NOTE: The first 100 people to use the promo code STATQUEST at devra.ai will get a free credit worth $10 US. Support StatQuest by buying my book The StatQuest Illustrated Guide to Machine Learning or a Study Guide or Merch!!! statquest.org/statquest-store/
@shrutiiyyer27839 ай бұрын
My question is, how then when most entry level work is automated, will people be able to learn to get from 0 to 60 when they don't know how to get to a 20 or 40 because that bit is something that AI has taken care of?
@I.II..III...IIIII.....9 ай бұрын
They don't give a shit at the moment
@firstlast68139 ай бұрын
One does not need to know how to file files, to be able to analyse what is written in that files? Is my experience as a cashier crucial for my management position in a supermarket? Of course all experience is helpful incl. the experience from entry level jobs, but I have the impression that in many cases experiences from entry level jobs are not truly necessary for higher level positions. And also: Training and education.
@shrutiiyyer27839 ай бұрын
@@firstlast6813 i see your point but how does it work for a transition so quick? I mean the work that entry level folks were doing wasn't too basic. It was something that made them get their hands dirty, play with the data and learn from it. We learnt DS this way back in our days and we still need those strong fundamentals to be better at the science of data.
@devra_ai9 ай бұрын
Great question and I think the answer is people are going to adapt and learn what is needed. There have been many sea changes in programming that obviated previous skills. We don't have to know assembly language thanks to compiled code; debugging is far easier due to stack traces; code quality is far easier with IDEs; versioning and collaboration is almost trivial now thanks to source control like git. Each of these advances got us past these frustrating parts of development and let us focus more on the truly fun parts of coding. I've worked deeply with AI for development and can say for certain it is similar. It is *fantastic* to write unit tests, refactors, boilerplate -- boring things! -- but it struggles with code that requires true creativity.
@iskrabesamrtna9 ай бұрын
I really enjoyed the conversation. My pythonic background never stopped me to write the search engine in Rust - so I can relate and confirm that LLMs are exceptional in coding. Of course it depends of an LLM and the general understanding of the concept implemented. I wish you luck with the future development, and I'm interested in testing DevraAI :)
@statquest9 ай бұрын
BAM! :)
@BrianRisk9 ай бұрын
This was a great time! Do you think my first answer is classifiable as a "monologue"? 🤣
@statquest9 ай бұрын
BAM! A dramatic monologue.
@FlokiGuy8 ай бұрын
Hello! Can we expect a video on structured state space models (SSMs) architecture of neural networks and Mamba and S4 model, it would be very cool to see an explanation from you. Thanks!