halo]

That's not a new prediction. It's also not likely under the current AI paradigm. The last AI paradigm shift was 2017, and there's no guarantee that there will even be another one.

It feels like there's been rapid development in the last three years, and in many ways there has been, but it's all been iterating within the current paradigm.

[manic trail]

take a look at claude opus 4.6.

[peak halo]

I'm aware.

[manic trail]

aware, have you used it?

[peak halo]

I'm a computational linguist and I analyze LLMs professionally.

[manic trail]

whenever i hear "i'm a professional" from someone, i feel the opposite feeling the statement is meant to invoke.

[peak halo]

Okay.

I suppose you're right that you should want a better reason than "I'm a professional, trust me bro"

But I don't feel like writing one right now because I'm on my phone and because I'm a bastard.

[open ivy]

AI isn't "taking dev jobs away en masse". However, AI slop can clog up job applications. The task is well suited for AI (said task isn't challenging or interesting, it has been done many times, and the result is mainly text rather than a difficult technical task or software project archetecture). So that can make it harder to get a job even without "replacement".

[balmy spade]

"Have you used it" feels like such a weak entry-point of an opinion. Yes, many of us have used it and plenty of us are still unconvinced. Is the next question "What have you done with it" or "Show me what you've done" as it has been asked by so many others before? And if I produce that proof, it moves to "You aren't using it correctly" or "You don't understand prompting" or any other number of equally weak stances that never go anywhere (except to another weak stance).

It's a heck of a prediction that LLMs, as they are today, could (replace|take away) (nearly)? all (dev)? jobs. I think the safer prediction is that the technology will change the landscape, but maybe that's why so many don't take that stance. It's safe. Boring. Not enough hype or emotion in it.

Long winded ramble to say; We still need devs, so it's worth learning those skills if you're interested in it.

[manic trail]

it's not an entry point of an opinoin. it was a question.

the rest of your message was tl;dr

[balmy spade]

Well, thank you for saving me the time to explain myself further. Have a great day.

[peak halo]

It's not even that long. And it was well written.

[mortal wedge]

I don't remember the exact seminar, but I attended a seminar that was given by one of the founding engineers of Uber.

I was frustrated because I don't think he actually covered the seminar topic, he just talked about how great he was during his time at Uber.

[open ivy]

Frusterating! Because most clubs/events/meetups/etc are about what they say they are. But tech seems to be one of not that many topics where you can go and have very little tech in the meetups.

[still condor]

@Grok summarize this message

[hushed crane]

I'm pivoting into writing

A gem from a book that was written 20+ years ago.

Like it or not, market is always changing.

No matter how stubborn you are, you can't get mad for not getting perl job in 2026....

[manic trail]

@peak halo Your pinned repos are all spaCy wrappers and CRFs. that's ~2019 classical NLP, not GenAI. no LangChain, LoRA, or RAGAS work.

[hushed crane]

My friend from uni contributted early in Lang Chain 🤯

[astral ermine]

What's the title of this book?

[fast fossil]

personally, I can't wait for that to happen
in the meantime though...

[hushed crane]

Pragmmatic programmer

[astral ermine]

Thanks

[manic trail]

i read a summary of this book

[hushed crane]

Why

[manic trail]

so i can learn from it

[manic trail]

same reason as you i'm guessing

[turbid bobcat]

which book is that? looks to be pretty good

it's "Pragmmatic programmer" just saw it in later messages

[hushed crane]

Pragmatic programmer

[turbid bobcat]

ty

[craggy rapids]

That's interesting if market is changing ,we can learn more.

Because learning never ends

[fast fossil]

It does for you after you die

[minor beacon]

Yes please tell me Hey 👋 sure! Nice to meet you. What would you like to talk about?

[opal shoal]

Right

[vestal bay]

Scam

Give You Only demotivation

[peak halo]

Yes, because I've been working for a company for five years where everything is closed source.

Except for the one project that was open source for some reason.

[manic trail]

Closed source

[peak halo]

Are you just trolling at this point or what?

[manic trail]

I'm not a fan of closed source. Is that bad?

[peak halo]

That's fine, but you'll be hard pressed to find a company that just lets everyone see what they're doing.

[next plover]

Most companies are closed source

[manic trail]

Most companies hire for custom software used internally.

[next plover]

Yeah?

[peak halo]

Though my company does maintain the CVE as a public service.

[sturdy hedge]

yeah honestly most professional python work stays behind closed doors anyway

[manic trail]

i'd say this is not software that is being distributed publicly so i wouldn't have the same ethical concerns about it. it's custom software used internally. the company who uses it has full access to the source code.

[modest bluff]

I am an automation developer and web scrapper, i am trying to find clients but fail everytime, does anyone have advice?

[peak halo]

the market for freelancers is always competitive, but right now it's exceptionally so. Unfortunately, I think you might need to look into a different source of income.

[modest bluff]

Do you mean there is less demand this time ?

[peak halo]

no, there's too many people trying to be freelancers.

particularly tasks like webscraping, which really aren't that difficult.

[modest bluff]

Oh okay

Well i ll try to do my best and see , thanks !

[peak halo]

do you have a contingency plan? how long can you afford to keep trying to get freelance jobs before you run out of money?

[modest bluff]

Money isn t a problem now , i still live with my parents, for the plan… i don t really have one, i ve seen that in platforms like fiverr , upwork…etc there is lot of competition so it s very hard getting clients there so i started posting on reddit but nothing serious yet .

[peak halo]

if you're serious about being able to support yourself financially as a programmer, you should probably get a degree in computer science. the job market is currently bad, but it might improve by the time you finish the degree.

By no fault of your own, I do not think you're going to be successful in your current endeavor.

[modest bluff]

Well , i ve learned automation and web scrapping just to make some money so that i can afford my robotic projects , which is the main field i am interested in , i never tried making money by freelancing from robotics , i don t even know if it s possible or not , and if demand is present and high or not

[sacred basalt]

Hello everyone.
I'm PHP web developer(Laravel/Wordpress), Lately, I've been drawn to developing and starting a career in Python. I began my journey in programming with this language, writing simple Telegram bots and automating tasks using Selenium. Recently, I tried the Flask framework and created an simple website for an old client with a CRUD admin panel.
Do you think it's worth switching to Django and FastAPI now? How is the market doing, and should I consider focusing more on LLM and data processing rather than web development?

[dire compass]

I found this article on Medium.
I would start with learning FastAPI since you already have experience in web development. You could even develop LLM-powered apps for your portfolio, gaining experience in both areas.

[dire compass]

Hey guys, I have a Bachelor's and a Master's in Electrical Engineering, but am currently looking for SW Engineering roles. Most of my 9 years as a full-time professional were spent developing and applying software to EE problems. However, I have several gaps in my programming knowledge, as I never took any computer science-specific classes (other than a beginner C programming course). I have been coding in Python for 17 years, but when faced with a complex problem to be solved from scratch, my mind often blocks. How should I approach filling those gaps to help me get and succeed in a SW Engineering role?

[turbid bobcat]

you'll be fine

[fringe sphinx]

We can't really do much but guess at what 'gaps' you might have. Can you be more specific about where you think you are lacking?

[mortal wedge]

What @fringe sphinx said, but I was in a similar place, I had spent years applying software to Biomedical problems and found myself needing to overcome gaps in my knowledge. Probably best to look over some overview/flow chart and try to identify stuff. For me it was DSA and fundamentals. Took some courses on coursera to bridge the gap

[worthy oasis]

Hi guys, I'm a neat and i was wondering would it be best for me to go back into education to get a software engineering degree , go into a boot camp to rehash my skills and jumpstart my learning or to continue making projects until i get myself to a level of a junior role (18 from the uk)

and if so would open university be a wise option?

[fringe sphinx]

What's a neat?

[worthy oasis]

not in education or employed

[peak halo]

it's NEET -- not in employment, education, or training

[grim monolith]

i train aikido does that count

[peak halo]

not for NEET

[grim monolith]

https://tenor.com/view/aikido-gif-27343233

[bleak vortex]

Have you guys ever landed a job in the jewelry industry?

[slate anchor]

cybersecurity is better, but no..I haven’t

[bleak vortex]

might be i just wanna focus overall for jewelry

[slate anchor]

you do you, chase your dreams man

[bleak vortex]

cybersecurity for jewelry sounds dope tho

i been thinking about it for counterfeit and tracking, not so sure how much cybersecurity involved in it

[slate anchor]

well, tracking numbers and data involvement

(Serial numbers) —> Trackable

[edgy yarrow]

what % of SWE internship interviews ask leetcode style questions

[bleak vortex]

does it count if it involves unique ids engraved for every piece so its more secure for anti faking?

[slate anchor]

not really public so not a lot of people know

[slate anchor]

well, it depends

branding and logos exist but idk if they have serial numbers for “jewelry industry”

[bleak vortex]

they do just micro engraving it, there is a bunch of interesting stuff i wanna combine for jewelry, inventory, tracking and security for it just no clue how public it is those kinds of jobs for jewelry

[slate anchor]

other brands have serial numbers attached to the bottom of jewelry but in a small fragment

[bleak vortex]

considering security, antitheft and RFIDs is it a very cybersecurity thing or do you not need that much cybersecurity knowledge for this kind of stuff

[slate anchor]

not really

You don’t need to be a hacker or pentester to deal with it, it’s more physical/security engineering; simple digital verification than hardcore infosec.

[bleak vortex]

yeah so i bet cybersecurity is not really the path to go then for that

[slate anchor]

spot on, cybersecurity isn’t really the main path or a deep requirement.

[bleak vortex]

thats good to know, that helps alot too thanks

[slate anchor]

No problem

[grand pumice]

Hello, I am a thirteen year old who is interested in coding python, i know absolutely nothing about coding but it interests me, any way I can learn through this server?

[queen tundra]

same here, and yes you can learn tons of things

[grand pumice]

is there a certain tutoring bot or ppl that help you?

[queen tundra]

Depends, what are you looking to work on?

[grand pumice]

ehh idrk, are there sections on python coding u can suggest for me to work on?

[queen tundra]

I started out with making games in unity, I would use VS Code and install the python extension

Then I moved to bots then websites

@grand pumice really just dive into whatever sounds fun to you

[grand pumice]

alr, thx a lot for the advice!

[queen tundra]

ofc

[vapid violet]

I have had 4 interviews for SWE internships. 0 of them have had me do an in person coding assessment, or really talk about anything technical at all. I have done some "on your own time" coding assessments which I assume are similar to leetcode.

[lilac yoke]

the vast majority, if the company is large its safe to assume they alwayas will

[shrewd finch]

i also want to learn python from scratch, after that probably i wanna learn other languages too but currently since i am a beginner i am confused, i reached dictionaries, is there good place where i can learn basics, i want to build a strong foundation, please anyone, i am 20 yrs old currently

[smoky quest]

In terms of career, a degree is the path of least resistance and with the most opportunities and compensation

[slate anchor]

depends what templates your wanting

[mint oracle]

I think 20 is too old!! as a hobby is good but as a career i think u r too old!!

[balmy mural]

20 is definitely not too old to start a career in tech. I know multiple people who only started their degrees in their mid twenties and are now successful software engineers.

[near ocean]

I had people 40+ years old in my classes in uni, at 20 youre barely even a person

[quick summit]

And somehow being 40+ is almost too old for anything these days, it's dumb.

[scenic juniper]

hire me

[turbid bobcat]

wdyvm, anyone under 21 literally still looks like a teenager

[sturdy hedge]

lol for real, starting at 20 is actually way ahead of the curve

[still condor]

!cleanban @pulsar silo spam bot

[inner wren]

:incoming_envelope: :ok_hand: applied ban to @pulsar silo permanently.

[turbid bobcat]

I think they might've meant too young

wait no, I looked again at the context and it just dont make sense

[sturdy hedge]

yeah parker was for sure trolling, ignore that nonsense

[sturdy hedge]

true, but either way 20 is a solid age to get into it

[barren lotus]

in a hypothetical situation where i would like to get a referral for a python related job in the next week or so, am i allowed to ask here?

