Haha yeah! I feel like it would help if they actually hinted at the cool applications of the boring basics you learn. But yeah, they will be useful in future haha Keep it up!! 🚀
@@borismeinardus how did you know you wanted to research? I love AI but also am a realist and know I'll need a stable career and I'm under the impression the research isn't needed- nor will it be able to make physical changes to the world (ie the differrence between the researcher behind autonomous vehicles and actually the engineers who made the Tesla). I'm only a first year but I have no idea if I should start a career transition (ie start as software engineer/junior internships/govmt roles) or a research transition (get lecturers on my side, read papers etc) - idk if you have a whole video on this lol
Well, you definitely need both! Most of the innovative, new, and cool technology that the engineers deploy into production come from research. Also all the Deep Learning models used in Tesla Autopilot! But without the engineers, the research will never really have an actual impact on the real world. The cool thing in AI specifically, the turnaround time from research to actual deployment is much shorter than in other industries. As an ML researcher, you still need to be a good engineer! When it comes to stable jobs and good salary, if you work at a top company, you will earn a decent amount of money either as a pure SWE/ ML Engineer or Researcher (in fact, perhaps even more as an ML researcher). I personally just love the exploratory nature of research, standing in front of a white board and trying to draw out a new idea, and reading a lot of papers. I feel like, learning (reading papers and taking very up to date lectures) is part of my job as a researcher and is just stuff I have found to really enjoy. I found this passion over time. As you saw in the video, in the very beginning, I wanted to build robots haha - so be more on the hardware side. But as long as you keep exploring new areas, you will at some point figure out what you enjoy more and what less. :) I hope this somewhat helps 😊
@@Uncreeperble Hearing this is still crazy to me haha Just trying to push the community in a good direction and share my experience :) Thank you for the support!!
I finished my Physics degree over 20 years ago, and I’m starting an MSc in AI this year, after almost 2 decades in software engineering. Thanks for telling us about your journey and your work!
Amazing! My fifth semesters feels like last week haha. Time flies by so insanely fast... I am really happy to hear that my simple story can shed some light into what it might take and even inspire the one or the other 😊 Thank you ❤️
I graduated from my Computer Application Degree in 2021. Right after that I started working in IT Support to be financially independent. Now after 2.5 years, I feel very lost and believe that I am not using my full potential anymore. I have started coding in my free time and found my passion for it again. I wish to be a ML Engineer. This video brings motivation. Thank you
Amazing! As long as you find a passion and consistently put in the time, it is pretty much impossible, that at some point you will be really good at it :) Keep it up and all the best!!!
00:01 6-year journey from studying Computer Engineering to working at favorite AI startup 02:35 Gained diverse experience beyond standard computer science courses 04:58 Learning curve plateaued after 8-9 months 07:36 Transition from bachelor to master's focused on machine learning and reinforcement learning. 10:03 Transition to applied scientist student researcher in ML Department 12:40 Steep learning curve in first computer vision project 15:19 Striving for top ML research opportunities after learning from project experiences. 17:46 Struggles with ML internships and shifting focus to research 19:56 Continuing education in Machine Learning with focus on computer vision and deep learning. 22:14 Completed Master's thesis in final year of university, focused on multimodal learning and research. 24:32 Overcame challenges to secure dream AI startup job
Hey Boris, It was so nice hearing about your journey into the world of ML. I am in my third year(5th semester) of engineering and pursuing electronics and communication engineering (somewhat similar to your computer engineering), interested in AI, embedded systems, and Machine learning. I found your story quite relatable...kudos man! love your content !
