"back in the years I thought AI was scary, data science required PhD, and computer vision was magic. now it's all just 5 lines of code in Python" - so relatable
Why learn the details when you only need to write the upper-level abstracted code to get what you want out of the computer? import THE_SHIT_I_WILL_NEVER_UNDERSTAND; do_what_I_want_lol();
Me: **Thinks I hate Python.** This Guy: **Mentions R as an alternative.** Me: Python is a gift from the gods. Let us sacrifice everything at the mouth of the s n a k e .
I mean Tidyverse is pretty neat, and beats pandas and matplotlib hands down, but R as a language is pretty much unusable for anything more complex than a few lines of code.
Imagine a parallel universe where Python were not invented and all data science machine learning stuff are based on R and small companies use R for production.
R is great for making figures with plotly (but you can do that with python or other languages). God have mercy on anyone that is actually using R to do anything more than that. Even a simple string is handled inconsistently and differently by packages. From my point of view, each package is effectively a different language.
" Of course we only do our research in Python, our prediction model is then running in C!" wink 😂 You bloody nailed it for the major part of the industrie though
Well, not really. When you're using Python, you're most likely using C. Most databases and data science libraries are under-the-hood, written in C, C++, or some other languages that are meant to be fast. Python is a great language to glue these really well-written libraries together into a useful product, but CPU-bound computational problems aren't its forte. Instead, Python is really powerful because it makes it easy to call into off-the-shelf libraries written in other languages to solve practical problems, rather than spending time trying to reinvent every little thing in Python.
Love it! Not sure my frail ego could handle a similar roast of R, but please do! Just the hate wars between different dialect fanboy groups (old, grumpy professors sticking to base R and mail lists like it’s 1996, sweet and naive tidyverse beginners, masochistic data.table users obsessing about speed and efficiency with no regard for code readability) could take a whole video. The only thing everyone can agree on is to universally hate the proprietary statistical software packages :)
i started to love this series. python used to be such a quiet and nice community, where it was about finishing the code in simple readable manner, and of course we always talked about how to replace stuff with C :D but the closest i came was writing something in Cython. not complaining, that now i easily get python jobs tho. lately it feels like every second job searching java programmer coming into python jobs brings in more and more "required" architecture ideas. i try to accomodate their requirements, write nice type-hinted dataclasses, build layered architectures, find ways to express interfaces, but the nagging does not stop. Damn you, Java.
This reminds me of people speaking languages significantly different then their native tongues, I wouldn’t be surprised if we needed just as much effort mastering a new programming language as any other natural language.
I was one of those Java guys who decided to switch to Python before it (Python) becomes the next COBOL. And I did moan about the lack of type info, and laughed at how everyone now says you need the if __name__ == '__main__' ... and basically as time goes on Python becomes more like Java. But I have to say I've come to like it. I'd reach for Python before Java these days and I never thought that would happen. I disagree with everyone who says it's a 'simple language' though. It's hellishly complicated compared to Java, but with the complexity comes a lot of good stuff.
Looking forward to listening to senior x86 asm programmer, junior Ruby developer, Rust nightly team lead developer, Cloud platform infrastructure and GoLang engineer
you can use Pycharm, VSCode, or any IDE/text editor to write and debug the essential part of your code and use Jupyter Notebooks just to present your result.
As a person who recently started picking up Python for AI development, I will probably be using a lot of these jokes You have no idea how many times I quote the Senior JS video
@@sazk4000 that’s fair, the package ecosystem is getting there for some cases but is lacking in others. Still, only been around for 10 or so years so definitely going to be improvement over time!
@@sazk4000 Keep checking back in. I felt the same way when I first tried. Went back 6 months later and there was coverage for all the utilities I need. The ecosystem is building out extremely quickly.
unsure about julia. i don't like the idea to have a language with so high and many goals, and some of the disputes i read were red flags for me. i loved the concept early on and wish it all the best. i rather went towards rust. I did the same thing back then when it was ruby or python, in 2004. I found python was more grounded and humble, so i went for the snake, even if everybody was talking about ruby.
