Hello everyone! I'm creating a brand new channel dedicated to in-depth guides and courses in AI & ML 😎! Subscribe to my brand new channel 'School of Machine Learning': www.youtube.com/@SchoolofMachineLearning
Haha, same thought exactly ;) "When the hype will die down.."? Are you kidding me? It's only gonna go higher and higher. With current direction of development in societies that's the only way. And right now it's probably THE ONLY way that will improve our lives in literally every field.
You should learn machine learning even if you don't intend to be a machine learning engineer. ML is a tool, an important one that can help make sense from data. In the future, knowing how to use ML as a tool will be equivalent to knowing how to use Excel today. It doesn't mean that everyone needs to know how to build a spreadsheet application but one would be highly handicapped if he or she doesn't know how to use one.
Nah bruh future is much more of easier life.... The past builds the future but yeah machine learning is good to learn(haven't learned it yet) but I think her point is just to focus on what truly matters to you and makes sense because learning is not just in seconds in minutes,days.... You know you are not a genius and your brain would just be prioritize the ones that seems useful to you/interests you..... But if people have excess time then yeah sure they spend that on something they enjoy do whatever they do
If someone generates data without purpose like verification of system designs, then better do not use ML to find "meaning" in data. To find meaning in data afterwards is the most ridicilous thing I have ever heard about data.
The first and foremost thing to keep in mind before doing anything.. whether its learning or base jumping whatever that is .you should never ever listen to anybody ...do if you want to do something...follow your passion and path...
Alot of people jump into cool souding stuff for the wrong reasons all the time, I think people shold have a clear understanding of what they're getting into so they don't give up midway and waste alot of time and money.
"Never listen to anybody" Sorry, but, Thats st*pid. I want to learn machine learning to do image regcognition on my project. I wont go into detail. Without listening to anyone, I wouldn't be able to know what programming code to use with. I was not sure whether i should java or python but thank to the words from many experts, i pick python, and yep, i made the right choice. The point is, listen to people while using ur brain. If u have a real plan, and know what u are going to use it for, then do it. If u only interest because of the hype, then dont. Not listening to anybody for idea is stu*id. Even one of Elon Musk tip is about asking people for idea.
actually, no. That's the complete opposite way of thinking, and it's just as toxic as the message this video wants to deliver. Never ever listen to anybody = taking a lot of risks, in an arrogant and definitely ignorant manner. Ignore me if you want to die happily while 'following your passion and path'.
@@gtv3582 you know the old-timey hollywood clapper board? black board with white stripes. in silent era it was just a board with scene number for editors to keep track when they're cutting film, when sound film was invented they added the clapper for a/v sync
ML reminds me Big Data's hype few years ago, some senior software engineers were scared of being overtaken by junior if they did not learn it, so they wanted to "train" themselves puting an Hadoop stack in every single projects they were working on, until they eventually understood it was not the holy grail of data processing. It just allows to solve some problems (and not all) in a more efficient way than other methods. I think people in tech industry should all learn what ML is, what it can do, hardware requierments and what gains it can provide, so they can know if they really need it. Then, maybe, they should learn it in depth, but first search if there is an existing product that meet their requierements. I learnt how to make, train and use neural networks 20 years ago as part of my academical training. Did I make NN since then ? No. Did I have used libs or services relying on NN ? Yes, quite often. That's quite a different thing.
It's annoying seeing these type of "Why you should NOT" videos. When its actually a big successful job, and she knows it. More competitors in Machine Learning might show she isn't as good as she thinks.
It is crucial to learn machine learning and basic programming. Otherwise you won’t understand what is going on in the world. You don’t need to become an AI researcher but you need to at least understand where to use it, how to train it, and how to assess its predictions. It will concern every businessmen sooner or later by becoming an indispensable too
Sorry, but I disagree. Our entire global society is going to have a hyper-competative market resulting in high levels of competition, especially once web3 becomes the norm. Sure, learn how to navigate the world which includes the technologies that are contained within it's infrastructure... but that's going to be done mostly through products and services. Just like it is today. Hardly anyone is building their own things, be it physical or digital and software/applications. And if they need to literally build something in the future, they will purchase a 3D printer, purchase a blueprint (or design one themselves if they are inclined to learn) and then print it. This can all be done without the individual needing to know anything significant about the technology that surrounds them other than how to interact with it/use it. ML-containing products and services in the future will need a normal part of life. Humans will simply navigate the changing times as well always have.
@@kestonsmith1354 Ok bro, thanks 4 the advice. btw, how did you use machine learning skills in ur project? do you create smthing like a data analytics or prediction system?
Yes indeed. Gradient Decent is sooo cool 😁. I my life I been intressted about how technology works. Electronics when I was a small boy, programming C64 at the 80 back in the days. Working all my life with controllers and embedded system devop.. Starting get some depression about the fact that we get a flood of cheep hardwares sensors and data... how could we ever get use of all this when we don’t have any tools to get use of all this in automations.. Have heard about Neutral Networks but no clue. And then around 2015 I start come in contact with SGD and fell in some tech ”love” with that things. SGD, Convolution, ResNet and Reinforcement learning are sooo cooool things. Backpropagation is Almost like magic
Bad advice in the video. It was just in my recommendation as I used to look at ML videos now and then. There is no argument that ML is hyped. But the hype is for a reason. The reason is that as the compute power increases ML can do things we haven't even imagined. Also, not all ML jobs require PHDs, one can make a way up the ladder from being an engineer to a researcher. If you are passionate about maths and how ML can be the fourth industrial revolution and want to be part of it, jump in, work your a** off and you will be able to get a job no matter what. Remember just be persistent and don't quit.
But she has a point, if you start learning something that you don't like, just for the money, your career path will be ruined. But if you liked it, you're interested on it, then go ahead!
Very true I agree with you. AI, ML is just a fancy term to lure people. Ultimately everything has to be programmed to make certain decision and execution.
