Can We Learn Generative AI Without Knowing Machine Learning And Deep Learning?

hello all my name is Kish Nayak and welcome to my YouTube channel so guys from past 2 months I have been getting this specific question they're asking Krish can we directly move to generative Ai and probably start using this llm models API for performing most of my task without learning machine learning or deep learning now this is the question I'm telling you guys many many people have asked me they have dropped me a lot of mails with respect to that and it is obviously true right now see if I talk about machine learning deep learning you know the concepts that you are specifically using over there and when you compare it with something like an llm model which can probably do almost every NLP task or you probably consider large image model where you are able to com probably do all the computer vision task so people will definitely have this question in mind can we just directly move into generative AI right instead of just learning machine learning you know starting from Basics probably it may take another 6 months to learn everything from Basics so so I'm going to answer this specific question in this video please make sure that you watch this video till the end because you will be able to get lot of info out of it right now to answer this question I would definitely like to say depends from People to People right like if you say me okay Krish would you like to just like let's say I'm starting machine I'm probably starting to learn data science right now if you ask me Krish would you like to directly jump towards generative a right I'm I I I also like coding okay so it is better that I can probably jump into generative a and start doing all the task that is related to any business use case and probably solve it but if I probably answer it I will never do it because see my perspective is to probably learn everything from basic and keep on building Concepts on top of it now also anything probably that comes new in the AI market right first of all I'm able to relate it from where it is probably coming right what what was the base what are the building block behind that specific algorithms right and this gives me the best feeling to learn something you know now there will be other people also who are very good at development and they probably don't want to learn everything from Basics they don't may like maths you know uh and I'll be also talking with respect to interview okay now in interview the concept is completely different so let me talk about the other type of people who are very good at coding who just want to learn AI to implement use cases and they are really not interested in maths for those kind of people you can directly jump into generative AI but again there is a drawback in that also right it's okay so you can directly use the apis you can directly probably build uh uh use cases you can build business use cases you can solve them you can do the deployment you can probably do each and everything okay and considering that if you are a good coder right you can definitely do that right prerequisite is definitely Python programming language you really need to be a good Python programming language I'm telling for the people who are not even starting from machine learning or deep learning if you directly jump to generative AI trust me you really need to be a good coder right so that is the reason why I say python right with respect to any applications that are probably coming any llm models that are going to come in the future right python is must you go and see it right uh every Library comes up with its own python SDK where you can probably access it in the form of API with keys uh even open AI right open AI many people are using it you just need to have the openi key later on with the help of python SD with the help of openi libraries you're able to do or Implement any use cases so python is definitely the future right and many many companies are specifically using python so I'm not saying that okay you always have to learn machine learning from Basics but it is always good to have why I'll tell you in just in some time right if you tell me Krish go and directly learn generative I will never do that you know I will definitely go from Basics and that is what I've actually done from 2014 and that way only I have conveyed all my information in YouTube but again there are many people who are still working in companies who are developers who are doing another role and they given some kind of llm task because they already a code developer over there right and with the help of APs they able to implement all the things over there so is it a good idea again it depends on you try to find out which kind of person you actually are whether you are the first category person or the second category person right uh things uh like all the basic things I probably think if you're not good at maths if you don't like learning all the things but at the end of the day when you're solving business use cases like ml use cases when we discuss in the team you know you will not be that presentable because you'll not be able to answer anything because your Basics is not strong at all right and this is the drawback obviously you'll be keeping quiet okay they'll say that okay go ahead and implement this go and do the fine tuning of this right but when it comes with respect to idea generation when it comes to core thinking when it comes to understanding about the use case trying to apply techniques in the use case right there I think you'll probably fail right and this is the drawback I will be very much Frank with you I'm not saying that you cannot go through that path you can definitely go through it you can also go and work in the companies that is of no you no problem at all now considering this there is also one drawback in the interviews if you are a person who have learned probably everything from machine learning Basics to deep learning Basics you have build that Concepts in interview they mostly ask about this basic building box right they don't directly ask you about API they they they come to that point on the lat Point like when they ask you about projects at that time they may probably ask you with respect to that how how to find tuning with their own custom data set what kind of projects you have actually developed but at the end of the day basic building blocks is super important okay I'm not against any any kind of people I'm I'm saying that these are the drawbacks that may happen in most of the interviews that I've have seen right now all my students who have probably uh made transition in this specific year they are working in some kind of llm projects and all but they were asked in interviews in the uh with respect to basic things like statistics machine learning deep learning Concepts like algorithms all these things were specifically asked and you can probably watch my successful transition playlist also there also have spoken to people right it it was a kind of a podcast and you can also understand that yes what kind of questions was specifically asked over there yes you can Implement any llm models you can probably use it for solving any business use cases but this is one of the major drawback in interviews I'm very much Frank with this guys if you don't agree with me no go ahead and see any podcast of mine or anyone who have made successful career transition llm and all are something that is built on top of everything right and right now the way that we are specifically getting llm apis openi apis Lama index anything all these libraries that you're specifically using it has really become easy to use you just need to probably use insert a python hdk component over there try to use that API and try to just write the on the code right and try to implement an end to end project so these are two ways the reason why I'm making this video because people may be starting data science right now and they may be seeing all the generative AI task that is happening right all all the tasks are very much easy and just by writing a snippet of code but will that be just sufficient right to clear the interview anyhow they will ask you things that are something related to the basic part with respect to machine learning statistics linear algebra or deep learning and later on once you probably get into the company it won't matter that much you can probably directly do the projects because over there the focus is more to solve the business use cases people will see that how you able to solve the business use cases and if any company is doing something from scratch then the intensity the complexity will start increasing at that time you'll not be able to even talk about things you will not be able to suggest any ideas itself right so for all the people who think that they can directly learn generative AI please go ahead with it but at the end of the day for clearing interviews really need to focus on the basics part focus on the projects projects is also very much important mlops is also very much important soon I will be creating a road map for the data science in 2024 for learning data science in 2024 to get jobs right and there will be two major things that I'll be discussing about for people who have already started like let's say one year back two year back or 6 months back they starting from machine learning deep learning and for people who are about to start and they directly want to move into G so it is always a good idea to make your Basics strong right and again it depends from People to People if let's say if you get a specific job um the JD is just with respect to llm models and all so there you can directly probably just learn regarding generative a and then go for the interview but I hardly feel that they will at least start something from the basics so if you agree with this particular point please do make sure that you comment down in the section uh in the comment section itself I let let's go ahead and see like how many people agree with my points that I'm probably saying and this is from all the experience that I've got by talking from people who have already made this kind of transition so I'm not telling a kind of lie I've just understood like what is favorable for you right so yes this was it for my side I think I hope you like this particular video I will see you all in the next video have a great day thank you one all take care bye-bye have a great day

Channel: Krish Naik

Views: 34,352

Keywords:

Id: 0R1mA_MVItI

Length: 9min 24sec (564 seconds)

Published: Sun Dec 10 2023