The thing that Julia needs is (1) better documentation, (2) better & longer support & maintenance for it's libraries/packages, (3) better onboarding for newcomers as to which packages/libraries give them the utilities they need for a given project, and (4) one of the FAANG companies to spend like 5% of their R&D budget on doing something in Julia over Python so (a) they see the insane C-like performance gains with similar simplicity of Python-like syntax that Julia offers, and (b) so other people think, "Huh, a FAANG company invested in this language, maybe I should take a look... 🤔"
1:26 Automation! Innovation! Revolutionary!.. PAIN! I spilt my beer. The cadence, the cut, chef's kiss! Really, the feeling of the vid being a selection of the juiciest bits from hours-long ramblings, while knowing it was all planned, adds so much to the enjoyment😅
@@qwertyuuytrewq825 Basically lists indices in Python start with 0 (just like almost any other programming language) and indices in Matlab start with 1. So one starts with a zero (Python) and one is a zero (Matlab).
The symbol python comes python python making python works but whom make python that make python works ? thats another library we need to import , or is it native supported ? or is only works in jupyter , the path , the env , so much option , so little standard ,everything is everywhere , the global variable
I have one question, why did you say that “we also don’t use python for production code, do we?” What does it mean actually. Why we can’t use python for production
this is brilliant. just to add: python is amazing. anything to be done should start with: why not in python? lots of reasons why not (does it need to be faster? does it need to run on a specific platform?), but anyway great to start with python, get the logic right, then transpose that logic to faster/platform-reason to use a different language. it's like bridging language
@@Lodinn def. readability is always emphasized and can't be emphasized enough. even if *you can't/don't want to/don't have time to, optimize; other people can more easily see the intention and improve where necessary
@@yds6268 in the health sector, and other sectors I've worked in, models built by industry are typically in python (prolly cos of PySpark, TF and Keras) whilst acamedics continue to use R. Some models brought across from unis might survive as R but many will be rewritten, particularly if they are require high compute. I've only seen one model built from scratch using R where I work, and that was by someone who'd just graduated with a stats degree. As an aside, I personally think R syntax is bonkers, but people still love it!
"back in the years I thought AI was scary, data science required PhD, and computer vision was magic.
now it's all just 5 lines of code in Python" - so relatable
I've felt that
Raw genius
5 lines? are you doing linear regression or something
Why learn the details when you only need to write the upper-level abstracted code to get what you want out of the computer?
import THE_SHIT_I_WILL_NEVER_UNDERSTAND;
do_what_I_want_lol();
So true 😂
"Now I sound like a PHD, but don't worry, I am only a... data scientist for quantum physics operations"
Hello, I wanted to ask if you could do an interview with a Linux Enthusiast, I think that'd be pretty interesting to watch!
YESS PLS we would love that
i love to see that
hell yeah man
I use Arch, btw.
I use Arch btw
Me: **Thinks I hate Python.**
This Guy: **Mentions R as an alternative.**
Me: Python is a gift from the gods. Let us sacrifice everything at the mouth of the s n a k e .
R is great though
I'm pretty new to data sci, but I prefer R currently
I mean Tidyverse is pretty neat, and beats pandas and matplotlib hands down, but R as a language is pretty much unusable for anything more complex than a few lines of code.
Imagine a parallel universe where Python were not invented and all data science machine learning stuff are based on R and small companies use R for production.
R is great for making figures with plotly (but you can do that with python or other languages). God have mercy on anyone that is actually using R to do anything more than that. Even a simple string is handled inconsistently and differently by packages. From my point of view, each package is effectively a different language.
Ive once tried to read a piece of python code from a mathematician and it was like trying to read a doctors handwriting.
What was it like?
@@biblebot3947 I'm guessing they would skimp on variable names and use individual letters instead. And probably filled with lambda functions.
missing line; "you can run a jupyter notebook in vs code now --- It's still a pain to debug."
I think most of those people (including me) have never seen VS Code... or even Windows more recent than XP.
Fun fact: Jupyter notebooks are used for production code. It includes quite advanced rich-output facilities, after all.
Can't wait for the interview with the Matlab developer. That's going to be a rib breaker for sure..
"One starts with a zero, and one _is_ a zero". Best line ever.
“Matlab developer” is somewhat of an oxymoron
@@Austin-hm6qq what about 'Mathematica developer' ???
@@rapidreaders7741 Please explain!?
They all use python matlab is demode. With matplot lib and jupiter who pays for matlab
"pip install machine_learning"
"pip install HAL"
insert someone yelling at you to use conda
the underscore is 🤌
that is the PyThoNiC way to name a var
Amazing humor, I really love these videos, keep doing great things!
"Imagine someone writes skynet in a Jupyter notebook!". That line killed me. I don't exactly know why but damn that's funny.