@@techconnect1775 we need someone to push programming further to the point where programming could become highly intelligent and capable of self diagnosing and giving the best alternative to execute in action. It may not be possible now but we need people to think this way otherwise we lose hope for brighter more clean future ahead.
Follow your passion folks, if you have started your journey then you already know all this stuff, you start your journey by keeping your destination in your mind and enjoy all the ups and downs experiences that add value to your journey once you complete it.
It’s not harmful to learn any tech, especially new tech or advancing tech for that matter. I don’t see why people shouldn’t learn ML irrespective of the points you’re trying to make. Learning something new, which is new itself shouldn’t be discouraged. I don’t agree with you.
Well, as someone who likes research in AI, I give major credit to Software Engineer. The math is pretty intense to always remember so Software Engineers are very good at making black boxes that are reusable. So I think everyone in technology is just as important.
Nah lol. Software engineers can do machine learning as well as data scientists do (if not better, as we understand the underlaying components, not only specific to the AI but generally regarding the system). Data scientists are essentially upgraded secretaries that believe that they are scientists.
The math isn't that hard for me, and I am in high-school. I actually learned some multivariable calculus for a project that I did, and it honestly was pretty straightforward. 3blue1brown's videos on neural networks really helped me grap the topic. Now of course, this is not all of machine learning so I don't know if there is more math to it, but please let me know! I find it fascinating!
You missed the best part, there's already API's that do everything. All that's needed is a background to explain the parameters used, favorable complexity, and the accuracy of the results.
Here we are 2 years later and this video and all of these comments are already obsolete....buckle up my friends....everything is gonna change and its gonna change real fast
The second point is very important. I started learning machine learning _because_ I had a plan that involved it, but, at one point, I gave up on that plan and went several months without doing _anything_ related to machine learning. I realized that it was not machine learning itself that I was passionate about, but rather, my plan that involved it. A few days ago, I returned to my old plan, and now, I'm going back into machine learning full-force. Additionally, since I dropped out of high school, I am self-taught, and, from the get-go, I knew that I needed to specialize in Natural Language Processing to achieve my goal. So, I watched loads of RUclips videos on that particular subject. At one point, I picked up a beginner's guide to machine learning, but I couldn't get past the 4th chapter because it was like putting one point into each of 5 skillsets rather than 5 points into one skillset, if that makes sense. (I already knew everything that the book taught about NLP and the rest was irrelevant to me.) To this day, there's still this nagging feeling in the back of my head that I can't resell the book that I barely used because I doodled all over one of the pages and trying to erase the doodles would probably just crinkle the paper and leave marks regardless. Anyways, good video! "Don't do it for the money" is a good moral for pretty much any career path! I have an extension enabled to see RUclips dislikes, which I usually use to avoid poorly-made tech tutorials (as stated earlier, I'm self-taught off of RUclips), so I was a bit worried at first, but, after watching this video, I have no idea why it got so many dislikes! All of the points made were great!
Hey Man. It's almost what I am suffering right now. I have been into the creative process of learning and making the websites but sometimes I feel I have a gist for learning ML algorithms. Do you mind telling your plan?
I've started my current job thinking I would get into machine learning to solve a problem. I studied the problem. Machine Learning looked like the solution, but it wasn't. STUDYING THE PROBLEM is the solution.
True. Sometimes you feel that ML can be put in there, only to realize that the problem could be solved with another AI technique such as expert systems or even with no AI at all.
Actually true. There are so many ml engineers now in the market that its very difficult to get a job, or even a call from a potential recruiter! This (lucrative) bubble has bursted.
I wish people would understand her better. Remember, all the people who got disappointed, whose ML-plans failed (I have seen many bachelor and master theses of that sort) will not brag about that. ML often fails, but no one writes s blog article about that. Also, many optimistic evaluations can be found out there, with people not understanding that their claimed ML success is actually a failure.
i took an introductory crash course and some of the simple projects that we did were so cool that i want to learn more about ML and make more exciting projects. thanks for guidance :)
It depends on the kind of work you do. For example, if you are a developer working in machine learning you should have probably already learned about it. I mean how are you supposed to work in machine learning if you don’t learn about it? If your job is to stock shelves (by hand) learning machine learning probably isn’t going to help you at that job.
Tldr: 1. You are just in because of the hype and the hype will die down and the field will likely turn into something you wont enjoy 2. You dont have a plan therefore you will be more likely to give up. 3. You believe it will give you a high paying job. That may be true but it also has a high barrier of entry. You might want to invest your time elsewhere 4. You believe it will be easier to get a job. Spoiler alert: its not.
Everybody throws the word "AI, Machine Learning, etc" as if it's nothing. Your basically just doing probabilities the entire time and the first time I studied it, just by looking at the math equations I kept banging my head against the desk and stared at it as if it was an alien language.
Thank you for the info. However I think this advice applies to just about every career. Especially in the fast changing world we are in even accounting has to be looked at.
Thank you. This video did help me focus! I'm taking ML courses now. I feel like a "natural" at this. I'm an unemployed Electrical Engineer trying to get back in my career. I want to work more with information, rather than just transporting it.
Me too! an unemployed Information Engineer trying to get back in my career after got broke from being a trader. But now ML may not suit me due to my situation, I need to get a job asap.
Wow. This made my day. I am also Electrical engineer (currently unemployed). Joining Masters in AI next month. Looking forward to take up the challenge ☺️
The biggest takeaway from this video is just to learn, "how to create a quick attention of public to watch a video?". Because Negative catches everyone.
I've a very specific problem I'm working towards solving. I quit school, and it's been 3 years of treading in deep-end. I know now how to set up an environment and get an agent to learn. I also know nothing about statistics or ML. If I do succeed, I'll still have no degree or formal education but what concerns me daily less and less is what if I don't succeed. We'll see. Ty for the videos
I'm simply just curious and I'm starting to learn more about it. Who knows, I'm 38 years old and I might just go back for a master's degree focused on machine learning.
I study Information and I can tell you that programming aka machine learning is not for everyone.. learning Information technology made me hate computers. So don’t unless you really have the passion and devotion and a background information about it.