[near ocean]

I mean, the average dev will be half way through a degree by that age so youre not ahead of the curve by starting one at 20 but also it's not over

[sturdy hedge]

fair point, just meant they still have plenty of time to pivot

[mint oracle]

I am not trolling!! but i have a different pov!!
I agree , U can learn anything!!
but market is bad and competition is high!! saturated!!
Yeah u can learn!!

[barren lotus]

!rules 9

[inner wren]

Rules

9. Do not offer or ask for paid work of any kind.

[analog sun]

@supple parrot your message was removed for advertising. Please re-read the server rules you agreed to when joining: #rules

[turbid bobcat]

I entered college at 19. I literally started studying at 18. Went through 12 years of school just going like: BRRRRRRRRRR

made literally 0 difference

matter fact, Id do it again cuz it mightve been beneficial in its own twisted way

[next plover]

That's just rude smh

[near ocean]

But is it wrong

[sturdy hedge]

kinda true but starting at 20 gives you so much room for error

[next plover]

I'm basically a fully functioning adult i even got w2 in the mail

[near ocean]

Post prefrontal cortex

[next plover]

????

[turbid bobcat]

you are a person at least since the moment you were born, obviously lmao

the brain keeps developing until 25, and we all develop at different paces

there is no right or wrong answer as to when to start or finish or wtv

[vast shoal]

What does it even mean for the brain to "keep developing until 25"? I mean, the brain obviously changes after that as well, neurons rewire when you learn new things and such.

[astral ermine]

If you trust someone enough with that, I don't see why not. Just make sure they can speak positively to your abilities and not be detrimental.

[next plover]

Google apparently says the age 25 thing is a misconception

However it's safe to assume my prefrontal cortex has a ways to go before it's on par with old people

[sturdy hedge]

lol fair, honestly i'm 27 and still feel like i'm winging it sometimes

[balmy mural]

Just got my first rejection in my current jobhunt that happened after there's already been at least one interview and not during CV screening or external recruiter process. They gave detailed feedback on why they didn't move forward with me, so that's pretty cool. Hurts a bit since it's the company I was the most interested in for my current batch of applications that haven't been rejected yet

[placid ember]

Hi, i'm python developer with 9 (almost 10) years of experience, i want to start job as a freelancer, cs i just lost my job, it would be nice of u if u could give me some orders

[peak halo]

you can't ask for work here, but you can ask for job-hunting advice.

or maybe "orders" is a mistranslation of "advice".

[fast fossil]

or maybe a mistranslation of "direction"

[severe mural]

Rule 1

[peak halo]

what about it? that's the one that says to follow the #code-of-conduct

[mental parrot]

Hello, i wanna try to learn python, what is the best and fastest way to do it?.

[peak halo]

forget about learning it "fast". it's a long-term journey.

[mental parrot]

ahh man, is it worth learning

I just get bored after 1 week

[severe mural]

How those the python community operate ?

[peak halo]

I do not understand your question. Try rephrasing it in #community-meta.

[mental parrot]

Anybody here makes malware?

[peak halo]

No, and I'll ban you if you ask again.

[viscid kite]

Hey! Where can I find some dev practice/experience (not necessarily commercial) just to learn how real projects work? Any tips?

[sturdy hedge]

cool yeah check out github repos with the good first issue label

[viscid kite]

thank you I'll check it out

[fervent acorn]

Guys, I am currently learning python from code with Harry [youtube]
Can you guys tell me , whether I should purchase a course or stick to my youtube Playlist?

[sturdy hedge]

free stuff is fine, check the official tutorial then just start building projects

[leaden canopy]

hello, i want to be a python automation AI engineer, or a security engineer since i come from cybersecurity, what projects do you think would be good to have to be competitive when i apply?

[sturdy hedge]

maybe build an automated vulnerability scanner with scapy, it bridges security and automation perfectly

[fervent acorn]

Thanks bro for advice!!

[leaden canopy]

how to implement AI?

[sturdy hedge]

maybe use scikit-learn to classify the findings or an llm for log analysis

[leaden canopy]

alright

[fringe sphinx]

Variety. Build a broad foundation first before trying to specialize. Web development, game development, cli apps, etc are all good ways to build core skills

!kin has project ideas

[inner wren]

Kindling Projects

The Kindling projects page contains a list of projects and ideas programmers can tackle to build their skills and knowledge.

[sturdy hedge]

curious what everyone else built as their first project that wasn't a tutorial

[supple parrot]

Does anyone here have a project idea or an ongoing project ?
If you have any difficulties or need a developer, let's talk.

[mortal wedge]

I'm not sure about age 25 being a misconception (been working in neuroscience). But what I can conclusively say is that 18 is definitely not when brains stop developing. I don't have access to the research projects I worked on anymore and everyone is different, but I do recall a trend of somewhere in the low 20s?

[sturdy hedge]

low 20s sounds right, like @turbid bobcat said earlier it doesnt even matter when you start

[mortal wedge]

Aside from like childhood/teen development, age related effects on the brain pale in comparison to all the other factors we've discovered and the ones we haven't.

But obviously from like age 0-3 it's all age related lol

I share that though, primarily for older adults who worry their brains aren't good enough to keep up or whatever, that while that is a factor it is such a small factor compared to others.

Keep learning, keep your brain active, treat your body well and you'll be alright!

[turbid bobcat]

https://pmc.ncbi.nlm.nih.gov/articles/PMC3621648/

[mortal wedge]

Yeah, the 18 cutoff is definitely just a cultural/arbitrary one

Also, I don't consider any age too high to learn

[turbid bobcat]

it's useful for legal stuff, it doesnt match the science cuz the gov wants taxes and bodies for the army

[mortal wedge]

Right, but the legal stuff is just a norm. Either way I agree, it doesn't match the science.

[turbid bobcat]

nah leetcode is nonsense

lots of ppl here told me to just study data st and algo book + its exercises, and Ive come to agree w them

[solid parcel]

LeetCode is far from a test of creativity. It's 100% evaluating an ability to pattern match (as well as, frankly, if you're someone prepared to grind something that can be pretty tedious in order to get good at it). Can you recognise when to apply known concepts to novel scenarios? Can you identify how efficient your code is, and iteratively improve it?

Frankly, most interviewing generally can be boiled down to effective pattern matching, including things like culture fit rounds.

[sturdy hedge]

lol true, its basically just training your brain for interviews, not actual coding

[deep shadow]

Kinda sucks that that's the case. I guess I still have to do it because I don't want to be in a situation where I have to cram LeetCode to find a job. I rather do it when I'm employed so I can build confidence now.

[sturdy hedge]

yeah its a slog but the confidence definitely helps when you're actually in the room

[deep shadow]

I thought it could be a fun thing I could do on the side as a creativity thing.
But tbf, it might still make me a better programmer, so that's something.

[deep shadow]

Makes sense. Gotta make that money lol

[mortal wedge]

The more senior you are the less you'll be asked to do leetcode style interviews. But it being infrequent doesn't matter, what if the job you really want asks it? Even if only 1 out of 10 interview companies or 1 out of 20, don't you want to still pass it?

That's my take, anyway. Although I get that you have to decide how to prioritize your time between skilling up, self care, applying, researching companies, etc.

[sturdy hedge]

true, better to be safe, but for seniors system design usually matters way more

[mortal wedge]

Can I just say that unemployment in this market takes more work than anything I've ever done while employed, lol

[sturdy hedge]

lol too real, job hunting is a whole other level of grind

[mortal wedge]

Yeah, you'll absolutely get better/best mileage from making sure you can pass design questions. It's more of once you're good on that and want to increase your odds in the case of this situation.

[turbid bobcat]

If I ever encounter leetcode outside of faang ever again imma just call them posers and exit the interview

[deep shadow]

Yeah, no, I agree 100%
That's why I started doing them. I don't want to be in a situation where I need to do those questions and can't

Also, the interviews are genuinely always harder than the work 😭

[mortal wedge]

Also, at least for me, it's less about being able to solve leetcode as it is to solve quickly under time pressure without referencing outside resources

[sturdy hedge]

if you want a break from leetcode check out advent of code, it's way more fun

[turbid bobcat]

except that im not gonna cuz the next thing imma do is make a business myself cuz everyone else is doing it wrong, for ex by demanding leetcode

[mortal wedge]

and ofc you'll have someone watching you and commenting on you the whole time

[turbid bobcat]

"vibe code facebook in 20min"

you kno is coming to a tech interview near u

[mortal wedge]

I'd exit out of that interview

IF that's really what our entire disicpline has come to, I give up

[turbid bobcat]

nah, no one trusts vibe coded code

I do think humans suck at coding tho

like just ship it, jeeezus

[deep shadow]

I get that
I guess it's just not testing the skills I thought it would be. Funny enough, I think right now it is, but if I want to get good at it, I need to stop using creativity and start thinking differently 😅

I'm fine with it though, I'm giving it like 15 minutes a day. I also genuinely really enjoy coding, sometimes you gotta do things you don't love so that you can enjoy the rest of it, yk?

[near ocean]

They call this the system design interview

[turbid bobcat]

tbh dude, I'd just build stuff and learn more useful nonsense like going in depth about docker and the linux shell

I regret every sec I spent on leetcode

[smoky quest]

It's just datastructures and algorithms, the fundamentals of CS/coding.

I would suggest to not just grind leetcode but also pick up a resource on DSA to learn about it beyond continuing to practice. It will be more efficient, save you time and make you a better engineer

[sturdy hedge]

yeah learning the actual logic makes it way less of a grind honestly

[deep shadow]

I'm doing that too 😅

Funny you mention it, because I'm working on my first big project while learning devops right now. I can code, but I know next to nothing about devops and cloud hosting, so I'm on that too. But yes, thought I'd do some leetcode along with that because I also want to make sure I never find myself in a position where I don't have employment.

[smoky quest]

you don't learn DSA by grinding leetcode, the same way you don't learn physics by throwing a bunch of apples in the air

[deep shadow]

🫡
Any recommendations?

[smoky quest]

https://mitpress.mit.edu/9780262046305/introduction-to-algorithms/

[deep shadow]

I did my bachelors in CS btw, so I know data structures, I'm just extremely rusty at them. I haven't done that shit in 2 years 😭

[deep shadow]

Thank you!
I appreciate it!

[smoky quest]

Best time to refresh

[deep shadow]

Will do
and I appreciate you guys taking out the time to give me some perspective 💙

I really like being a programmer and I want to try and level up as much as I can, in all aspects of this field. (All relevant aspects, of course)

[sturdy hedge]

no worries, that mindset is exactly what sets great developers apart

[solid parcel]

Docker is all Linux 🙂 If you want to understand how it works under the hood, I massively recommend looking up Liz Rice and one of her talks on Building a Container from Scratch. It's all just cgroups, ulimits and namespaces

I'd started building out a minimal container runtime in Go... Will get back to it once I'm through my dissertation

[turbid bobcat]

I kno, ive been doing devops for a living

[sturdy hedge]

tried building a simple namespace isolator in python once, go is way better for low level stuff

[solid parcel]

At the moment I'm lucky if I get any time to code at all, I'm ending up spending all my time in meetings and writing flipping Jira tickets and requirements 🙃

[deep shadow]

🗒️ 🖊️

[turbid bobcat]

im so glad I jumped ship b4 getting to that

I hate meetings, Im a do stuff typa dude

[solid parcel]

I'm interviewing internally for an SRE/platform engineering role that should be more hands on if I land it. So with any luck I'll get that, and then look to pivot to Google SRE in a couple of years.

I had checkout.com reach out yesterday too, which would something I'd consider going for if it wasn't for my dissertation. Payments experience would be pretty great to get.