It is so cool to see the many different paths people follow that can all lead to something similar in the end :) As long as you know what you want and work for it, you'll get it! Thank you for the support and keep it up! ❤️
Hey boris congratulations and all the best, Thanks for video man I'm halfway through in the domain and had lots of doubts and needed a perspective to see things which you gave so nicely and really well presented. Learnt a lot I mean this is what I actually wanted and needed I just want to say imma be so grateful for this tysm
I am really glad to hear that! In the end, we all go through similar struggles, but as long as we push through, I am sure we will all "make it" in one way or the other Keep it up!!! ❤️
Your video inspired me. I shifted careers and I know ML is hot right now. That's why I plan on differentiating myself by applying ML to what I finished in college: Chemistry. I think it's still an uncharted territory with very little literature. So I am excited. And just like you, RL does something to me. There's something interesting and awesome in building agents that can decide on their own. I will probably specialize in this part of machine learning. 😀Hoping for the best!
AI in science in general is an insanely fascinating, exciting, and revolutionary field! (Look at Alpha-Fold) There definitely is a lot you can explore there and research to become an expert in that field! Keep it up ❤️
Man you’re such a down to earth and personable guy, and Its great how you discuss the emotions and failures along the way. It makes things more relatable even though you’re at such a level ahead of the majority of people
Damn Bro, was ist passiert? Das erste mal, als ich deinen channel entdeckt habe hattest du noch 23 Abonnenten?! Und siehe da fast 53.000! Hast du dir verdient, toller content! Beste Grüße fellow TUBler :D
Congrats for the channel Boris. Here a 6y experience Business Analyst who's graduated in Bachelor Economics and he's starting Master in Data science this year cause I want to get more into depth of IT stuff and I'm literally learning many different things from anywhere before even starting the university. As I work 80% remotely and the company I work for accepts well studying workers, I'm investing my "spare time" to build something that could allow me a better future, like working as Business Analyst/Project Manager on ML/AI fields. I already developed some skills in data managing using SQL and attending several online courses such as Data Analytics and IBM Data engineering (Coursera) thus I'm very eager to start. Even if I got Python basics and tried some ML scripts, I don't feel much proficient to start by my own without external help but I'm gonna learn
Best wishes for your future Boris🔥 Thank you for your videos and advice. They have motivated me tremendously and have lead me to finally get a freelance job as a ML Engineer😊 I wish you even more success in life!
@@borismeinardus Did your deep learning lectures covered Tensor computations with higher GPU tasks in multiple dimensions or higher-dimensional arrays?
I really appreciate that you put emphasis on your trials and tribulations before ending up where you wanted to. As a second year CS student at a top university, I often feel the pressure to do everything at once. However I agree with you and the message I got from all of my academic mentors: Building real expertise in a research field takes time! Grüße aus München;)
100%!! Every person has different circumstances. Some might get there faster, some slower, but no matter how much you rush to get the experience, in the end, if you have 10,000h of experience you will be quite good at what you do :) Bleib am Ball! ;)
Bro wtf we are so similar in our thinking. It was the same for me when I was in high school I wanted to go into robotics and chose computer engineering for the same reason as well. Didn’t know ML existed and was in a very rough spot honestly. But then I finally found ML and AI and decided to do Masters. I want to learn more and catch up for all the lost years but it has been really difficult. Hopefully I can reach the research engineer position as well.
Please make a video on salaries in Berlin (Munich & Frankfurt) for ML engineers beginner to senior level. You're just as amazing as your content! Big fan!
I'm a mechatronics technician. Six months ago, I got my first role as an ML engineer after two years of hard studying Data Science and ML to solve problems related to money. Currently, the money problem has been solved by my current job, and I have been questioning myself about what comes next with ML and DL. I have been studying a lot of calculus to understand how the models learn and work, but I still don't know exactly what it means to be an ML researcher. Your explanation will help me continue enhancing my journey. However, I have a big problem related to my path carrier: whether to study physics or a mathematics bachelor's degree. Physics will help me explore the world and model physical phenomena, while mathematics will help me understand probability theories and calculus deeply. I'm not sure which career to choose. The only thing I know is that I will continue studying to become an ML researcher and, eventually, a scientific researcher.