The #1 reason you should NOT learn machine learning is because you won't get rich knowing ML. It is a tool that should be in your programming toolkit. Learn how to sell and make something that provides value instead.
People should do ML if they truly love complex, hard problems that can be time consuming and sometimes unresolved...and that is not a lot of people. But it is worth learning about it for fun. Most interesting field in CS imo. Making a career out of it is another conversation though.
I m going for masters but i m into dilemma What should I choose ??? MSc data science Or MSc AI and ML note - please don't say , choose your passion or intrest
I just saw this first video saying no need to learn this language Even Php is outdated by some people On other hand still people say that php is not dead yet and ML is the hot topic but this lady saying you shouldn’t learn 😂😂😂😂😂 Epic one
You should not start learning machine learning, without a plan that should be based on basic knowledge of machine learning. But "I'm a man with a plan, the plan is to push as much as i can"
I was curious about this topic because I am doing my first internship right now where I am doing python programming to do economic analysis! I think it's so cool, and it makes me so interested in learning more programming languages. I know I am going into data when I graduate, so I am looking at other paths in data. I am so glad I was born in this era. So exicted!
It's impossible to learn everything. Like she said "you need a plan" so by the time you are capable of specializing in ML you should be looking at a specific area of focus.
"don't get on the hype" while i kinda agree with this, you need to remember that the whole tech industry is about hyping new APIs frameworks and fields, when companies like Tesla and Nvidia pour millions into machine learning it simply means there's a huge potential for it beyond the initial hype
Typical click bait 😅... I only agree with the first reason, all other reasons aren't why you shouldn't learn ML, but why you should find the right way to learn, which often requires doubling down on efficient effort.
I think this video was made for mostly clicks. A typical go against the grain title to get people to click on it. I’m not sure “hype” is the correct word for a ground breaking field where people are exploring the possibilities of modeling the human brain digitally.
I absolutely agree with you. From my experience toward my research, machine learning is only useful if there is data available to extract features. However, if you are extracting features from unknown observation then model-based approach is very useful. These days a hybrid between model-based approach and machine learning techniques are useful
FOMO IS REAL IM ALL IN! Not gonna be sorry if the bubble pops. Let me explain: You learn a great amount of mathematics. And this can be used not only in machine learning, but in all CS related subjects. You will learn a programming language. No escape from that. You will learn to think to solve problems not only use products to solve your problems. Its amazing what it can do. I believe we are just in the begininng. Think about what is going to happen when you can train a complex model in 5 minutes.
Having been in ML and Data Science for 8 years, I agree with these points. ML is really hard, unless you're doing no research, and the pay is only high if you can sell yourself well and work long hours and think a lot about how to create business value. Sell yourself well: this is really important and is everyday, you need to communicate really, really clearly and simply with management and teams doing the work before and after you about what is needed and what is possible. Before: the data collection experts. After: might well have Software Engineers and Developers deploying and managing some implementation of your algorithms.
Correct way to get into machine learning: do a PhD in it. The North American way: send around your CV, suck up to hiring managers, complain, do bootcamps, anything that's a fast, quick and easy way
I would recommend renaming this video to "Should you really learn machine learning?" or "You might not want to learn machine learning." These aren't reasons we shouldn't learn it; these are examples of incorrect reasoning for those who DO want to learn.
Like an alternative AI verison of Joe rogan is who is much better than rogan in everyday and everyone just listens to the AI version instead of the real guy?
I have been for 15 years in Machine Learning field. I have seen lots of young data scientists and machine learning engineers with their masters' degrees but none of them actually can do real machine learning jobs at all. They all have heard about machine learning algorithms and all they did was all about introduction to machine learning. The problem is that with their degrees, they believe they know what it is and they pretent they know what they are doing and it makes me sick and really sick.
I really appreciate the fact that you are trying to help people understand that a future in AI/ML is not for everyone. However, many of your defacto style assertions are simply opinions. For example, about solving the world's problems (and creating new ones); although we are in what is somewhat analogous to the late '70's when I got into traditional computer science (i.e. AI is fairly primitive still-- especially where it will be in just a few years at its current growth rate of about 4-times Moore's Law), we are already seeing explosive advances across many fields, from robotics, finance and the environment to what we are starting to see like alternatives via AI to Quantum-genetic algorithms and canonical form expressions via AI for things areas like infection prediction, cancer diagnosis, protein folding and many other areas. So, not sure what you mean that it will not change the world (and in a MAJOR way). If I could tell you what we are doing with .mil budget development, you would think what we created was decades away, while actually it is currently in military testing. Also, compared to traditional computer science in its current state, it is in fact simple. While not trivial, the state of the art commercially is only about 6 years old in most respects, while with traditional computer science, Donald E. Knuth 1, 2 and 3 were just the bare beginning of study. Lastly, you are wrong about starting salaries and jobs. Like many , we have to pay a major premium to a good candidate (no BS in CS necessarily required, let alone an MS or Doctorate) compared to any other type of computer position. Plus benefits, plus signing bonuses, plus moving expenses, plus a starting salary of about 30% higher than other CS areas.
Question: If you are a gap year student going into college who is planning to study Computer Science, what is the best thing to study after taking Intro to programing: CS50? //What would be the best thing to learn to get an summer internship before your first year of university? //Would Intro to AI: CS50 or Intro to Web Development: CS50 be a better option? //What courses would you take to prepare Intro to CS in Java? Where would you learn math? // Can you get a job as a ML engineer with a Bachelors of Science in CS with a specialization in ML?
theme: Focus on your why If in doubt, follow 5 why's approach reflectively humbly examine your reasons goals I think it can start with curiosity Just know it's a journey Overnight success is not reality For everyone It's okay Be a learning champion Of your own journey Start small and simple Build and adapt Been in industry 25 years Nothing is linear Enjoy the ups and downs Keep learning Driven by focused goals Update them regularly Enjoy the journey So you can last Most of all Take care of yourself Sleep Excercise Diet Mental health habits
"Why You Should NOT..." it is little bit sensationalist, but you touch the point, you don't need machine learn to a problem that have small variables to deal with.