It's the delivery. He 100% means what he says, only to be 100% in disbelief.
Me: so how do you store variables in python?
Junior python dev: What's a variable?
Wanted to upvote but it said 69. Abort mission!
senior c++ developer : you wouldn't get it.
@@mesaber86 You can do it now, someone took the bait already.
That's the great thing, you don't have to know, you're just assigning values to objects
@@balazsh2 well, knowing how data is stored and the types of that data is definitely useful.
" Of course we only do our research in Python, our prediction model is then running in C!" wink 😂
You bloody nailed it for the major part of the industrie though
To be fair, that's the rational thing to do, considering how C is
Well, not really.
When you're using Python, you're most likely using C.
Most databases and data science libraries are under-the-hood, written in C, C++, or some other languages that are meant to be fast.
Python is a great language to glue these really well-written libraries together into a useful product, but CPU-bound computational problems aren't its forte. Instead, Python is really powerful because it makes it easy to call into off-the-shelf libraries written in other languages to solve practical problems, rather than spending time trying to reinvent every little thing in Python.
@@yvrelna Yes, that is the joke.
"Import numpy... and then import the rest"
Every indian Python tutorial in a nutshell.
These are some of the funniest videos I've seen man, I would really love to see your take on Linux!
Thank you!!!! I've been waiting for the new upload for a long time.
"We've created skynet" 😂😂😂
The Python dev sneaking a little Julia in their free time to taste what a good language is like XD
"Use python to predict inhouse coffee usage" really made me laugh
"what can I say: automation innovation revolutionary PAIN"
Hit me so hard
Skynet in a jupyter notebook really made me laugh… 😂
love it! Especially the Jupyter jokes xD
I had to lift over code from Matlab to python and I love the "one starts with a zero and one is a zero".
As someone with 20 hours of programming knowledge I find your jokes motivating to keep studying!
Seriously loved your senior JS developer video alot!!! Please bring more JS videos
Love it! Not sure my frail ego could handle a similar roast of R, but please do! Just the hate wars between different dialect fanboy groups (old, grumpy professors sticking to base R and mail lists like it’s 1996, sweet and naive tidyverse beginners, masochistic data.table users obsessing about speed and efficiency with no regard for code readability) could take a whole video. The only thing everyone can agree on is to universally hate the proprietary statistical software packages :)
OMG! That sentence in the brackets just perfectly describes R community!
data.table is lit. so is collapse. better than pandas.
Thank you so much for enable subtitles
i started to love this series.
python used to be such a quiet and nice community, where it was about finishing the code in simple readable manner, and of course we always talked about how to replace stuff with C :D but the closest i came was writing something in Cython. not complaining, that now i easily get python jobs tho.
lately it feels like every second job searching java programmer coming into python jobs brings in more and more "required" architecture ideas.
i try to accomodate their requirements, write nice type-hinted dataclasses, build layered architectures, find ways to express interfaces, but the nagging does not stop. Damn you, Java.
Goddamit this is so relatable
This reminds me of people speaking languages significantly different then their native tongues, I wouldn’t be surprised if we needed just as much effort mastering a new programming language as any other natural language.
I was one of those Java guys who decided to switch to Python before it (Python) becomes the next COBOL. And I did moan about the lack of type info, and laughed at how everyone now says you need the if __name__ == '__main__' ... and basically as time goes on Python becomes more like Java.
But I have to say I've come to like it. I'd reach for Python before Java these days and I never thought that would happen. I disagree with everyone who says it's a 'simple language' though. It's hellishly complicated compared to Java, but with the complexity comes a lot of good stuff.
Thank for you for making these videos, to give me something to laugh about. And to allow me realize how sad my programming life has been.
Love the Matlab jokes 🤣
Would love to see a C developer interview
Pointers!
Pointers!
Pointers!
you mean resource constraint, memory leak, recursive headers, seg fault.
@@vitaminncpp its the programmers job to manage memory. Not my fault you pointed to the wrong spot
@@ImperialFool in java there is no such thing as memory.
@@ImperialFool Don't foget to use malloc and free!!!!
Looking forward to listening to senior x86 asm programmer, junior Ruby developer, Rust nightly team lead developer, Cloud platform infrastructure and GoLang engineer
Amazing - please do functional, Scala, spark :)
As someone who first started programming in MATLAB, this hits, especially the ending
Matlab is the superior product.
@@Person-jw7mk yeah but only because python isn't a product
@@AirAtNight6977 Superior tool. Whatever.