Wow, this is so cool! Congratulations!!! So, to the question of what a ML researcher does... it's not that crazy, really. If you read a lot of papers, you get a feeling for what they actually do. A researcher focuses more on asking good questions, coming up with hypothesis about those unanswered questions, and then implementing models and a lot of experiments to prove or disprove their hypothesis. The workflow generally focuses less on the deployment and product side, but more on this exploration side. That said, this varies significantly depending on whether you do research in academia or at a company and then again varies a lot depending on the company. In the end, if you feel like you love reading papers, start to ask questions no one has asked before, and poke around these state-of-the-art models and find limitations and come up with potential solutions, you are on a great way of getting into research! I hope this somewhat helps 😊 Happy learning and keep it up!!! ❤️
I'm just a civil engineering but I want to learn about ML and AI and your story is really really inspired me to hard work to achieve what i want you deserve more man Wish you the best of luck ❤️❤️
Thank you so much! ❤️ I am a true believer in the 10,000 hours rule. If you put in 10,000 hours into any skill, you will master it. Some people split those 10,000h into 10 years and some into 6. As long as you know it takes time and you enjoy the process, your enthusiasm will be more likely to stay alive for that long :) Keep it up!! ❤️
Please share your video making process if possible. I really love your minimal content style and video editing. I would love to see how you make your videos like how you write scripts, plan video topics, and tools you use to edit them. Of course only if you want to share the journey.
@borismeinardus, I notice a strong resemblance between our approaches. Like you, I am in my final year of my master’s program and working on my thesis. However, unlike you, I live in a developing country where I face numerous obstacles in learning new concepts and finding remote job opportunities. Since my teenage years, I have always enjoyed creating and inventing, which led me to explore AI. I have a question: Do I need to implement research papers from scratch? With the rapid growth of AI and the constant influx of research papers-almost on an hourly basis-starting a paper implementation can be very time-consuming and frustrating, and there’s always a risk it may not work out. My proposed approach for my thesis is to use the existing code from a paper I’m interested in, fix any bugs, and make adjustments for improvements. Do you think this approach is sufficient?
Always! I think it is really important, because if you only see those „perfect people“ you get a false sense of what you need to do to get to the job you want :) Of course, there are insanely talented and hardworking people, but there are more „normal-ish“ people that get the same job :)
Yeah.. If you really want to understand ML and get good at it, it takes a lot of time, let‘s say 10,000h. You can cramp these 10,000h into a tighter schedule, or take your time and spread it out over a few years. Either way, you need to invest the time to really get good at something :)
@@borismeinardus can you tell this to my university xd I never said you can learn this really in two semester but that's what the study plan says.. Tough to develop a real understanding in this speed
Hi! Great video :) I was wondering what is your set up for coding ML? Do you use VS code with anaconda, use some cloud environment or something else entirely? Thanks and have a good one :)
Hi Boris! For the AI startup interview, did you prepare data structure and algorithms or has been enough the knowledge of both deep learning and your experience? Thanks for the answer!
Hi Boris, congratulations for your hardwork and dedication! I am also thinking of learning RL and I was wondering which courses did you take in order to learn robotics and how were you able to correlate RL and robotics. Can you please guide me about it? Any suggestions for robotics course as I have zero knowledge about this field? And as for RL, I am thinking of going for CS234 by Stanford.
Hi Boris, I just enrolled in Data Science course in my college (3rd year), I feel I've no clue what to do. I have basic knowledge about python and I think to get selected I need to become better but I don't know what to be better and how to be better at it. Do you have any advice? How should I start my journey and what tips do u proivde?! I have seen many videos of yours, incredible content.
Congratulation Boris I am also a computer engineering final year master student from Ghana, doing a research in machine learning with goal of becoming an ML research engineer. Would be nice for me if you can share how you went about completing your thesis, what resource you looked at and any general advice you have for someone in my position. Thank you.
Sounds cool and thank you very much for the support ❤️ So, there of course is no one size fits all solution, but in general, one of the best things I did for my thesis was to give myself enough time to read a lot of papers and explore which ideas could make sense or work. Once you real a lot of papers, you get an overview of what was done and where the limitations are of those previous methods. Once you have a feeling for that, you have a point of attack where you start to do janky experiments and poke around to see if your hypotheses of where the models break is true and thus where you can come up with ideas of how to fix it! After you have a general plan of what to try out, the rest is just hustle hustle hustle to actually implement the ideas haha I hope this somewhat helps and keep it up!