This video mean to me like, you want to go viral by doing anything. No matter what it is. I came with a hype and immediately it went away(like you said) when I saw the timestamps. Sorry if I'm wrong
Solid Algebra like advanced levels and intermediate calculus, some experience with Matrix algebra and calculus. These are just prerequisite for understanding basics and getting feel of it, after that once you know basics you can learn advance libraries that are available in python or any other languages to solve whatever you want, there are some even higher level abstractions like Google's tensor flow for example that may look alien but if you know basics it's like writing 10 lines of code instead of 300.
Smitha , if you are in Canada you are in the wrong country. The barrier for anything tech there is bad because of lack of economic saturation . People are too few and far apart and the Tech economy runs pretty much to service the US .
I think the downvotes come from people who didn't watch til the end. She's saying like anything do it for the right reasons, and prepare for it not to be easy
People on internet be like : don't be a software engineer don't learn ML don't learn Ai 10 reasons why you should not learn math Everything in this world is hard if you are sucks at playing flute no band wants you to play flute in their band and if you love playing flute then you would be great in it and everyone wants you SIMPLE!
A very good point was raised, and I have heard it said this way. A learning programmer who already has a problem to solve will progress faster. Find the area of ML that excites you. Voice and image recognition research is amazing!
I'm studying courses regarding Machine learning at university at the moment, taking kaggle courses and reading some deep learning books on the side. This video actually gives me MORE confidence because I did not recognise myself in any of the pit-falls. 1. I got interested in ML after taking a AI course, not because of the hype. I didn't really consider it before actually trying it. 2. I have a idea and niche I'd want to work with. Combining ML with Automation engineering which I have been interning with for the last few years.
Such poor reasons not to begin a field. Either you want to expand your mind, learn a relevant skill, or you don't. I'd never base learning something on FOMO. Ridiculous.
Wow, massive learnings, thanks! I wonder if you wouldn’t mind discussing the kind of limitations a hobbyist would face? What kind of projects are possible?
Hello everyone! I'm creating a brand new channel dedicated to in-depth guides and courses in AI & ML 😎! Subscribe to my brand new channel 'School of Machine Learning': www.youtube.com/@SchoolofMachineLearning
because you dont want competition? nice try
Hahaha
Yes, she dont want other competition. Lol
Lol
True
Haha, same thought exactly ;)
"When the hype will die down.."? Are you kidding me? It's only gonna go higher and higher. With current direction of development in societies that's the only way. And right now it's probably THE ONLY way that will improve our lives in literally every field.
You should learn machine learning even if you don't intend to be a machine learning engineer. ML is a tool, an important one that can help make sense from data. In the future, knowing how to use ML as a tool will be equivalent to knowing how to use Excel today. It doesn't mean that everyone needs to know how to build a spreadsheet application but one would be highly handicapped if he or she doesn't know how to use one.
Valid point !! I also think it's going to a common tool in various fields of study. So learning it will always help.
Nah bruh future is much more of easier life.... The past builds the future but yeah machine learning is good to learn(haven't learned it yet) but I think her point is just to focus on what truly matters to you and makes sense because learning is not just in seconds in minutes,days.... You know you are not a genius and your brain would just be prioritize the ones that seems useful to you/interests you..... But if people have excess time then yeah sure they spend that on something they enjoy do whatever they do
If someone generates data without purpose like verification of system designs, then better do not use ML to find "meaning" in data. To find meaning in data afterwards is the most ridicilous thing I have ever heard about data.
Why not just use tableau? Secret: it has ml models built into it... Don't tell business people...
Or we should learn ML, because we may have to fight them for the survival of humanity. Better an army of Neo's than One.
The first and foremost thing to keep in mind before doing anything.. whether its learning or base jumping whatever that is .you should never ever listen to anybody ...do if you want to do something...follow your passion and path...
Alot of people jump into cool souding stuff for the wrong reasons all the time, I think people shold have a clear understanding of what they're getting into so they don't give up midway and waste alot of time and money.
"Never listen to anybody"
Sorry, but,
Thats st*pid.
I want to learn machine learning to do image regcognition on my project. I wont go into detail. Without listening to anyone, I wouldn't be able to know what programming code to use with. I was not sure whether i should java or python but thank to the words from many experts, i pick python, and yep, i made the right choice.
The point is, listen to people while using ur brain. If u have a real plan, and know what u are going to use it for, then do it. If u only interest because of the hype, then dont.
Not listening to anybody for idea is stu*id. Even one of Elon Musk tip is about asking people for idea.
actually, no. That's the complete opposite way of thinking, and it's just as toxic as the message this video wants to deliver. Never ever listen to anybody = taking a lot of risks, in an arrogant and definitely ignorant manner. Ignore me if you want to die happily while 'following your passion and path'.
Thank you, now I finally have the confidence to fulfill my dream of being a mass murderer! I won't listen to those pesky "lawmakers" anymore
Agreed! People can be background noise
Tip to align the voice track: just clap your hands in the beginning of the clip, this way its easy to align the audio (peak) with the video track.
bro idk why but i was struggling to figure out a way to do it, and on some random video i found the solution lol. thx
@@gtv3582 that’s awesome :)
Use machine learning.
@@gtv3582 you know the old-timey hollywood clapper board? black board with white stripes. in silent era it was just a board with scene number for editors to keep track when they're cutting film, when sound film was invented they added the clapper for a/v sync
Her machine learning didn't do much for it lol. But srsly, ML can be the future.
ML reminds me Big Data's hype few years ago, some senior software engineers were scared of being overtaken by junior if they did not learn it, so they wanted to "train" themselves puting an Hadoop stack in every single projects they were working on, until they eventually understood it was not the holy grail of data processing. It just allows to solve some problems (and not all) in a more efficient way than other methods.