[turbid bobcat]

Ive had that job title and still dont kno what it means. Just felt like a do everything role

platf eng that is, idk abour SRE

[solid parcel]

Really comes down to the org. This particular internal position would be a supporting function helping multiple application teams embed SRE practices, as well as building out golden paths and tooling for them to consume (particularly when it comes to observability and monitoring). With a sprinkle of your standard incident response thrown in for good measure.

[turbid bobcat]

yea true, it always seems to vary by role

thats why in my resume now I just stick the job title "Software Engineer" everywhere and let the bulletpoints refine it

otherwise I get nonsense like "Platform Engineer" and then a bunch of backend, frontend, and AI in the hulletpoinrs

🤷

[solid parcel]

The roles I get people reaching out about are pretty much all cloud, DevOps, infrastructure, SRE, platform engineering. So far more on the operational side of the fence rather than hardcore dev roles

[turbid bobcat]

vary by org*

[solid parcel]

LinkedIn is convinced the role I'm the best fit for is an infra SWE at OpenAI, which is amusing, albeit woefully inaccurate

[sturdy hedge]

linkedin recommendations are basically just throwing darts at a board honestly

[turbid bobcat]

lmao apply, cashgrab those oligarchs

[solid parcel]

There was a period where it was showing me nothing but cleaning gigs, lol

[solid parcel]

Alas I'm far from qualified for that currently. I'm debating looking to target Anthropic after Google, but we shall see...

[sturdy hedge]

lol that is one hell of a career pivot

[turbid bobcat]

anthropic probly harder than openai

at least last time I checked their salaries were all half a mil and upwards

[solid parcel]

Perhaps! If a company has to win the AI race, I'd much prefer they do rather than OpenAI.

[mortal wedge]

From what I've heard, a Google SRE does mostly nothing, but when google goes down everyone and the CEO yells at them until it comes back up

[turbid bobcat]

those r all top of the food chain companies. good luck, get that bag 💰

[solid parcel]

I'm in the UK and the going rate looks to be about £300,000 base + a fair chunk in options, so that maps, yeah. Last I heard the big AI firms are basically paying Google L+1 for any given level

[turbid bobcat]

sigh, I'll never see such salaries any time soon cuz I cant stand big companies and their nonsense

300k pounds goes to 400k.USD. which matches my memory more or less

[solid parcel]

Ha, assuming there hasn't been a huge bursting by the time I'd get there (which feasibly would take me 4-5 years from now), perhaps I will.

I may also decide that I'm happy with the c. £240,000 I'd be on as a Google L5 rather than working to push further. Got a fair ways to go to get there though, that would be c. 2.4x my current comp, haha.

[turbid bobcat]

whats the diff at this point anyway

unless u paying crazy rent, u can just stick most of that in the stock market and be goose farmer

[solid parcel]

Right, this is pretty much the plan. At that point, I don't really know what the marginal value of another £100,000 - £200,000 gross would be. Particularly given the amount I'd have to grind to get there. It will fall right around the time I'll likely be looking to start a family, so I suspect my priorities will have shifted from chasing Anthropic.

[turbid bobcat]

half of that extra margin goes to the gov

but stock market wise, it does make a diff tbh

not in the short term ofc, but the classic compounding effect. probly a good idea to simulate it using lile a, say, 8% / year ROIn

I recall the averages per year are predicted to be lower than what we are currently used to

[solid parcel]

FYI after accounting for inflation, the global stock market returns an average of c. 4.5%. I use that when modelling growth (which if anything is fairly aggressive, given it would be a 100% stock allocation)

[turbid bobcat]

4.5% far too low unless u doing high inflation and/or u doing bonds or wtv

the US market is saturating rn but just looking at the past 3 years or so, it has been a crazy bull market

like, interest rates are hovering at ~3% and idk if they'll be dialing them back to 0 again

so the ROI you are using is really just the fed rate rn

ah u in the UK, which would mean rates are naturally higher to keep the pound within its historical strenght

[solid parcel]

Looking at the msci adjusted for GBP and inflation from 1900 to now, the average has been 5.6%. Forecasts going forward are expected to be meaningfully lower as valuations are so flipping stretched currently (Based on industry modelling and consensus, not just me going off gut feeling!)

[turbid bobcat]

ah u mean the all world indexes ?

[turbid bobcat]

would be believable ya

[solid parcel]

Yup. Likely to be more representative of long term returns. The US average is skewed by the crazy returns across the past 20 or so years

[turbid bobcat]

I either buy europe or US

everything else is suss. for ex emerg markets can be susceptiable to governance issues that cause share.dilution without protrction for shareholders

is all about having proper institutions and regulations

[vapid violet]

I applied for a summer 2026 internship a couple months back (4-5). I got rejected. The company has posted some internships to linkedin 2 hours ago. They appear to be the same internships they were previously offering. Do I apply again? Or should I wait until the summer 2027 ones come out? Will it be seen as "spam"?

[sturdy hedge]

honestly just go for it, recruitment can be a total coin toss sometimes

[vapid violet]

actually it looks like it was just a linkedin repost not an actual listing repost.

[pale flower]

ok

[short robin]

What career paths are their in coding with python can i do freelance coding? is that a good decision im tryna get a lil side job that i can do on a laptop or from home

[edgy mesa]

I need a new job screw this one

[mortal wedge]

Mood

[hearty sierra]

I wouldn't do freelance coding, you'll have way too much competition

[mortal wedge]

If you want to be competitive freelancing you need a lot of the same indicators that would be competitive in a full time role, if not more.

College degree, impressive portfolio, etc

connections

[hearty sierra]

Yeah just not a good idea especially if you're just starting out

[mortal wedge]

It's more of a "I've been doing this for a while and want a break from corporate" instead of "it's hard to get an entry level job and this will be easier" hack

[mystic whale]

But the job market seems equally competitive, of course, this depends on the country, but it seems that all sectors are "saturated" by high competition.

[mystic whale]

So it's a market made up of experienced people, there's no room for beginners, I understand.

[mortal wedge]

Essentially yes. In normal times I'd say go for entry level positions instead, but I'm also not sure those even exist anymore.

(True entry level, not an "entry level" position looking for 2-5 YOE)

[mystic whale]

I know close friends who got good beginner jobs, and yet they seem to have less knowledge than I do. I try every day to improve my skills and invest in new ones, but it seems that those who make "more noise," stand out more, are better regarded.

[mystic whale]

Yes, I've been studying for a year, and I've never even had an interview opportunity, and I can already do projects with CRUD, but it doesn't seem to have the return I expected.

[mortal wedge]

Hiring pipelines are chaotic messes right now that do not seem like they're succeeding at reliably surfacing the strongest candidates to positions. Things like when you apply are worth a disproportionate amount of significance. (You either want to apply within hours to get in the first batch of candidates or you want to apply after a position has been reposted. Everyone in between is essentially dumped.)

And individual pipelines are so disparate. Even a simple question like "should I upload my resume as a docx or a pdf" is already fraught because some ATS systems will be able to do well at processing one and not the other. Some ATS's will rank applicants by keyword usage and others will just rank them in chronological order by time someone applies. Advice is all over the place.

The market definitely is not behaving like it has previously.

[next plover]

I can't recommend a website and blog enough it's literally a cheat code

[mystic whale]

That's precisely why I've been thinking about creating something, whether it's a SaaS product or freelancing for a company, and also building a network of contacts.

[fringe sphinx]

Allocate time to the non-technical aspects of University. Clubs, social, talking to professors, etc are all part of the experience.

It's great to build different things and gain experience, but it's also important to learn from other people. Talk to people who've recently graduated and hear their experience

[mystic whale]

I do, and agree it's important, I have a professor who is my mentor and prepares me for activities more complex than those in my course. Whenever there are IT events in the city, I try to go, but I still feel like something is missing. It seems like I'll only really enter the field after I start earning money from it.

[fringe sphinx]

Any idea what is missing? What do you feel?

[mystic whale]

I agree that it's important to stay in touch with people who have the kind of life I want to have.

[mystic whale]

Overall, regarding money, it seems that unless I receive something for my effort, I continue to feel like I'm doing something wrong.

[fringe sphinx]

I guess it's tough until you land that first job. Sounds like a fairly normal self doubt: the reality is we all don't know more than we know 🙂

[mystic whale]

When I look at my classmates, I seem to be way ahead, but when I look at someone who is successful in their career, I feel like I'm doing something wrong. It's this feeling of uncertainty that makes me seek advice.

[fringe sphinx]

Classic imposter syndrome, if I may be an internet doctor. You're in good company.

I bet if you ask your peers, they feel the same

[mortal wedge]

Yeah, impostor syndrome is a hell of a condition.

I'm getting it hardcore while being unemployed for a minute. Logically I know that I'm in a niche that is getting hit particularly hard and even if I wasn't, everyone is struggling.

[mystic whale]

Yes, maybe I need to focus more on attending job interviews than on studying, not that I need to stop studying, but rather focus on what will help me get a job.

[fringe sphinx]

My only real advice is: allocate more time to listening to other people's journeys.

[mortal wedge]

Also, bear in mind, that in many contexts people will be putting their best foot forward.

As opposed to their actual journey/situation. Like if you go on facebook everyone is living their best life and having the most thrilling experiences but that's not indicative of the average experience

[mystic whale]

Exactly, that's why I wanted to try something new, to distance myself from what is conventional.

[mystic whale]

Yes, I know some people who earn well and work for foreign companies; I'll see if I can have that kind of conversation with them.

[mystic whale]

Without a doubt, social media is another problem; distancing myself from it a bit would really help alleviate this syndrome, and I've been trying to do so.

[mortal wedge]

One thing I did was use a browser extension to disable the linkedin feed and replace it with an inspirational quote/message

[mystic whale]

Often, the problem lies in how we view the achievements of others; regarding many things, I try to think "One day I will achieve that," but it always remains just on paper.

[mortal wedge]

The feed really is just not great for mental health, at least not for me

[hushed crane]

For anyone with a functioning brain really

[still condor]

!rule 6 9 @burnt hazel @tulip thistle Make sure to read our rules. this is not the place to look for work

[inner wren]

Rules

6. Do not post unapproved advertising.

9. Do not offer or ask for paid work of any kind.

[ionic nacelle]

hello does server have any general

[still condor]

If you want to talk about Python, #python-discussion. If you want to talk about things unrelated to Python, we have three off-topic channels ( #ot0-psvm’s-eternal-disapproval #ot1-perplexing-regexing #ot2-never-nester’s-nightmare )

[ionic nacelle]

i want to ask are u a mod and if u r can i use emojis

[still condor]

I don't see why we'd ban emojis

[ionic nacelle]

not spamming

[slate anchor]

yes they are a mod (check someone’s roles if you suspect this) and emojis are available for a reason, but it is irrelevant to the topic on the channel

[solid parcel]

Interesting, what kind of impact have you seen from having one?

[solid parcel]

It could be worse, it could be Blind 🙃

[sturdy hedge]

lol blind is a different level of toxic honestly

[solid parcel]

It really, really is

[frigid isle]

Hello everyone, I'm learning Python again 🙂

[next plover]

Gives people a chance to read about what you do beyond a resume that looks exactly the same as the last 400 applicants

And forces you to learn concepts deep enough to be a be able to explain them.

[solid parcel]

Do you get much traffic to it? I'd be surprised if most recruiters could be bothered with checking it out, particularly at the point where they're just scanning through a whole ton of CVs. Even once at interview, I find that even the contents of my CV isn't focused on much, let alone anything extraneous to that.