@@Person-jw7mk Nope.
I did GUI programming in MATLAB once.
Let’s just say, a matrix-centric language is not suited to GUI programming.
Python is riddled with problems while MATLAB IS a problem
Just finishing the video, an ad pops up: “Right now there is a shortage of AI and Machine Learning Professionals…”
Raking in that sweet pro add revenue.
Oddly enough, there is. Too bad all these people fresh from online courses are quite useless.
I dropped out of an AI course when I was in uni and I mostly blame Jupyter notebooks for that
what's wrong with it?
you can use Pycharm, VSCode, or any IDE/text editor to write and debug the essential part of your code and use Jupyter Notebooks just to present your result.
I hope you find the motivation to go back at it! Don't let the tools hinder you from having fun :D
i don't even know what it is and i already hate it
use vi
Please do a video about data engineers!
Coming up: interview with junior COBOL developer - would be 5s long
Most of the junior Java guys I meet remind me of the junior COBOL guys (like me) from back when I first started.
As a person who recently started picking up Python for AI development, I will probably be using a lot of these jokes
You have no idea how many times I quote the Senior JS video
They are more true than they should be
Such a messy language, I love it! No, I do not recommend
"Guys I get the jokes too! Ha!" Who asked?
Documentation?
hehehe
"I LOVE IT!"
The rolling R had me ded 😂
AUTOMATION, INNOVATION
gets me yet again lmaoo
Bro the Julia joke had me dying laughing. I did not see it coming the way it panned out.
“Let’s not mention Julia” 😂 it’s a really good language, I definitely recommend python devs try it out
i did.. went back to Python... it just had more utility
@@sazk4000 that’s fair, the package ecosystem is getting there for some cases but is lacking in others. Still, only been around for 10 or so years so definitely going to be improvement over time!
@@sazk4000 Keep checking back in. I felt the same way when I first tried. Went back 6 months later and there was coverage for all the utilities I need. The ecosystem is building out extremely quickly.
unsure about julia. i don't like the idea to have a language with so high and many goals, and some of the disputes i read were red flags for me. i loved the concept early on and wish it all the best. i rather went towards rust.
I did the same thing back then when it was ruby or python, in 2004. I found python was more grounded and humble, so i went for the snake, even if everybody was talking about ruby.
The thing that Julia needs is
(1) better documentation,
(2) better & longer support & maintenance for it's libraries/packages,
(3) better onboarding for newcomers as to which packages/libraries give them the utilities they need for a given project,
and
(4) one of the FAANG companies to spend like 5% of their R&D budget on doing something in Julia over Python so (a) they see the insane C-like performance gains with similar simplicity of Python-like syntax that Julia offers, and (b) so other people think, "Huh, a FAANG company invested in this language, maybe I should take a look... 🤔"
Anyone knows the IDE and settings at 1:47? I've never seen that top panel before
Spyder
Next, we want to see "An interview with Richard M. Stallman"
I have a spill-resistant keyboard (Corsair K68).
Today it saved me because I spilled half of my coffee during the Julia "this is not for recording".
Ahahaha “Imagine someone accidentally writes the Skynet in Jupyter notebook” I’m crying!
Automation, innovation, revolutionary ....... pain
I'm ded
Make a interview with the linux guy who shouts use linux everytime you face slightest glitch in your non linux OS
vim
I would like to interject fo a moment, what you are referring as Linux is in fact GNU/Linux...
The rest is history 😅😅😅😅
Hey, that's me
@@tablettablete186 See this is the guy I'm taking about 😂
omfg why is it so funny when u jump cut repeating words. jupyter jupyter jupyter ahahah 😂ffmpeg 🤣🤣
Amazing! Waiting for the pure C developer interview!
an interview with an empty chair?
1:26 Automation! Innovation! Revolutionary!.. PAIN!
I spilt my beer. The cadence, the cut, chef's kiss! Really, the feeling of the vid being a selection of the juiciest bits from hours-long ramblings, while knowing it was all planned, adds so much to the enjoyment😅
Missed opportunity - "Cython whats that ?"
I died at the matlab finisher. Mortal Kombat has nothing on that upper cut.
Maan, that guy is already a LEGEND!
Most of those are really spot on, especially Python and JS. I think C++ one could have been a bit better though.
@@reav3rtm true
"Let's not mention Julia" oof. I've been hearing that for the past 4 years at work.