I think people in tech industry should all learn what ML is, what it can do, hardware requierments and what gains it can provide, so they can know if they really need it. Then, maybe, they should learn it in depth, but first search if there is an existing product that meet their requierements.
I learnt how to make, train and use neural networks 20 years ago as part of my academical training. Did I make NN since then ? No. Did I have used libs or services relying on NN ? Yes, quite often. That's quite a different thing.
Will I need to learn AI or machine learning if I want to build a software that can generate animated movie from an input image?
Same can be said about OOP!
It's annoying seeing these type of "Why you should NOT" videos. When its actually a big successful job, and she knows it. More competitors in Machine Learning might show she isn't as good as she thinks.
It is crucial to learn machine learning and basic programming. Otherwise you won’t understand what is going on in the world. You don’t need to become an AI researcher but you need to at least understand where to use it, how to train it, and how to assess its predictions. It will concern every businessmen sooner or later by becoming an indispensable too
True
Sorry, but I disagree. Our entire global society is going to have a hyper-competative market resulting in high levels of competition, especially once web3 becomes the norm. Sure, learn how to navigate the world which includes the technologies that are contained within it's infrastructure... but that's going to be done mostly through products and services. Just like it is today. Hardly anyone is building their own things, be it physical or digital and software/applications. And if they need to literally build something in the future, they will purchase a 3D printer, purchase a blueprint (or design one themselves if they are inclined to learn) and then print it. This can all be done without the individual needing to know anything significant about the technology that surrounds them other than how to interact with it/use it.
ML-containing products and services in the future will need a normal part of life. Humans will simply navigate the changing times as well always have.
You want to learn something, don't just watch RUclips videos alone and do small examples, start projects on your own. It helped me with programming
i love testimonies, hypes me up to practice and believes it. many people tell you what to do
Hey, do you start projects and just keep them with you when you finished it, or use it as a part of your portfolio?
@@leonhardsatriasuharjo1644 Some of them I use as part of my portfolio. You want to show some of your work but not everything.
@@kestonsmith1354 Ok bro, thanks 4 the advice. btw, how did you use machine learning skills in ur project? do you create smthing like a data analytics or prediction system?
Machine Learning is a art. Everyone will enjoy when understood. Learn ML for fun if not for career.
Yes indeed. Gradient Decent is sooo cool 😁. I my life I been intressted about how technology works. Electronics when I was a small boy, programming C64 at the 80 back in the days. Working all my life with controllers and embedded system devop.. Starting get some depression about the fact that we get a flood of cheep hardwares sensors and data... how could we ever get use of all this when we don’t have any tools to get use of all this in automations.. Have heard about Neutral Networks but no clue. And then around 2015 I start come in contact with SGD and fell in some tech ”love” with that things. SGD, Convolution, ResNet and Reinforcement learning are sooo cooool things. Backpropagation is Almost like magic
Strongly agreed.
THIS ONE OF THE FEW REAL TALKS ABOUT ML.....SIMPLY PUT: "IF YOU DO NOT HAVE A GOAL IN MIND FOR LEARNING ML DON'T LEARN IT"
Bad advice in the video. It was just in my recommendation as I used to look at ML videos now and then. There is no argument that ML is hyped. But the hype is for a reason. The reason is that as the compute power increases ML can do things we haven't even imagined. Also, not all ML jobs require PHDs, one can make a way up the ladder from being an engineer to a researcher. If you are passionate about maths and how ML can be the fourth industrial revolution and want to be part of it, jump in, work your a** off and you will be able to get a job no matter what. Remember just be persistent and don't quit.
I have no idea what machine learning is, but i will start learning to get a high paid job. U won't change my mind
But she has a point, if you start learning something that you don't like, just for the money, your career path will be ruined. But if you liked it, you're interested on it, then go ahead!
Even worse is when they use AI just to sell a product. I actually didn't know you had a YT channel. These are very good and valid points.
Samzee Codes thanks Sam! Yes I felt like these needed to be addressed!
Even worse is when they think "if else" is AI in a product.
@@SmithaKolan Did you work in DLSS (Deep Learning Super Sampling)???
Very true I agree with you. AI, ML is just a fancy term to lure people. Ultimately everything has to be programmed to make certain decision and execution.
@@techconnect1775 we need someone to push programming further to the point where programming could become highly intelligent and capable of self diagnosing and giving the best alternative to execute in action. It may not be possible now but we need people to think this way otherwise we lose hope for brighter more clean future ahead.
I thank you so much for giving a summary in the description.
I did not watch the video and now comment to compensate.
Follow your passion folks, if you have started your journey then you already know all this stuff, you start your journey by keeping your destination in your mind and enjoy all the ups and downs experiences that add value to your journey once you complete it.
Well said!
It’s not harmful to learn any tech, especially new tech or advancing tech for that matter. I don’t see why people shouldn’t learn ML irrespective of the points you’re trying to make. Learning something new, which is new itself shouldn’t be discouraged. I don’t agree with you.
Well, as someone who likes research in AI, I give major credit to Software Engineer. The math is pretty intense to always remember so Software Engineers are very good at making black boxes that are reusable. So I think everyone in technology is just as important.
Nah lol. Software engineers can do machine learning as well as data scientists do (if not better, as we understand the underlaying components, not only specific to the AI but generally regarding the system). Data scientists are essentially upgraded secretaries that believe that they are scientists.
The math isn't that hard for me, and I am in high-school. I actually learned some multivariable calculus for a project that I did, and it honestly was pretty straightforward. 3blue1brown's videos on neural networks really helped me grap the topic. Now of course, this is not all of machine learning so I don't know if there is more math to it, but please let me know! I find it fascinating!
@@EliteCubingAlliance there is more math. You do differential equations, statistics, linear algebra and some physics
Understanding Machine Learning is different from taking Machine Learning Career.
You missed the best part, there's already API's that do everything. All that's needed is a background to explain the parameters used, favorable complexity, and the accuracy of the results.
Then what? Data Engineering? MLOps?
Can you give some examples of the APIs?