[solid parcel]

Agreed this is a fantastic way to learn 🙂

[next plover]

I try to post my blogs on linkedin and I pretend I have an audience outside of that and it's very fun to me lol

[sturdy hedge]

honestly that's the best way to learn, if you can explain it you know it

[next plover]

However people who interviewed me for my actual security position looked at it and a few people I've worked with have read it as well

[sturdy hedge]

that is sick, actually proves the extra effort pays off sometimes

[solid parcel]

Interesting to hear. Once I'm through my dissertation and have some capacity, I'm planning to look into raising my profile. Getting a website in place, actually posting on LinkedIn, starting to give a few talks... All that good stuff

[next plover]

There's some quote like learn at arms length so you can teach at elbows length

[sturdy hedge]

giving talks is massive for networking, good luck with the dissertation

[near ocean]

Based and linkedinpilled actually

[solid parcel]

It's the flip side of why AI can cripple learning speed. Teaching forces you to engage deeply with a lot of minutiae that you may skim past otherwise

[solid parcel]

Thank you! Only been going for a month or so, and I'm feeling pretty wiped... Currently juggling work full time, interviewing for a new internal role, a colleague with a huge attitude problem and then the dissertation on top 🙃

[sturdy hedge]

for sure, just copy-pasting code doesn't actually teach you the underlying logic

[sturdy hedge]

damn that sounds exhausting, maybe try blocking out weekends just for the dissertation

[solid parcel]

Ha, yeah that's what I'm trying to do. I've got a proof of concept in place, at least! Doing clever graph things to shift privilege escalation detection left to pre-deployment (while maintaining awareness of live environment state, including resources not managed by infrastructure as code). So it fills a gap that no existing tooling does.

[sturdy hedge]

shifting security left with graphs is sick, you using networkx

[solid parcel]

I'm using AzureHound (an Azure collector for a privilege escalation detection tool called Bloodhound) to capture live state (which crucially can include resources managed outside of Terraform that existing pre-deployment tooling would be unaware of), parsing the Terraform plan and integrating that with the AzureHound output. I can then upload that artifact to BloodHound and query it for privilege escalation pathways. The composite graph of live state + Terraform plan output allows me to 'project forward' to the expected future state if the plan were to be applied. So I can detect if a planned change would introduce a privilege escalation pathway 🙂

I've been pleasantly surprised at how well it works. This weekend I'll be putting together a test set. Got some in place, but need more scenarios in order to derive statistically significant results for my dissertation.

Current pre-deployment tooling only has context of the Terraform code itself, while post-deployment tooling is reactionary by nature. So I'm looking to shift the rigour of the latter approach to the pre-deployment step.

Under the hood, BloodHound uses neo4j

If I get time, I may look to use networkx or similar for visualisation, as the default manner BloodHound displays info isn't brilliant for what I'm trying to do.

[next plover]

We always be agreeing on everything 🧠

[solid parcel]

fools seldom differ 😆😉

[fringe sphinx]

Always felt this was a silly expression, given the infinitude of ways to be wrong 🙂

[near ocean]

<@&831776746206265384> ad & scam?

[fringe sphinx]

!ban 1028936761583337544 Scam

[inner wren]

:incoming_envelope: :ok_hand: applied ban to @silent cape permanently.

[frigid isle]

Anyonne here got a job that requires knowledge of multiple programming languages?

I'm seeing so many jobs that doesn't use Python, so I feel like man.

Am I missing out on something

[sturdy hedge]

python is huge for backend and data, but most jobs use a mix of languages

[frigid isle]

How long does it take to be good in multiple languages do land a job, does anyone know?

[solid parcel]

Oh it massively is, haha

[sturdy hedge]

just focus on one for now, picking up others gets way easier after that

[frigid isle]

Understandable

[near ocean]

Thats most software jobs

[frigid isle]

Yeah, what journey must I go through to land one

[sturdy hedge]

build something you'd actually use, that's what makes you stand out in interviews

[near ocean]

Usually degree, some projects, some leetcode, some practice talking to strangers about tech stuff

[frigid isle]

True!

[frigid isle]

Got it!'

[fringe sphinx]

How many roads must a man walk down?

There's only one answer, ofc.

[frigid isle]

Literally one?

[fringe sphinx]

42

[frigid isle]

What?

[fast fossil]

might be the wrong question though

[frigid isle]

May you bless me with the right question

[solid parcel]

He was making a reference to The Hitchhiker's Guide To The Galaxy, haha. The answer to The Ultimate Question of Life, the Universe and Everything is found to be 42, but they're none the wiser as to what the question is

[frigid isle]

I see

Okay wait am I missing the point? Did that somehow anwser my question in a way? 😂😂😂

[sturdy hedge]

lol nah just a joke, just focus on building one solid project for now

[frigid isle]

hahaha alright thanks mate!

[fringe sphinx]

It's a reference to a book that many programmers have read: The Hitchiker's Guide to the Galaxy.

42 is the boomers 67.

[solid parcel]

My life has become subtly but measurably worse for having read that sentence

[fast fossil]

is 69 the millennial 67 then?

or maybe gen z's 67
since 67 is more of a gen alpha thing, isn't it

[lethal kettle]

yes

[turbid bobcat]

ah yes, back when words had meaning and jokes needed to be marginally clever

jk jokes werent funny

[next plover]

My dad loves 42

[south basin]

What are your top alternatives to LinkedIn?

[peak halo]

for what purpose?

because there isn't another "facebook for professionals", because that kind of thing only works if everyone goes to the same one.

[south basin]

Search for job postings

[peak halo]

I would use Google Jobs (which is for jobs in general, not at google specifically). it might be marginally helpful to just use that to figure out where there are jobs, and then to apply on the actual website for that company.

[south basin]

Did you try to scrape ATS systems for new openings?

[peak halo]

No

when you do automated applymaxing, you're contributing to a problem

[south basin]

Sure

I guess direct apply automation for thousands of company sites will be too labor-intensive for one person

Not the same thing as easy apply spam on LinkedIn

[vast shoal]

Flooding companies with applications push them to use automation to manage incoming applications as well, which dehumanizes the applicant and makes it way harder to get interviews if you don't have qualifications that look great at a glance and can be filtered for (like a university degree). If it feels like applying for jobs today sucks, this is partially why.

[peak halo]

I also don't think applymaxing increases ones own chances of getting a job. it just makes it take longer for whoever does end up getting it.

[south basin]

Well, if some job receive 1000 applications and 990 belongs to your bot farm, guess what will happen with other 10 candidates

[vast shoal]

I do think it can in some cases decrease the chances of getting a job for some people who are qualified but don't have the kinds of upfront qualifications that recruiters filter on though.

Which is sad.

[jagged locust]

It’ll probably be easy for them to realize someone applied 990 times though - I don’t see the point

[south basin]

The point is these 990 different people applied automatically

[smoky quest]

For your purpose, there are gotchas to automatically filter spam applications.
So that means you may be sending yourself tons of applications, but they will have low probability to reach actual humans

Improving your resume, getting feedback, working on projects and networking would have a higher ROI than trying to automate it

[south basin]

Well you can get the feedback by sending 990 different resumes with fake credentials and see which one passed through the ai filter

[fast fossil]

rather selfish

[peak halo]

what do you expect to happen if you get to an interview because of a fraudulent resume?

[south basin]

You can send real resume with real credentials, but built in a right way not to be wiped off by ai

[fast fossil]

I think it's not a terrible* way to find out what properties make the resume go through far enough for you to get an interview

[peak halo]

that doesn't answer my question

[jagged locust]

I have no idea how you can make that work @south basin

[smoky quest]

That's still lower ROI.
You would have to spend a lot of effort to make sure they are different enough, yet similar enough so that you can have attribution of change and impact

[smoky quest]

I bet you cannot

[jagged locust]

For starters they'd all need to have your Last name and First name, so it's trivial to see they're all yours. If they have random Last names and First names then you're gonna be selected but then what, it's not you

[south basin]

Then you know the structure that actually penetrates the screen and send a right person with this structure

[fast fossil]

no, in theory this just tells you what properties make it go through, so you adjust your resume and send that in

[smoky quest]

and then they see the very same content/structure and think you are a bot and get yeeted

[jagged locust]

I mean we all know what properties make resumes go through

Go to a top university, have a high GPA, ... is this breaking news? do we need to make bots to discover that?

[peak halo]

I don't know what one would expect to learn from "resume AB testing" that you can't learn from just... reading the job listing.

[south basin]

I guess that wouldn't be 1 in 990 that goes through, so there will be a space to change some things if you get the pattern

[jagged locust]

But what pattern are you looking for, they don't screen for people using a particular font size
There's nothing hidden

[fast fossil]

this would be your chance to find out if that's really the case

[south basin]

That might be not that simple if you have a long list of previous jobs and skills

[fast fossil]

perhaps, resumes that use ComicSans have a greater probability of being accepted than those that use Times New Roman

[jagged locust]

I find this particulary hard to believe

[peak halo]

it's a joke

though I would be mildly annoyed if I got a resume in a non-serif font

"Illinois" give me serifs pls

[fast fossil]

I joke but now I'm actually intrigued to run such an experiment for, uhh, learning purposes

[jagged locust]

Would you? I find non-serif cleaner / more low-key idk

[fast fossil]

now I definitely want to run this experiment

[south basin]

Well at least you can check if they reply to anyone at all or it's a ghost job

[peak halo]

I just hate how upper-case I looks in non-serif fonts. that's it.

[jagged locust]

I mean you could lol

not sure how that'd reveal anything useful, but i do loads of useless things for fun

[still condor]

That might be true. Not because of the font, but because a highly valued professional is more likely to goof around and format their resume with Comic Sans, knowing that employers will want them anyway

correlation not causation thing

[fast fossil]

hmmm, dammit, not really an easy experiment to run, tons of things to account for actually

[jagged locust]

Might be industry dependent, but I personally never applied for a job with a resume on a website, I've always been referred by someone from the company so I guess that's kinda how you should do things if you want actual advice lol

[south basin]

Another question. How do you decide if you will refer said person to the position

[jagged locust]

What does that mean

[fast fossil]

you think about it and then you have your decision
idk, do you like the person, do you find them suitable, are they even looking for a job right now

[south basin]

How will you decide if he is suitable or not?

[jagged locust]

I mean, you imagine them doing the tasks that'd be assigned to them. If they're likely to farm sign-ons and mess around, probably not a good idea to refer that person. If they're smart and have decent soft skills, go for it.

Basically common sense should prevail, no need to do any complicated analysis of your friend

[fast fossil]

you are friends with them, have worked with them previously (in a professional setting or on some project) and know what they are capable of
could also ask them

[south basin]

Do you have a lot of friends who meet all of those criteria?

[jagged locust]

It's not compulsory to refer people

[fast fossil]

not really

[south basin]

In that case the idea of staking your career on such a connections sounds unreliable, isn't it?

[fast fossil]

what

[jagged locust]

Then don't refer people that you think aren't fit for the position?? Again, you don't have to do it
Also I find it unlikely that your career would take a hit if you refer someone that turns out to be bad at their job. Worst thing that could happen is that they'll stop taking your referrals seriously.

[south basin]

You speak from the position of a person who gives referral, I meant the other side

[fast fossil]

I think they're trying to say that networking is worthless because since I don't have friends to refer to positions similar to mine, no one would in turn refer me to such a position
thus they conclude that mass-spamming resumes is the superior approach to landing a job
idk

[jagged locust]

You don't do any judging on the other side, they already have the job, they'll be judging you, i don't understand your question

[south basin]

Nevermind

[jagged locust]

I mean if you want to mass-spam resumes I won't stop you, I just find this not as useful as actually listening to your lecturers and doing homework assignments (i'm assuming you're a student)

[south basin]

No, I never was, in this field at least. I also didn't say that I want to send a spam, i was looking for your opinion only (this room I mean)

[smoky quest]

There are two kinds:

• People you would stake your name for it. You vouch they are great and use your karma for that
• You just want the referral bonus and you will refer any living body in the chance to get it

The former has obviously far more value than the latter

[south basin]

I don't know for sure does this question violate rule 9 or not

[peak halo]

it does

[south basin]

Deleted!

[mortal wedge]

It depends. The best case is they stop taking your referrals seriously. The worst case is that it's a hit to your overall credibility, if someone is less credible in one area it's understandable why someone would decide that they're less credible in other areas as well. Bad hires can blow back on the person making the hiring decision. (Especially depending on how bad the hire is, outside of just doing their job poorly, like if they steal from the company there's a chance you could get fired for recommending them).