It’s good, and the ecosystem of packages is getting there too
Lol this guy is soooo funny. I actually like this channel now. Please keep 'em coming Lol!!!!!!!!!
I've watched this a couple dozen times already. Man, please, come back with new videos. We need more *revolutionary pain*:) Cheers!
do one of these for the Julia language next! :D
Can't wait for the senior Go developer. We all are waiting for that, actually.
The ending made me laugh so hard!! 😂😂😂
I definitely need assistance understanding last joke about zero
@@qwertyuuytrewq825 Basically lists indices in Python start with 0 (just like almost any other programming language) and indices in Matlab start with 1. So one starts with a zero (Python) and one is a zero (Matlab).
@@chiroyce Thank you! Got it )
I literally finished a Monte Carlo Radiative Transfer Simulation course this week...using Jupyter.
*Back in mah days, we did it in C*.
Was it fun though?
Love those videos like it's insane keep going man !
This needed a Part2.
I actually use Jupiter as a debugger.
Ite very useful to have inherent breakpoints and keeping objects in memory.
spot on with everything, especially the julia part
Everyone is so focused on the Python discussion that he thinks that no one will notice that one of his shirt buttons aren't done properly.
This is sooo funny. I'm going to make Senior Python dev videos now
1:45 - what ide is he using ?
The symbol python comes python python making python works but whom make python that make python works ? thats another library we need to import , or is it native supported ? or is only works in jupyter , the path , the env , so much option , so little standard ,everything is everywhere , the global variable
"revolutionary....pain"
So fun!
And I am technically a Python developper. Had more fun with the Monty Python, or you content, than with this language 😉
pls do neovim next with the aspect of building your own IDE from scratch with plugins ... LSP, cmp etc
Please do Solidity next.
The "Revolutionary PAIN" got me good, ngl xD
As someone who writes JupyterLab extensions ... Jupyter is a pain.
1:46 What font is that?
I have one question, why did you say that “we also don’t use python for production code, do we?” What does it mean actually. Why we can’t use python for production
I never liked the idea of Jupiter Notebooks and have refused to try or figure them out
First, import numpy then import the rest: keras, small ins, outs .. and we've created skynet!
:)
Oh no…this is what I must sound like to my colleagues (!). I think I’ve said every line in here lmao
Please do FORTRAN sometime soon!
Can't wait for "Senior Docker Engineer" :D
this is brilliant. just to add: python is amazing. anything to be done should start with: why not in python? lots of reasons why not (does it need to be faster? does it need to run on a specific platform?), but anyway great to start with python, get the logic right, then transpose that logic to faster/platform-reason to use a different language. it's like bridging language
More like, you can test things for cheap (development-wise) and only optimize what worked out in the first place.
@@Lodinn def. readability is always emphasized and can't be emphasized enough. even if *you can't/don't want to/don't have time to, optimize; other people can more easily see the intention and improve where necessary
@@johnc5258 in other words, pseudo-code that computer can understand!! fucking magic
Nice touch, the professional here dressed exactly like his PhD student, except that his shirt is tucked in.
well jupyter is in the sky so skynet makes sense
Hope you do one for R
Definitely - it's wierd how massive it remains in academia compared to industry
@@azursmile I always thought R was more prevalent in the industry
@@yds6268 in the health sector, and other sectors I've worked in, models built by industry are typically in python (prolly cos of PySpark, TF and Keras) whilst acamedics continue to use R. Some models brought across from unis might survive as R but many will be rewritten, particularly if they are require high compute. I've only seen one model built from scratch using R where I work, and that was by someone who'd just graduated with a stats degree.
As an aside, I personally think R syntax is bonkers, but people still love it!
@@azursmile You can use tensorflow with R
@@maythesciencebewithyou sure, and Spark, but it's not as common as with python.
Sr. Ruby interview pleaseee!!!
Interview with a QA Engineer please:)
In house coffee usage lol
I always thought that I was bad at debugging jupyter notebooks, but it seems that is true for everyone XD
"Let's not mention Julia" - this made me lol hard
I love these videos!
And we don't do production code in Python ...
UA-cam takes a deep sigh ...
julia rulz... Somehow, being called a zero compared to Python is a little harsh on matlab...
Do you have someplace we can make donations?
the fact that there's no ruby video makes me sad about the relevance of ruby these days
Can you take on LISP
Those small C pointers always make me giggle... ;)
Pascal would like an interview
hey, why so short?
will it be a part 3 then?