@@dingding4898 keras does literally everything for you haha
@@mynameisjapesI assumed he was talking about pretrained models by cloud services. My bad!
@@dingding4898 keras is indeed an API
@@MARTIN-101 Yeah, my bad. It has become quite expansive, such that it feels like a framework.
Here we are 2 years later and this video and all of these comments are already obsolete....buckle up my friends....everything is gonna change and its gonna change real fast
The second point is very important. I started learning machine learning _because_ I had a plan that involved it, but, at one point, I gave up on that plan and went several months without doing _anything_ related to machine learning. I realized that it was not machine learning itself that I was passionate about, but rather, my plan that involved it. A few days ago, I returned to my old plan, and now, I'm going back into machine learning full-force.
Additionally, since I dropped out of high school, I am self-taught, and, from the get-go, I knew that I needed to specialize in Natural Language Processing to achieve my goal. So, I watched loads of RUclips videos on that particular subject. At one point, I picked up a beginner's guide to machine learning, but I couldn't get past the 4th chapter because it was like putting one point into each of 5 skillsets rather than 5 points into one skillset, if that makes sense. (I already knew everything that the book taught about NLP and the rest was irrelevant to me.) To this day, there's still this nagging feeling in the back of my head that I can't resell the book that I barely used because I doodled all over one of the pages and trying to erase the doodles would probably just crinkle the paper and leave marks regardless.
Anyways, good video! "Don't do it for the money" is a good moral for pretty much any career path! I have an extension enabled to see RUclips dislikes, which I usually use to avoid poorly-made tech tutorials (as stated earlier, I'm self-taught off of RUclips), so I was a bit worried at first, but, after watching this video, I have no idea why it got so many dislikes! All of the points made were great!
I only love computer vision , surveillance , to track humans under social credit points ,
Hey Rowan404, what was your plan that involved ML? Let’s hear it 😮
Hey Man. It's almost what I am suffering right now. I have been into the creative process of learning and making the websites but sometimes I feel I have a gist for learning ML algorithms. Do you mind telling your plan?
By having a plan. It's important to know where and what data is going to come from. Without data, there's no need to learn to analyse data.
I've started my current job thinking I would get into machine learning to solve a problem. I studied the problem. Machine Learning looked like the solution, but it wasn't. STUDYING THE PROBLEM is the solution.
What was the problem?
True.
Sometimes you feel that ML can be put in there, only to realize that the problem could be solved with another AI technique such as expert systems or even with no AI at all.
Actually true.
There are so many ml engineers now in the market that its very difficult to get a job, or even a call from a potential recruiter! This (lucrative) bubble has bursted.
imo this is the most motivational video to learn ml for me
I wish people would understand her better. Remember, all the people who got disappointed, whose ML-plans failed (I have seen many bachelor and master theses of that sort) will not brag about that. ML often fails, but no one writes s blog article about that. Also, many optimistic evaluations can be found out there, with people not understanding that their claimed ML success is actually a failure.
i took an introductory crash course and some of the simple projects that we did were so cool that i want to learn more about ML and make more exciting projects. thanks for guidance :)
where did you take this course?
@@deesee3622 I took it as a part of internship for my college although google offers a similar introductory course which is totally free
@@azmatkhan7195 can u please share the link
How’s it going two years later?
@@lenderzconstable pretty good i would say, currently doing GSOC'23
It depends on the kind of work you do. For example, if you are a developer working in machine learning you should have probably already learned about it. I mean how are you supposed to work in machine learning if you don’t learn about it? If your job is to stock shelves (by hand) learning machine learning probably isn’t going to help you at that job.
Tldr:
1. You are just in because of the hype and the hype will die down and the field will likely turn into something you wont enjoy
2. You dont have a plan therefore you will be more likely to give up.
3. You believe it will give you a high paying job. That may be true but it also has a high barrier of entry. You might want to invest your time elsewhere
4. You believe it will be easier to get a job. Spoiler alert: its not.
Everybody throws the word "AI, Machine Learning, etc" as if it's nothing. Your basically just doing probabilities the entire time and the first time I studied it, just by looking at the math equations I kept banging my head against the desk and stared at it as if it was an alien language.
Thank you for the info.
However I think this advice applies to just about every career.
Especially in the fast changing world we are in even accounting has to be looked at.
Everyone should know ML since its a tool for the future, not to become a specialist. What a stupid video. It was only made for the clickbaits.
I was searching "should i learn AI or machine learning in 2023?" And this is the first result came up
Thank you. This video did help me focus! I'm taking ML courses now.
I feel like a "natural" at this.
I'm an unemployed Electrical Engineer trying to get back in my career. I want to work more with information, rather than just transporting it.
Me too! an unemployed Information Engineer trying to get back in my career after got broke from being a trader. But now ML may not suit me due to my situation, I need to get a job asap.
@@thaenalpha umm try web development
Wow. This made my day. I am also Electrical engineer (currently unemployed). Joining Masters in AI next month. Looking forward to take up the challenge ☺️
@@beluga.314 good luck to you!
The biggest takeaway from this video is just to learn, "how to create a quick attention of public to watch a video?". Because Negative catches everyone.
I've a very specific problem I'm working towards solving. I quit school, and it's been 3 years of treading in deep-end. I know now how to set up an environment and get an agent to learn. I also know nothing about statistics or ML. If I do succeed, I'll still have no degree or formal education but what concerns me daily less and less is what if I don't succeed. We'll see. Ty for the videos
I'm simply just curious and I'm starting to learn more about it. Who knows, I'm 38 years old and I might just go back for a master's degree focused on machine learning.
That interior design is super cool and very relaxing. Well done
I think it's a virtual background
Thankyou very much your
video was vey informative 💛
I study Information and I can tell you that programming aka machine learning is not for everyone.. learning Information technology made me hate computers. So don’t unless you really have the passion and devotion and a background information about it.
The #1 reason you should NOT learn machine learning is because you won't get rich knowing ML. It is a tool that should be in your programming toolkit. Learn how to sell and make something that provides value instead.