The inverse of this is also true, though. Good hiring decisions can count in your favor, strong referrals can increase your credibility. And in many cases, there are referral bonuses and if you're shown to have good referrals your referrals are more likely to get in.

[smoky quest]

it's less relevant to this discord given the members, but it's also an expectation of higher level positions: they do come with their network

[jagged locust]

Yeah I guess if you refer someone that eventuelly reneges on their offer, that might blow back on you

[fringe sphinx]

I don't think I'd ever blame the referrer for someone reneging on an offer.

Or anything negative.

[merry ledge]

just had some guy call me 10 minutes after I filled out the application pissed off that i applied for a position without a security clearance 😂 i thought they supplied those?

[jagged locust]

Has anyone ever reneged on an offer? I've been offered a few times recently after saying I couldn't continue the process with firm X because firm Y was pushing for an answer, and I thought it was considered the rudest thing ever but it seems it's quite common since I've been suggested to do it by literally the HR of those "firm X" that couldn't plan on-sites fast enough?

[peak halo]

If they job requires an active security clearance, you have to have already gotten one from a previous job

[merry ledge]

smh

im never getting a job.

[peak halo]

Jobs that will help you get a security clearance are unfortunately less common than ones that require an active one. But sometimes, there's literally nothing they could have you do while waiting for you to go through the clearance process (and potentially not get one)

[merry ledge]

how do i get a security clearance

[peak halo]

You have to work for a company or agency that has some, but not all, classified work. And if there's a reason you could support something that company does that is classified, they'll help you start the application process

[merry ledge]

im so cooked. I am out of options

i should have never done CS

[spice bridge]

What are some projects I can start that will help me learn, I’ve only done school projects up until now and i’ve kind of hit a wall

[fringe sphinx]

What major is safe these days anyway

[merry ledge]

our engineering career fairs have >70% of companies accepting civil

[fringe sphinx]

Fair, AEC firms do have a shortage of positions... but entry salaries are significantly lower

[merry ledge]

any salary beats CS salary

[fringe sphinx]

That's not reality, overall, tho.

[merry ledge]

ive applied to hundreds of companies i will literally starve in 7 months

[gritty torrent]

I'm literally starving now. 🤷‍♂️

[merry ledge]

i might have to start applying in europe. That is how cooked i am

[fringe sphinx]

Where are you?

[merry ledge]

USA

[smoky quest]

Some people have and would usually imply burning some bridges

[jagged locust]

Meaning the company would put you on a black list but that'd be all?
Or would there be some sort of bigger repercussion?

I ended up not reneging but I'm still curious if that was to happen to me again

[smoky quest]

something like that. Though people talk and move companies too

No one will hold you accountable if you withdraw before signing, but after signing, it would be no different than a company withdrawing their job offer to you after you both signed; you would be pissed and for good reasons

[jagged locust]

My understanding of reneging was saying yes to the offer informally (email or phone call with HR) and not actually signing

[smoky quest]

if you haven't signed the offer, then it's not reneging

[jagged locust]

you'd say that's "less bad"? that could still lead to you being on a black list though right? technically you said yes whether it was informal or formal doesn't change much does it?

[smoky quest]

reneging means you go back on your words. If you haven't signed, you haven't agreed to anything and there is no stigma or anything bad to it

[smoky quest]

it's not even bad. It means there was no agreement and no commitment

[smoky quest]

there wouldn't be a reason to be on a blacklist because you got a better offer somewhere else and you did not even sign with them

If you withdraw before signing, they may even keep the door open for you in case it doesn't work out in that other company

[jagged locust]

Because the thing is, for example, when I got the offer, I was told you have until X to say yes or no by email. I ended up saying yes but I signed the contract only a week after X (they drafted the contract, I read it, etc ...) so in that week in between X and X+1 you're saying it's fine saying actually company Y offered me more so I won't sign?

[smoky quest]

it's absolutely fine

[jagged locust]

Damn, I didn't play that very well

[smoky quest]

you could even decide to not sign because of the weather, mood or whatever

[jagged locust]

I guess I was being overly cautious, I felt like after saying yes by email it was sort of impossible for me to say no without burning bridges

[smoky quest]

a verbal acceptance is a bit more dicey, but it's not signed until it's signed

[jagged locust]

I guess a better way of doing things would've been to pick something in the contract that I didn't like as a pretext
instead of just saying Y offered me more money and sound as a money-maximiser person

[smoky quest]

I did that once too and they sent me flowers and told me I could join them any time after

[smoky quest]

don't give excuses that they can address. Because what do you do if they decide to change the clause for you?
To that end, keep it short, professional and non-specific

you don't want them to try to find a solution for you

[jagged locust]

Wouldn't they ask what made you change your mind
Would you say "personal reasons"?

[smoky quest]

they can.
It's also up to you to choose whether to answer and how to answer. You can keep it vague in terms of how the other company is a better fit for your career [growth]

or you could throw a line about how they had a bigger offer. Worst case they match it 🙂

[jagged locust]

In my mind the whole "can you match that offer" was before saying yes informally

I honestly kinda got strong armed by the HR person I think lol I didn't even ask for a deadline extension - I was kinda scared they'd rescind

[smoky quest]

ideally, it should all happen before you say yes informally

[smoky quest]

sales people gotta close that deal

and yes, they might rescind that deal at any time and for any reason

It's not signed until it's signed

[jagged locust]

The only thing I negotiated was no non-compete if I leave within 6 months

I guess I'll be more prepared next time

[merry ledge]

are there any european countries in need of software engineers or are they just as saturated?

[jagged locust]

That was some fast moderation

[merry ledge]

they lurk

[south basin]

Well, if you are okay with $2500 net per month, you might have a chance

[fringe sphinx]

Part of my point is don't misinterpret an economic downturn as saturation. The effect may appear the same but the cause is very different

[smoky quest]

Note also that as long as it is not signed, both parties will continue to look around.
So until they make an offer, they might make an offer to someone else.

Time kills all the deals

[merry ledge]

economic downturn causes saturation

[south basin]

And don't expect there is no saturation just because there is a downturn, I would say

[smoky quest]

Given the push for more independence, I would expect more demand coming up

[merry ledge]

wha does that mean

[fringe sphinx]

There's just a lot of misleading analyses out there

[jagged locust]

Well once they make an offer informally by email it's usually a almost-sure thing for you right?

otherwise the reputation of the company might get a hit or whatever if they start throwing offers and rescinding a lot

[fringe sphinx]

From 'ai is taking our jerbs' and ' tech is cooked' to calmer 'tech companies are reacting to economic and political craziness with austerity measures'

[smoky quest]

Sure.
But let's say you appear washy washy or unsure or hard to get. And there is another candidate who comes up and is more motivated to join them.
Nothing stops them from withdrawing their offer to prioritize that other candidate.
Given nothing was signed, there is no harm done

[merry ledge]

is ai not taking our jerbs? the entry level positions are virtually non existant

[fringe sphinx]

Why do you attribute this to AI?

[merry ledge]

ai out performs my classmates in their coding and my classmates are entry

[smoky quest]

that speaks more to the skill level of your classmates than the AI

[fringe sphinx]

Ai outperforms in schoolwork because of the simplicity and commonness of th schoolwork

I am not arguing the converse (that ai is not a factor , just the absence of evidence for the thesis that ai is the dominant factor

[south basin]

I would expect more supply as well, given the hordes coming out of CS diploma mills and talented self-taughts everywhere on the one hand and manufacturing decline on the other hand

[pine turret]

Hey everyone, new to the server! I'm always up for a chat ❤️
Who’s down for some fun ?

[fringe sphinx]

!kin

[inner wren]

Kindling Projects

The Kindling projects page contains a list of projects and ideas programmers can tackle to build their skills and knowledge.

[merry ledge]

i haven't spent time finding actual sources. Just anecdotal evidence

[smoky quest]

Depends on the country.
There are also a few factors at play:

• Some EU countries put more weight on education than the US
• If a self taught can compete with someone with a degree, the person with a degree has bigger problems

[fringe sphinx]

I hypothesize that AI replacing engineers has had little impact on the overall job market

And that other economic reasons have been the bigger factors

[merry ledge]

too late for me to switch to civil

[merry ledge]

new grads are cooked

[smoky quest]

until the layoffs subside

[merry ledge]

your document shows layoffs went down already

[smoky quest]

going down does not mean it's done 😉

though I do hear people expecting more hiring coming up

[merry ledge]

before or after my 7 month deadline

[south basin]

My top:

1. Software market saturation in general. Was a thing in ~2018 at least, but people tried not to see the elephant in the room because stock market had no other story to sell to the public for further growth
2. Companies that had no chance to learn why overseas outsourcing is dangerous because they didn't know how to operate remotely in large scale got this chance and went this road (and it will take time before they learn it in a hard way and their successors will take their market share)
3. AI Ponzi scheme with all things attached
4. Actual efficiency growth because of AI
5. Fed rate and other financial and taxational BS

[smoky quest]

let me get my crystal ball :p

hopefully sooner for you

[merry ledge]

ill continue trying ❤️

[fringe sphinx]

You can post an anonymized resume here for critique. Your resume can always be better

[mortal wedge]

A lot of people have tried to draw comparison between this and other recessions. One convo I had the other day was about whether this market or previous ones being worse for recent grads.

My stance was that this market is uniquely bad for recent grads as CEOs/corporations are going forward with using AI to replace workers and entry level is the easiest target.

It's still bad for everyone, but I do think recent grads are disporportionately affected.

And unfortunately, while a lot of people are trying to flock to other industries, the unfortunate reality is that those other industries are generally saturated and will easily give preference to those with 5 YOE in the industry over someone trying to break in.

I've had little to no luck trying to move to adjacent roles. I don't have hard numbers as it was too depressing to track, but I would estimate I'm 20x to 100x more likely to get a response to roles within my niche industry than outside of it, especially depending on industry. (I've been in medtech/medical science, I've had numerous defense contractors reach out to me based off of my experience with AI for weapons development, but once I mention my work history is in medtech it has ended the conversation 100% of the time, lol)

[mortal wedge]

There's also the elephant in the room that with a changing tariff landscape, companies (in the US but also overseas who trade with the US) are often far more hesitant in making economic investments that could backfire if a new tariff goes up. SWE is not immune to this, if for no other reason than most computing infrastructure (the physical parts) come from overseas even it's deployed and maintained domestically.

[merry ledge]

sad to hear

[south basin]

I guess we can put it in the general downturn basket, it's not like there is something special for tech sector. Actually I think real sectors like manufacturing, agriculture or raw resources are more affected by that

[mortal wedge]

While true, I don't think most Americans appreciate how reliant we are on the east Asian electronics/semiconductor supply chains. Our capability to manufacture stuff domestically is highly limited and even if we invested we'd still have to ship the ingredients/parts in.

Either way, the economy is rough, everyone is hanging on to their jobs for dear life, and employers are generally hesitant to invest in workforce expansion and new product development.

(Also, a lot of people, myself included, have their jobs partially tied to physical devices. Anyone doing embedded work or making software that goes on medical devices, for instance.)

[still condor]

!rule 6 9 @supple parrot Please read our rules and the description of this channel. You cannot look for jobs on this server.

[inner wren]

Rules

6. Do not post unapproved advertising.

9. Do not offer or ask for paid work of any kind.

[fringe sphinx]

I'm unconvinced that AI is a significant driver or changer for SWE jobs. I don't buy the claimed productivity gains in corporate SWE-land: I could be wrong but I think what we're seeing is more about opportunistic austerity measures . (I'm not saying AI assisted coding is not useful or helpful, just that it's not making corporate SWEs suddenly 10xers)

[peak halo]

I think what we're seeing is more about opportunistic austerity measures

[fringe sphinx]

Some people liken this to the https://en.wikipedia.org/wiki/Productivity_paradox

[mortal wedge]

I'm not even talking about productivity/claimed productivity, I just mean that many corporations are laying off thousands of workers and using AI leverage as the specific rationale. It's the worst in tech but I'm seeing pressure across many industries to follow similarly.

Whether there are actual gains from doing this, if it's wise, if it's sustainable is a separate topic.

[fringe sphinx]

For sure, it is the boogeyman. We're at a weird inflection point: poor economic outlook combined with a bubble... or at least rapid investment in a single hyped industry

[mortal wedge]

Yeah

[fringe sphinx]

Previous peaks and valleys were less schizophrenic

[mortal wedge]

This growth isn't sustainable if for no other reason than that the raw materials needed for data center construction isn't available.

[fringe sphinx]

Howard Mark's memo on this is really insightful and interesting: https://youtu.be/sAjGXSdDkhQ?si=ZE4zIEsl2Jpkc1WZ

[mortal wedge]

I'll try to remember to check it out, can't atm

[fringe sphinx]

He's an old school finance dude, made his name in the fairly boring fixed income markets.

[solid parcel]

Ha, I was talking about this earlier today, albeit the 'paradox' part is overstated

[mortal wedge]

I do suspect that companies will, in the future, realize they need more humans in the loop than they think they need now.

But for now, this is the market we have to work with/in

[solid parcel]

Even if AI can get 90% of the way there on a piece of work, that last 10% is going to be a mountain to climb. I'm feeling very relaxed about the future for developers and tech employees more broadly.

[mortal wedge]

I don't disagree with you, but assuming that's the case it's going to take employers a while to figure that out.

[smoky quest]

AI doesn't have to solve 100%

[mortal wedge]

I think Gen AI will also face a steep cliff if regulation expands worldwide to restrict training data to data that is specifically licensed for that purpose.

We are starting to see such regulation across Europe

[smoky quest]

There isn't a single cause to all of this market. Some companies will want to lay off people to replace them with people, some others need to free up capital to invest in AI, some others are just riding that excuse, some others are downstream of these changes, etc.

[mortal wedge]

I think one of the reasons we saw such rapid development is that with a lot of AI you can increase performance by throwing more resources at it, even without significant advances in developing the tech. But even that has limits that we're coming up against, even in terms of physical computing infrastructure.

But I've soapboxed for long enough. To tie it back into the channel topic, I think things are going to be rough for a little while, at the very least.

Plateau was the word I was looking for, not steep cliff

[open ivy]

Oh yes another (when) "will AI steal our jobs". For starters we can look at:

1. How much more work can be done in said field before most "useful tasks" are used up vs how much the speedup is.
2. What changes there will be to the workflow and how hard it is to adapt.
3. Will there be a shiny new "AI integration expert" degree that uses hype to woo HR and makes them avoid "traditional" programmers?

Programming seems a bit safer than DevOps, but both seem pretty safe.

[mortal wedge]

(My only response to that is when I hear people brag about being a vibe coder I want to throw things at a wall)

[open ivy]

It's like someone woodworking and bragging about using power tools.

And connects back to my previous point that many "tech meetups" do not involve deep discussions of tech problems or people working together.

[smoky quest]

the value is not in knowing how to hold a screwdriver, but what you can produce

[mortal wedge]

Honestly, all of my best/deepest spontaneous tech discussions happened with tech related people at completely unrelated events. Heck, I've met recruiters at bars while out drinking with friends.

[open ivy]

And AI is absolutely flooding the companies with job applications. This creates problems for applicants which is separate than replacing humans. Makes me wonder what mechanisms they are using to defend and how effective they are?

[smoky quest]

AI is not flooding companies. People using AI are flooding applications

[mortal wedge]

AI mechanisms at best, random/arbitrary at worst, and not very based off of all the things they're saying/general attitude.

[open ivy]

Yes that is a more precise way to put it.

Not all AI bots are equal. The subset who have the best bots (who get around the defenses and who can afford the best cloud servers, among other factors) drown out the rest of us.

[peak halo]

"5+ years of experience" is weird. Just say how many you have.

I'll try to give this a better review tomorrow

!remind 12h do it you bastard

[astral ermine]

Thank you
You know, years ago you also reviewed my resume

[smoky quest]

a bit too handwavy/unspecific in your experience

[still condor]

I might be a bit bad at math but it seems like you have less than 5 years of experience

[astral ermine]

Stuff before 2021 that didn't fit

[still condor]

ah

[astral ermine]

Noted, wanted to make it readable for non tech people. I'll try be more specific

[smoky quest]

Show, don't tell

[astral ermine]

I like sharing screen and showing my projects when I get a chance in interviews, or walking through problems if I get assignment
Usually if I get an interview, I'll get an offer

So, you're saying I should e.g. be more specific on implementation and outcomes?

[mortal wedge]

Making it readable/understandable by non tech/industry experts is also important. It's a balance.

The first human step will be a recruiter who likely doesn't have technical depth who is probably looking at resumes that include keywords that they were either told by the hiring manager to filter for or the ATS did it natively.

One strategy is to use more straightforward language and expand based on parentheticals.

Developed an ML production pipeline with versioned outputs (feature extraction, run configs, monitoring, and regression testing).

But the goal is to have the necessary keywords to ensure you're not filtered out before the human stage, and then be direct and clear enough for the non-technical human screening stage to put your resume in the yes pile for the rest of the hiring team.

[open ivy]

There is a LOT of guesswork. Because we don't know what the filters are! The people using bots in a stealthy way (again, it's unclear what works and doesn't) have a HUGE advantage.

The standard "A/B testing" approach is very slow because you may get little, if any, feedback from either A and B and even if you do the sample size is tiny.

It's not easy to sniff out AI if the user actually knows what they are doing. Selenium browser fingerprinting is another arms race.

So beyond the standard "make it look good, be descriptive, tailored (somehow) and have just the right amount of exaggeration" it's a lot of luck.

[mortal wedge]

Yeah it’s a difficult problem to solve or optimize. It’s also very important to get your resume in early, but it takes time to tailor it, so you have to decide how much time to spend tailoring and this is assuming you’re learning about the position pretty close to when it’s posted.

The filters/tools/etc are inconsistent. Hell we can’t even advise if someone should use docx vs pdfs because there are a lot of different ATSs and what’s better is dependent on which ats is processing your resume if you can even get that information at all.

I will say the only tactic that has gotten me significantly and meaningfully more mileage is leveraging networks. But that's sort of a.... you can try to build them but oftentimes you sort of either have them or you don't. It's definitely not a slam dunk like it used to be, but it does help cut through the noise

[open ivy]

The lack of feedback in cold applications is crippling. Networking has feedback, still far from perfect but it's something. The trouble is you still don't know how far you are.

[arctic granite]

hello!

[turbid bobcat]

In your first bulletpoint you say somethin like "Architected and deployed a machine learning framework"

Machine learning (ML) frameworks are software libraries and tools—primarily in Python—that simplify building, training, and deploying AI models by providing pre-built functions for complex algorithms, data processing, and optimization. They enable faster development by handling lower-level mathematical operations. Leading frameworks include TensorFlow, PyTorch, Scikit-learn, and Keras.

Did you literally rebuild something like tensorflow?

[arctic granite]

Hello, I been coding a code that tries to clean datasets for .csv files for any dataset (through common data cleaning methods) and I been trying to complete the start of code (I used AI tool for only .bat files since I don't know how to run it and I am trying to learn how data science works not how to write .bat file) and I want someone to rate my code since it getting to main course (the actual data cleaning part) and I want to see what mistakes have I done and how I can improve them.

https://github.com/Mohammed-Musab/Lazy-Data-Cleaner

Please note that I am kind of a beginner and I want to use the code for college application and I want at least a functional code before end of April.

note: i reuploaded it from #data-science-and-ml since I didn't get response and it kind of the same thing for this channel.

[vast shoal]

Try posting in #1035199133436354600, or possibly #1468524576479641744

[turbid bobcat]

I can't tell what you're trying to say in the second bulletpoint

firsr bulletpoint second job is also phrased in a confusing way

first bulletpoint third job, as a european, I do find the mention of regulations exciting, especially since your resume is so geared towards medtech, would be nice if u could mention which very briefly

[arctic granite]

done

[arctic granite]

just a quick question.. I use AI for .bat files, is there any problems in that? Since I don't have any ideas of what is it and not willing to learn it

[vast shoal]

As long as the majority of the project is written by you, and in particular the Python code, I don't think it's a big deal

[turbid bobcat]

recruiters literally do not kno what they are doing

I've seen their nonsense live in the flesh, they straight up just judge you by the name of companies in your resume, the job titles, look for gaps, etc

then they literally just look for keywords

it's actually mind boggling they are included in the recruiting pipeline at all

[smoky quest]

no one would know since neither you would understand it nor the AI can guarantee it is correct

[arctic granite]

It written by me but I only use AI as a last result if I got error and I am trying to learn not copy & paste

[arctic granite]

the only 2 files in bat are:

run.bat - which just mirror python terminal
setup.bat - which just download Python3.11.9, colorama and pandas

[vast shoal]

I'm speaking in terms of whether it's ok to post the code review request or not. Not in terms of whether it's a good idea in general.

To be clear.

[smoky quest]

There are good recruiters and bad recruiters. There are knowledgeable recruiters and not knowledgeable recruiters.
Assuming all recruiters are bad is like assuming all candidates are bad. That's just not a productive generalization

[smoky quest]

Sure. How does that change any of the things that I said?

[arctic granite]

I don't know...

[vast shoal]

I pointed out one issue with the AI-generated bat file in your help thread already

[smoky quest]

okay

[arctic granite]

where

[vast shoal]

In your help thread?

[arctic granite]

found it

[vast shoal]

In any case, this channel is for career advice, so we should move the discussion of your project there

[arctic granite]

I will be using it for college application so does that count?

[vast shoal]

That still doesn't make this the appropriate place to discuss the technical details of your project

[turbid bobcat]

A good recruiter is probably a lot more expensive.

Like it's so rare to hear people say good things about them.

There's even a youtube channel of a recruiter just showing what they do and giving tips. And its as bad as im making it out to be.

[vast shoal]

You could discuss whether it's a good credential or not here

[smoky quest]

It's so rare to interact with a good candidate.
Most candidates are trash and don't even make it pass first contact

Does that mean that we should assume all candidates are bad?

[turbid bobcat]

No, but I would advise against giving them a job and a paying salary. Don't recruiters face the same scrutiny when getting hired

[smoky quest]

no different than bad engineers being hired

That's why people talk about A teams, B teams, C teams, etc.

[arctic granite]

Anyways does my project help me in my college application?

[smoky quest]

what's your project?

[arctic granite]

https://github.com/Mohammed-Musab/Lazy-Data-Cleaner

[smoky quest]

Don't expect anyone to click on that link.
Can you summarize it in 3-5 bullets?

[arctic granite]

oh okay give me a minute

[turbid bobcat]

I suppose you're right.

Let me rephrase then, I'd rather automate resume selection using an LLM than to have the typical recruiter be included in the pipeline

[smoky quest]

That's your right.
Though that would be a monkey paw moment

[turbid bobcat]

yes it wouldnt be ideal.

[arctic granite]

I am working on a data cleaner that uses should use common data cleaning methods based on user input... It currently doesn’t have any cleaning data functionality due to me working on how to copy files (so it edits a copy instead of main file) and displaying information about dataset. I want someone to review the code so when I start the data cleaning part of the code, I shouldn’t worry about anything else. This project has 2 main goals:

• For publishing
• For college application

[smoky quest]

Two main things:

• Show, don't tell
• You need to articulate what demonstrate your expertise/skills. What makes you a super engineer. So talk about what was difficult or unique or cool

[arctic granite]

I am bit off on thinking part... can you explain it more about the second point?

[smoky quest]

Imagine the recruiter. They receive a thousand resumes. They can't talk to a thousand people, just the top 20. So which ones should they pick?
And why do you write about your project? You must want the reader to learn something about you and how awesome you are. Otherwise, if there is nothing to learn from about you from that project, then why putting it on your resume? So what is that project telling me about you and your skills?