The young lady is correct. You need to have a plan and love to code.
She is right in most aspects.
If it is just for high pay, there are many other fields, but ML is the cool field which also pays well.
What other fields have the best pay?
People should do ML if they truly love complex, hard problems that can be time consuming and sometimes unresolved...and that is not a lot of people. But it is worth learning about it for fun. Most interesting field in CS imo. Making a career out of it is another conversation though.
rather go to web development route anyday
I m going for masters but i m into dilemma
What should I choose ???
MSc data science
Or
MSc AI and ML
note - please don't say , choose your passion or intrest
I just saw this first video saying no need to learn this language
Even Php is outdated by some people
On other hand still people say that php is not dead yet and ML is the hot topic but this lady saying you shouldn’t learn 😂😂😂😂😂
Epic one
Domain knowledge is the key to application. You can't get that in school.
Could you make a video on Why you SHOULD learn Machine Learning?
she cant probably I feel like she might quit it at any time lol
@@bombrman1994 why do you think that's the case? lol
@@user77434 men does not give up and is not a men.
Jk
This kind of was why to do it. More like a tip sheet.
lmao the vid is going to be 2 min
You should not start learning machine learning, without a plan that should be based on basic knowledge of machine learning.
But "I'm a man with a plan, the plan is to push as much as i can"
I was curious about this topic because I am doing my first internship right now where I am doing python programming to do economic analysis! I think it's so cool, and it makes me so interested in learning more programming languages. I know I am going into data when I graduate, so I am looking at other paths in data. I am so glad I was born in this era. So exicted!
Everyone should know ML since its a tool for the future, not to become a specialist. What a stupid video. It was only made for the clickbaits.
her smile is so sincere, she convinced me of something i don't even know 🙂
You should learn everything in this era if you can!!!
It's impossible to learn everything. Like she said "you need a plan" so by the time you are capable of specializing in ML you should be looking at a specific area of focus.
@@photonboy999 Agreed
"don't get on the hype"
while i kinda agree with this, you need to remember that the whole tech industry is about hyping new APIs frameworks and fields, when companies like Tesla and Nvidia pour millions into machine learning it simply means there's a huge potential for it beyond the initial hype
Thanks for this. I was pretty confused about this, and you hit the nail on the head. I'll take these points into consideration
Typical click bait 😅... I only agree with the first reason, all other reasons aren't why you shouldn't learn ML, but why you should find the right way to learn, which often requires doubling down on efficient effort.
I understand, but what do you mean by having a plan? I mean what's your definition of having a plan in ML. And how do I make one?
I think this video was made for mostly clicks. A typical go against the grain title to get people to click on it. I’m not sure “hype” is the correct word for a ground breaking field where people are exploring the possibilities of modeling the human brain digitally.
I absolutely agree with you. From my experience toward my research, machine learning is only useful if there is data available to extract features. However, if you are extracting features from unknown observation then model-based approach is very useful. These days a hybrid between model-based approach and machine learning techniques are useful
FOMO IS REAL IM ALL IN! Not gonna be sorry if the bubble pops. Let me explain:
You learn a great amount of mathematics. And this can be used not only in machine learning, but in all CS related subjects.
You will learn a programming language. No escape from that.
You will learn to think to solve problems not only use products to solve your problems.
Its amazing what it can do.
I believe we are just in the begininng.
Think about what is going to happen when you can train a complex model in 5 minutes.
Having been in ML and Data Science for 8 years, I agree with these points.
ML is really hard, unless you're doing no research, and the pay is only high if you can sell yourself well and work long hours and think a lot about how to create business value.
Sell yourself well: this is really important and is everyday, you need to communicate really, really clearly and simply with management and teams doing the work before and after you about what is needed and what is possible.
Before: the data collection experts.
After: might well have Software Engineers and Developers deploying and managing some implementation of your algorithms.
Correct way to get into machine learning: do a PhD in it. The North American way: send around your CV, suck up to hiring managers, complain, do bootcamps, anything that's a fast, quick and easy way
I would recommend renaming this video to "Should you really learn machine learning?" or "You might not want to learn machine learning." These aren't reasons we shouldn't learn it; these are examples of incorrect reasoning for those who DO want to learn.
Yea agree, i think title is pure clickbait, there are no reasons why we shouldn't learn something
you don't get views with boring titles ;)
idk i'm learning it anyway, because the concepts are so exciting.
Even as an enthousiast, I get excited daily about new machine learning ideas I haven’t thought about
I know the feeling! Good luck to both of us on our explorations :)
Like an alternative AI verison of Joe rogan is who is much better than rogan in everyday and everyone just listens to the AI version instead of the real guy?
@@sumtingwong8768 omfg so I'm not the only one who thought of this?!?!?!
@@rishavjain1867 its no joke
Even if you try it just because of the hype, it is not a problem to try it out.
I see a lot of high paying jobs that need machine learning.
When we learn machine learning, do we learn how machines learn, or does the machine learn how we learn machine learning?
Learn everything you can.
I have been for 15 years in Machine Learning field. I have seen lots of young data scientists and machine learning engineers with their masters' degrees but none of them actually can do real machine learning jobs at all. They all have heard about machine learning algorithms and all they did was all about introduction to machine learning. The problem is that with their degrees, they believe they know what it is and they pretent they know what they are doing and it makes me sick and really sick.
Pretty good advices. Thank you for uploading this video
O lot of people also confuse pid and mpc controllers with machine learning. Everything is ml
I really appreciate the fact that you are trying to help people understand that a future in AI/ML is not for everyone. However, many of your defacto style assertions are simply opinions.
For example, about solving the world's problems (and creating new ones); although we are in what is somewhat analogous to the late '70's when I got into traditional computer science (i.e. AI is fairly primitive still-- especially where it will be in just a few years at its current growth rate of about 4-times Moore's Law), we are already seeing explosive advances across many fields, from robotics, finance and the environment to what we are starting to see like alternatives via AI to Quantum-genetic algorithms and canonical form expressions via AI for things areas like infection prediction, cancer diagnosis, protein folding and many other areas. So, not sure what you mean that it will not change the world (and in a MAJOR way). If I could tell you what we are doing with .mil budget development, you would think what we created was decades away, while actually it is currently in military testing.
Also, compared to traditional computer science in its current state, it is in fact simple. While not trivial, the state of the art commercially is only about 6 years old in most respects, while with traditional computer science, Donald E. Knuth 1, 2 and 3 were just the bare beginning of study.
Lastly, you are wrong about starting salaries and jobs. Like many , we have to pay a major premium to a good candidate (no BS in CS necessarily required, let alone an MS or Doctorate) compared to any other type of computer position. Plus benefits, plus signing bonuses, plus moving expenses, plus a starting salary of about 30% higher than other CS areas.
took the words right out of mouth
simply because you have an opinion , doesn't mean your word should become gospel
Question: If you are a gap year student going into college who is planning to study Computer Science, what is the best thing to study after taking Intro to programing: CS50?
//What would be the best thing to learn to get an summer internship before your first year of university?
//Would Intro to AI: CS50 or Intro to Web Development: CS50 be a better option?
//What courses would you take to prepare Intro to CS in Java? Where would you learn math?
// Can you get a job as a ML engineer with a Bachelors of Science in CS with a specialization in ML?
How about learning data structures and algorithms before learning machine learnjng
and then simply quitting coding forever
Nobody learns such things anymore. It is of the far past.
@@aabb-zz9uw bruh
@@aabb-zz9uw This is not even close to true.
@@noaho.5304 maybe he's talking about AutoML?
what problem do you solve as a machine learning Engineer
yep, i agree on the second point, having a plan is a necessity.
theme: Focus on your why
If in doubt, follow 5 why's approach
reflectively
humbly
examine your reasons
goals
I think it can start with curiosity
Just know it's a journey
Overnight success is not reality
For everyone
It's okay
Be a learning champion
Of your own journey
Start small and simple
Build and adapt
Been in industry 25 years
Nothing is linear
Enjoy the ups and downs
Keep learning
Driven by focused goals
Update them regularly
Enjoy the journey
So you can last
Most of all
Take care of yourself
Sleep
Excercise
Diet
Mental health habits
"Why You Should NOT..." it is little bit sensationalist, but you touch the point, you don't need machine learn to a problem that have small variables to deal with.
This video mean to me like, you want to go viral by doing anything. No matter what it is. I came with a hype and immediately it went away(like you said) when I saw the timestamps. Sorry if I'm wrong
I want to develop a project by my own. I want to know , What is the level of Mathematical complexity? What I need to know before starting?
Real ML calls for functional anal and measure them. Good luck.
Solid Algebra like advanced levels and intermediate calculus, some experience with Matrix algebra and calculus. These are just prerequisite for understanding basics and getting feel of it, after that once you know basics you can learn advance libraries that are available in python or any other languages to solve whatever you want, there are some even higher level abstractions like Google's tensor flow for example that may look alien but if you know basics it's like writing 10 lines of code instead of 300.
Smitha , if you are in Canada you are in the wrong country. The barrier for anything tech there is bad because of lack of economic saturation . People are too few and far apart and the Tech economy runs pretty much to service the US .
Hi @Smitha Kolan - just some feedback - the audio and video are out of sync by a few seconds :D
btw - great video !
She can use some machine learning to fix that 😂
1980: "internet is as useful as fax machine, it won't have a big impact on the world, don't bother with internet"
I started a toilet cleaning company...we integrated machine learning. 😛
make a poop knife which will detect and cut big chunks with the help of opencv.
Toilet: "you've been drinking too much wine Karen"
Don't worry, toilet cleaner worker will protest about your ML stealing their job.
I think the downvotes come from people who didn't watch til the end. She's saying like anything do it for the right reasons, and prepare for it not to be easy
This video suggestion came from the ML recommendations system of RUclips 😬😀
HhHHHHHHHHH
why do you think their recommendation system is machine learning?
Nah everyone knows RUclips recommendations is controlled by illuminati
@@gokuldinesh8851 Illuminati may own youTube,So Yeah, cant rule this out.
"Machine learning" has existed for at least a century in different forms.
Superb Video, very realistic. There is a one-second delay in the video from the audio (only in some parts)
I think you should change the title to "Why You Should Learn Machine Learning"
If I don’t know anything about a tool how should I leverage to solve my problems?
Totally in sync with my thought
People on internet be like : don't be a software engineer don't learn ML don't learn Ai 10 reasons why you should not learn math
Everything in this world is hard if you are sucks at playing flute no band wants you to play flute in their band and if you love playing flute then you would be great in it and everyone wants you SIMPLE!
Great overview! Something I also came up on personal experience over last 6 years in the industry
Thanks a ton !!!! It is an eye opening video, every success hungry person should watch.
I started as a data analyst on my way to ML. It definitely helps.
Same here🙌
Is there a difference?
You dont know just by putting that thumb nail you are making so many future aspirant anxious.
A very good point was raised, and I have heard it said this way. A learning programmer who already has a problem to solve will progress faster. Find the area of ML that excites you. Voice and image recognition research is amazing!
i use it a lot to for self adapting controllers in closed loops, it's the best!
I'm studying courses regarding Machine learning at university at the moment, taking kaggle courses and reading some deep learning books on the side.
This video actually gives me MORE confidence because I did not recognise myself in any of the pit-falls.
1. I got interested in ML after taking a AI course, not because of the hype. I didn't really consider it before actually trying it.
2. I have a idea and niche I'd want to work with. Combining ML with Automation engineering which I have been interning with for the last few years.
How’s it going a year later?
Such poor reasons not to begin a field. Either you want to expand your mind, learn a relevant skill, or you don't. I'd never base learning something on FOMO. Ridiculous.
Wow, massive learnings, thanks! I wonder if you wouldn’t mind discussing the kind of limitations a hobbyist would face? What kind of projects are possible?