[arctic granite]

That I at least have an understanding of data analysis and this project support how it may be used for AI training.. and I am applying to college not recruiter.

[smoky quest]

so I have a thousand people claiming to know about data analysis.
How does "data cleaner that uses should use common data cleaning methods based on user input" show me you are better than most of them?

[arctic granite]

It should - since it isn't completed yet. provide the user with accurate correcting data program that removes errors and outlier from real life datasets

which also can be used to help train AI and simplify the data cleaning process for any users.

[smoky quest]

the fact that you removed some outliers does not tell me you are better than the other candidates

it just tells me you claim to have removed some outliers for your specific dataset

[arctic granite]

it is used to work for any datasets (that are .csv) and should be accurate in correcting datasets

[smoky quest]

you are telling me it should work, you are not showing me it works.
It could very well be a simple linear regression or a dbscan or a sarima based algo. But you aren't demonstrating any prowesss, you are just claiming to remove some outliers for your dataset

[smoky quest]

imagine I receive thousands of resumes claiming to clean some data. How do I differentiate them? How do I rank them?

what makes you special?

[arctic granite]

I am since it isn't complete yet... and i am working on it.

my plan for it simple:

• first display all information about the dataset
• then detect duplicate rows, remove them and drop columns with high missingness
• after that it fill missing values with column mean/median (depending on user response) or fill in the missing values with a constant
• at the end check memory usage, identify outliers, correlation matrix and group by and aggregate

Then introduce a basic visualizations features and generate pdf report of the dataset. The goal is to make it fast, user-friendly, and compatible with any datasets no matter what is it.

(those aren't fully everything it should do but a part of it)

[smoky quest]

As a resume reviewer, I would want to see more depth.
This is first year level

[arctic granite]

ya I am still a high school student

[smoky quest]

oh that changes a lot

then it's a great start!

[arctic granite]

that why i said I want it for college application

[smoky quest]

my bad. I am so used to job seekers

[arctic granite]

anyways can you rate my current progress?

I will start on cleaning part of the project... so I need it the start to be functional and good

[smoky quest]

It's a great start.

Though "identify outliers" has hidden dragons.
Outliers is based on what is defined as normal and it may not be obvious and will depend on the nature of your data

it would be cool to detail the algos you used for that

[arctic granite]

I got a todo list... of more than 20 steps to do

but I want the structure to be good

so?

[smoky quest]

what about it?

[arctic granite]

I want the structure to be good that I don't need to go back and rewrite a part (btw I take a long time to update since i am still in school)

https://github.com/Mohammed-Musab/Lazy-Data-Cleaner/

[smoky quest]

might be worth looking at uv, pandas or click

also scikit-learn

[arctic granite]

i will use it later for the AI part & visualization… and what is UV? numpy is no use yet.

[smoky quest]

also in terms of outliers detection, I do like that book a lot https://www.amazon.com/Outlier-Analysis-Charu-C-Aggarwal/dp/3319475770?
Though while it's quite readable, it might be a worthy challenge to read as a high schooler

[smoky quest]

uv helps manage your project and dependency

related to that, you may want to look into testing / unit testing

[arctic granite]

what that and give me UV full name so i can add it to my todo list

[smoky quest]

it's literally its full name 🙂

https://github.com/astral-sh/uv

[arctic granite]

what is it used for?

[smoky quest]

uv helps manage your project and dependency

[arctic granite]

so?

[smoky quest]

so it helps you get rid of your setup.bat

[arctic granite]

Okay that good

[smoky quest]

since you can tell it which python version to use with which libraries, etc.

in a standard and cross platform way

so now your script could even work on GNU/Linux!

[arctic granite]

oh

I will do linux support later... it like final boss and since I already have ubuntus it might be easy

[smoky quest]

nice

[shadow cove]

Is a Ai engineer a good career for the future ?

[turbid bobcat]

yes

[shadow cove]

alr

[fringe sphinx]

Don't think about specializing when you're starting: aim to be a good engineer, and be prepared for different types of jobs

[inner wren]

It has arrived!

Here's your reminder: do it you bastard
[Jump back to when you created the reminder](#career-advice message)

[odd quartz]

what do i do after learning web scraping?

[peak halo]

!warn @dreamy crypt your message was removed for asking for a job, which is not allowed

[inner wren]

:incoming_envelope: :ok_hand: applied warning to @dreamy crypt.

[arctic granite]

send help guys... i almost finished my college application resume (or cv i don't really know what is it) and it turn out you don't need it for the college i am applying for...

[peak halo]

okay, so you can hold on to it and update it when you apply for internships

[arctic granite]

it will need alot of update that will make it like writing one from scratch but with the same front

[old carbon]

Is learning to do simple things with Python profitable nowadays (in freelancing)? I feel like AI replaced everything that could be considered simple. The only moment it becomes profitable is when it comes to more complex and advanced stuff that takes even more time to learn and master

Just randomly thinking before trying to get on freelancing someday

(Sorry for the long text)

[turbid bobcat]

there's a ton of ppl that dont know how to use AI to make the stuff they want

you do still need to know software to prompt the AI

[old carbon]

Isnt that a minority

[turbid bobcat]

it's not. for example boomers with businesses outside of tech and need a website done

ive met a good share that are not even aware of the capabilities of current AI

[old carbon]

Oh well
Thank you

[hidden ivy]

Is anyone looking for an AI full-stack developer to ship an LLM feature or chatbot into production?

[peak halo]

You can't ask for jobs here.

[merry ledge]

so i applied to hundreds of USA locations and didnt get an interview. I applied to 1 single canada place accidentally and got an interview 😭

would i need a work visa or something?

[peak halo]

@merry ledge make sure they're aware of your location and citizenship and see what they say.

[open ivy]

What do you think is wrong with your application? Did anyone give feedback you think is honest?

[merry ledge]

the issue with my application is i dont have senior level experience

[peak halo]

That isn't inherently an issue. If the position is for a junior, and it's budgeted for a junior, then seniors are overqualified.

[open ivy]

How do you know it's that? Did they tell you?

[sly adder]

Bots or ML

[main peak]

Is ML something good to learn via a bootcamp while being in high school?

Btw whats the difference between Data analytics and ML engineering is it like the same or are different fields cause they all are clustered up in my mind 😅😅

[vapid jay]

I’m starting to learn programming, though I’ve been in IT for about 4–5 years in general. Everything would be fine, but seeing how AI advances year after year scares me. By the time I learn to code deeply enough in 2–3 years, AI will be truly advanced. I get that seniors and mid-level devs will use it as a tool because they have the underlying knowledge, but won't it be so advanced by then that the number of programmers will be radically reduced? I’m worried it’ll just be a few seniors acting as architects, managing their own AI bots trained to write the code, while they just oversee the whole thing and connect it to other services.

[vast shoal]

Highly doubtful. People often vastly overestimate the usefulness of current AI architectures in development. The core difficulty in programming is not writing the code, but understanding the problems and the solutions. AI can do the former (to some extent), but it can't really do the latter, even at a junior level. It seems exceedingly unlikely to change even as LLMs grow larger and more powerful, because the issue is not how big the model is, it's the nature of transformer architectures. We'd need an entirely new technological revolution for that to change, and we have no reason to believe that is more likely to come along in 5, 10, 100 years or ever. There's just no way to predict if and when that will happen, so there's no reason to make career decisions based on assumptions about that. Besides, if companies stop hiring juniors, who's going to replace the seniors who retire?

[celest moat]

Can anyone share the roadmap of learning python?

[vast shoal]

The roadmap is just: learn the basics through an introductory course while doing practical exercises and small projects along the way, then continue doing projects of increasing size and complexity in areas that interest you, and learn other technologies as you need them.

You need to write code in order to become a good coder.

You can come here and ask and discuss if you want project ideas

[turbid bobcat]

ok no but my fingers do thank me that I no longer need to write code for several hours at a time

like I was about to get one of those special keyboards and whatnot

now I dont write code unless strictly necessary

[fringe sphinx]

Read this pinned message, too: #python-discussion message

[still condor]

!warn @dawn mango Please read our rules and the channel description. We do not allow looking for developers on this server.

[inner wren]

:incoming_envelope: :ok_hand: applied warning to @dawn mango.

[mortal wedge]

Oh, I know the answer to this question!

||Nobody. If anything they'll have to push back the retirement age or try to rehire the people who want to retire||

[untold iris]

hi

[unreal niche]

Why should I learn python

[true harness]

it's fun

[unreal niche]

Idk if i will need it
Im a front end developer looking forward to being a full stack developer

I didn't gotcha
Do you mean node js
Or python

That makes perfect sense Thanks for the suggestion

I think that learning nodejs would be easier cuz i have already learnt Javascript

[spring ivy]

dude waymo pays 230k for L3💀

man most of these companies pay more than google 💀
Ig it may be partially cuz of the prestige and partially cuz google has a kirkland office.. idrk

stripe, waymo, roblox, all pay more than google for L3, palantir didnt have Levels it had general SWE so idrk how much they pay

[wind plinth]

Don't work at Roblox 💀

[spring ivy]

tesla fucking sucks big time.. ideally dont work there lol

140k and they will make u work ur ass off like 70-80hrs/week from office, no hybrid/remote

[wind plinth]

Why not just work for a smaller company?

[spring ivy]

less stable, lower pay, not like the work their will be more enjoyable for me either

[wind plinth]

like 50K USD a year salary is really good (at least where I live)

[spring ivy]

yeah maybe but idk i was earning 80k$ in india so i would feel pretty bad that i left that job and havent been able to make up the opportunity costs

then again that was probably unstable too....

[wind plinth]

Personally, I try to think about what I actually need the money for, and I can't think of what you actually need that requires more than 60/70K

And jobs at large companies aren't super easy to come by, and I don't trust them not to fire you within the first year considering how they're trying to replace everything with AI that barely works

[spring ivy]

any other companies that pay more than google?

and that are realistic for new grad dude who isnt absolutely cracked amazing?
I found stripe, waymo, roblox till now..

[spring ivy]

tbf where i live u would easily need 80k-120k just to afford rent for a family of 4

and the rooms will still be small aF

but yeah in india it was chill.. honestly a huge mistake to come to USA.. its way worse..

[white relic]

I work at a smaller company in the US. I don't get paid Bay Area wages, but then, I also don't have to live in the Bay Area

[astral ermine]

If you plan to raise a family, or help your partner or parents, the extra money goes a long way

Also, retire earlier

[peak halo]

or at all

[lime badge]

idk if this question is allowed but - has anyone here dealt with a company called Mercor? I did this AI interview with them the other day, and they almost immediately followed up with a job offer for like $45 an hour. I smell a scam here but can't find anything conclusive on them

the whole interview was about API design and AI integration but the job offer was for "AI video annotation" and required that I have access to douyin and their "seedance" model, which is giving me bad vibes

[haughty slate]

Any suggestions on deciding what job to get? I’m kind of in a “I don’t care what I do as long as it’s programming” mindset, which proves problematic to actually focusing on picking up niche skills to be a more competitive candidate

Or at least minimally competent lol

[pine sleet]

i think its a scam. i got an email from a hiring agency that was trying to hire for them and they keep spamming me with emails. it just seems really shady

except in my email they were advertising $150/hr which just doesn't seem right for a remote position

[turbid bobcat]

the platform seems legit

they do have some questionable practices like AI interviews, and seems like there's scams that operate on top of it but without the platforms consent

tho, inconsistent contract pay of 100-150 USD/h not enough to compel me to help train my own supposed replacement lmao

-# I already do that by participating on social media 🥲

[buoyant flame]

Hey

I need help 😭

I don't know why but I find SQl is kinda hard to learn

[smoky quest]

sounds like a question for #databases rather than #career-advice

[buoyant flame]

Yeah.. also I've no idea what to do after my graduation...

[smoky quest]

I mean, finding a job is a common thread

[buoyant flame]

Yeah... But like I don't know what kinda job I can get... I know python only ..

[smoky quest]

your degree surely had more classes than just python?

[buoyant flame]

Well my degree is total trash.. it's useless , so I'm doing self-study

[smoky quest]

you can still look at the topics for ideas

try things and see what you vibe with

[buoyant flame]

I'm 19 and I don't know anything about outside world work..

[smoky quest]

oh so you haven't been through a degree yet

[buoyant flame]

Next year I'll finish my degree

[smoky quest]

what degree is that?

[buoyant flame]

Computer science

[smoky quest]

I figured as much

is it a bachelor, masters, something else?

[buoyant flame]

Bachelor..

[smoky quest]

how did you fit a 4-5 years degree before 21 years old?

[buoyant flame]

No it's 3 years here .... I started college when I was 18

[smoky quest]

where is here?

[buoyant flame]

India obviously

[smoky quest]

not sure what would make it obvious

nonetheless, congrats

[buoyant flame]

Why congrats mate ? That I'm stuck here 🥲

[smoky quest]

you are almost through your education and ready to jump into your career!

[buoyant flame]

Yeah. I've got family pressure on my back ... My parents told me to get a job after completing my bachelor.. they're saying that it's useless to do masters... They have high expectations 🥲 Really you know sometimes I just want to disappear

[smoky quest]

masters can be useful too if that's part of your career goal

[buoyant flame]

That's my goals but not my parents so

[wet knot]

hi all. not sure if i'm in the correct channel. i've made a simple pyside application that reads a camera feed, detects barcodes (using OpenCV), extracts the code, and automatically types it into the currently focused application. I built this for the company i'm working at. The source code is availavle on github, but they are willing to pay me for it. what price should i charge?

[vast shoal]

They want to buy full ownership of the code and IP from you?

[dull belfry]

Hi here is my resume
I need advice

please note that the projects done from 2023-2024 during a phase where I was actively learning about web development

I am currently out of college and is looking for internships/ jobs in general software development

i like backend development

You can see that I haven't done any new projects for a while now
Is my resume weak?

[dull belfry]

This is the full resume

[wet knot]

no. they just wanted to pay me for use the application (like for license).

[tough gale]

Are you from aktu ?

[dull belfry]

nope

Thats a different university i think

[tough gale]

No it's apj abdul kalam technical university only formerly it's uptac currently it's aktu

[dull belfry]

not technical its technological

different state

[tough gale]

Where are you from ?

[vast shoal]

Here's my suggestion. Figure out what the rough hourly price of hiring a software consultant with your experience level is at your location. Calculate the development cost based on this and the time it took you to create the software. Subtract ~30-40% from that to account for the fact that it's a license and not an IP transfer. Charge them that for a perpetual license (of the current version only).

Charge them separately for support, maintenance or change requests.

You may wanna start at a slightly higher price when negotiating, so if they try to haggle you don't end up at an unfair price.

[wet knot]

it isn't a professional level application. but i just wanted to know the standard way of billing. how do i set the price if i do maintenance and change requests?

[fast fossil]

all that dead space in the skills section keeps throwing me off

[vast shoal]

Charge the hourly rate for a software consultant of your experience level at your location.

[wet knot]

thanks ❤️

[ocean pawn]

do yall put phone numbers on ur resume

[true harness]

yes, it's expected in the US

[ocean pawn]

oh okay i randomly seen something one day saying not to, dont think i kept reading to see why but im updating mine today

[next plover]

Just don't make your resume public to anyone

I Only share it to specific people it's not available on my linkedin or anything

[dull belfry]

skill gap😔

[fast fossil]

literally

[ocean pawn]

i see

[inner ibex]

Do companies hire python backend developers

[next plover]

if they are looking for backend python developers yes but your question isnt really answerable

where I work they would be looking more for JS or C# devs

[fast fossil]

you just answered it

[next plover]

yeah but its not helpful

[fast fossil]

No bidding. No connects. Just strategy, proof, and visibility.
This reads as if an LLM generated it

[vapid jay]

Yes took some help from gpt. Anyways the story is authentic and i think that is what matters 👍

[nocturne harbor]

!tempban 1062259074894417941 3d You've been warned before about not advertising/recruiting here.

[inner wren]

:incoming_envelope: :ok_hand: applied ban to @placid creek until <t:1772050633:f> (3 days).

[flat cosmos]

guys can anyone teach me py?

[hearty island]

.resources

fahhh what

!resources

[inner wren]

Resources

The Resources page on our website contains a list of hand-selected learning resources that we regularly recommend to both beginners and experts.

[vernal pond]

Guys I’m bored gimme something to code

[formal rapids]

@vernal pond I'd suggest going to do a 3 kata on codewars

https://www.codewars.com/kata/54bebed0d5b56c5b2600027f/train/python

[vernal pond]

Nah just something to make

[formal rapids]

its hard to recomdned without knowing any of your goals have you already made simple web api or any database systems?

Django is a good use of time to learn, or fast api, or anothe route would be learning Python Arcade library or a twitch app might be fun, depends on your goals

[vapid jay]

ok

[true star]

im 15 looking to go into quant finance, what should i be teaching me self within the next 3 years?

[fringe sphinx]

Take math class seriously: take the most advanced math courses you can. If you can, take a summer college course.

Learn programming. Experiment. Don't worry about what you program, just get experience coding.

And, learn/listen about markets and finance and the economy

Follow your interests, but find the 'fun' in math.

I'm not a quant, but I work with them. I prefer to stay on the SWE side

[barren lotus]

how do you stand out in an interview?

[true star]

i do, im currently top of my year level, i've competed in math competitions and i do extremely well in school, what im mainly looking for advice wise is specific concepts to master, i've heard quant work is heavy interms of statistics, linear algebra and calculus, right now i've just been messing around with python and mastering current maths concepts im being taught but i was just curious if there was anything i can do even if its niche to give me a large edge

my end goal is to be able to create a working quantitative model or just general financial model that actually shows results within the next 3 years, do you think thats attainable?

[fringe sphinx]

'Edge' isn't really the way to look at it. Think of it as things you can do to build your foundation.

At your age, a strong programming foundation would be good... not specifically quant, just being able to tackle projects of reasonable complexity

[true star]

when i say edge i mean building that foundation which can allow me to have an edge if you know what i mean

[fringe sphinx]

Even building a game, or a web app or a calculator app for a class, would all be useful ways of building your core coding skills

[true star]

what projects you suggest working on, i know you said just code anything but i feel like theres specific concepts within python which more so apply to financial applications of it since it is a broad language

[true star]

i use to do game development not coding though but i've moved away from that

[fringe sphinx]

Theres a few different layers to this, ranging from 'being a fluent programmer' to 'building data analytics skills in Python' to strong DSA-type skills to being able to implement high performance algorithms using a stack that quant firms use

(The last may involve c++ or other languages)

[true star]

yeah i've done my research c++ mainly used for HFT quant roles rather then generalised ones, bit more niche

[barren lotus]

i was told that simply answering the questions well wont make one stand out

[fringe sphinx]

Yup, it's a big industry and there's different things people mean by 'quant'

[true star]

im younger so i dont have any real world experience but i've heard companies care a lot not just about your talent in the field you're working in but also general social and communication skills, people enjoy working with people who can communicate well

[fringe sphinx]

That's true in, well, every field. Being able to communicate a technical idea is very important.

[true star]

when you're looking at quantitative finance there are quite a few roles, do you reckon quant roles are safe from being replaced by ai anytime soon?

[serene kite]

hi, just wondering if this unit is good for a cs major:

https://www.qut.edu.au/study/unit?unitCode=MZB125

[true star]

@fringe sphinx where are you from id you dont mind me asking

[fringe sphinx]

I believe so. AI tools improve dev and quant quality of life: it's easier to... say... generate a chart for some data. Or, regress some data set. But interpretation and planning and design is not a skill of modern AI

[fringe sphinx]

US

[true star]

what about future ai? with the rate its evolving its hard to say

[fringe sphinx]

Not really sure, looks like a pre-calc type course?

[serene kite]

i have already done calculus in high school

i wanna end up having a good knowledge around machine learning

[true star]

you learn pre calc in university wtf 😭

[fringe sphinx]

It's hard to say, ofc, but LLMs are unlikely to significantly improve. What will improve is our ability to wield them: orchestration, planning, multi agent pipelines, etc

So, I see AI as a way of doing less boring work, and I like it.

[mortal wedge]

Also depends on what regulation looks like when it finally catches up to the tech.

[true star]

Nice im Australian so theres limited opportunities here compared to the US do you think its worth it trying to go to a US ivy league rather then an Australian uni or is it not worth the debt + inconvenience

[fringe sphinx]

Different countries have different programs. Some Uni's in US have pre calc for people who didn't take it in HS

[mortal wedge]

Even with LLMs rapidly evolving, the core architecture isn't changing too much, it's a lot of like... tools bolted on to the sides.

[true star]

do you think the quant job market will shrink or grow with ai?

[mortal wedge]

I'd argue LLMs haven't fundamentally improved a lot at all, but the companies have gotten better at using them and patching their shortcomings

I strongly suspect AI will shrink the job market as a whole.

I mean, it already is, I don't have to suspect anything.

[serene kite]

same here

[fringe sphinx]

I don't know AU's universities. I do know many people recommend a graduate degree in the US rather than an undergrad. But, if you can get into a top tier US school and can afford it? That's hard to pass up

[true star]

i beleive the bubble will pop in a few years lol its hyper inflated atm

[serene kite]

im thinking of getting my degree in au then going to US for jobs cause you can avoid the huge interests and debts on your student loan like that

[fringe sphinx]

Hard to say, the job market is changing not because of AI but because of other economic factors. The market ebbs and flows

[mortal wedge]

I agree that it's hyper inflated. But I also think there is legit value added there. I don't know how and when things will change.

The market has definitely been pulling an ebb in recent years though

[fringe sphinx]

I believe there'll always be good $$ for good engineers. The market may or may not reemerge for mediocre engineers.

[true star]

best uni here is melb which is roughly top 20 globally, we got a few others in the top 20-30s aswell, main issue is job market theres obviously going to be alot more quant roles in the US, do you think US based firms would hire people who want to university internationally or stick to hiring locally?

[mortal wedge]

Revised US job numbers for last year across all sectors was insane. net average of 300 jobs per state per month

[fringe sphinx]

I don't hire quants, so that's a little too specific for me. But, a US graduate degree is probably good enough for even the most 'elitist' of job opportunities

[true star]

yeah ofcourse it adds alot of value but the volume money being put in isnt being reciprocated close to enough

[mortal wedge]

I agree somewhat, but I think in practice this depends on companies being able to determine good vs bad engineers, which I am losing hope they'll be able to do so.

[true star]

what do you do if you dont mind me asking?

[mortal wedge]

Full agree. Current growth rate is unstainable if for no other reason than the raw materials for the physical components of the computing infrastructure needed for data centers just isn't there.

[fringe sphinx]

I do data stuff in Finance. Everything but a Quant, is what I say.

[true star]

yep when a $500b data centre is planned and starts constructions all in a year it raises concerns about logistics and ethics

[true star]

are you employed or work independently?

[mortal wedge]

How do you like finance? I might be moving into a finance heavy or at least finance adjacent role

[fringe sphinx]

The interesting topic is how it's funded. If you're interested in quant and economics, listen to https://youtu.be/sAjGXSdDkhQ?si=CJjsOREJMY3OzYAd

[mortal wedge]

(Interviewing for a staff AI position to try to improve software that helps providers get claims approved by insurance companies)

[true star]

what industry are you currently working in and how old are you?

[mortal wedge]

Medtech, late 30s

[fringe sphinx]

Honestly? I have zero interest in the field of finance. But they do have good data problems.

[true star]

i've already looked into that dw im pretty up to date on global economics, i understand how its being funded

[mortal wedge]

Honestly I would just love it if my job was making insurance companies lose money

[fringe sphinx]

Employed. Small tech firm, so it feels independent.