Congrats🎉 on the offer. This the first time I engage in the comments on a video. I have one question about what do you do when you feel fatigued after sitting too long Engineering something hard and is not working for you?Waiting for your answer❤ P.s I am an Informatics Engineering student specializing in AI and I will graduate in the next two months. I also followed you on LinkedIn recently ❤
Thank very much ❤️ Oh man, we have all had that feeling and will always have it... It sounds like a stupidly easy solution, but take a break and talk with people. Taking a break resets you mind and allows you to explore different ideas. If you sit on one problem for too long, your mind can't think of anything else than what is directly in front of you and runs against a wall. Talking to others and explaining the problem helps you understand it much better and will very often give you a magical "aha"-moments. Also, more often than not, they can help you by asking you questions about the problem that help you. It genuinely sounds like magic, but happens way more often than you think. (You don't even have to speak to a technical person, or even a real person, lol) You can also ask questions on technical discord servers or ask a question on StackOverflow. If you do that, and don't give up, I am sure you will somehow make it! Keep it up!!! 🚀😊
@@borismeinardus Thanks for the advice. 🌹 Interestingly that happened to me today I was working on a clustering algorithm for customer segmentation project. I took a break then the next morning I was able to solve it very easily. Can I ask you what activites you do beside working and AI in general? I am excited to know 😁
@@WalidBabiAI Haha amazing!! Well, let's see... I go to the gym 4 times a week, play volleyball 2 times a week and cook every day. That's more or less it haha
Hey i started computer science engineering when i was 21. Now im in my 2nd year. Am i done for? Am i behind? Ive just now started learning machine learning.
I am really happy that you have succeeded, but for me there is not so much opportunity to collaborate and work on projects. I also hope that one day I can work on different projects like you and gain experience.
It might be tricky to find something locally where you are, but the internet is am amazing place! I literally know a person, who added one amazing feature to the huggingface library and now got a job there! There are a lot of different open source projects you can have a look at, play around with, and build on top of. There are a few cool online communities (E.g. discord servers of MLST, Aleksa Gordic, Yannic Kilcher, ...) where you can find cool people. I can definitely recommend to look for cool projects on the internet, because there definitely are a lot! You just need to never give up looking and keep building stuff, no matter how simple they might seem in the beginning :) The internet opens a lot of possibilities :) I am certain that if you keep looking and keep going you will get amazing experience and make new friends along the way! ❤️
Hey Boris, I am a French student interested in machine learning and I think I can learn from your experience. So here I am, I have skills in computer vision and I wanted to get started in something brand new and exciting: Reinforcepent Learning! For 1 month I have been training on the subject except that I have reached a point of doubt. This field seems to be neglected in favor of new advances brought with generative AI (transformer, GAN, ...). I tell myself that developing my skills further with more complex and more recent architectures for computer vision would be more beneficial to me. I know that we never learn for nothing but I want to focus on what will bring me the most. So basically the question is RL or computer vision in Big 2024?
Sounds like you are doing great so far! In the end, if you are an expert in any domain, companies will want you. So the general advice would probably be to go for that what you can really passionately explore and become an expert in. Now, that said, you are right, that some people say that RL is not that great, but other say it is... IF I had to pick one of the too, I would probably also say CV is generally a safer bet. But again, if you do CV and more or less hate it, it won't help you either. I know this is not a perfect answer, but I hope it could still help you a bit :) Definitely keep it up no matter what you go for ❤️
Bro i need your guidance as you're an experienced person many RU-vidr are saying ML and DL is obsolete in 2024 it's mean a person wish pursue career in this field will have to struggle to secure any job as per them this field has no future. What is your opinion?
Hey Boris! Any tips on how to start with research papers? I have currently just started and it feels like they are too advanced. Im a 2nd year Computer Science student going into 3rd year. I have a huge interest in ML (I am making ML applications at a startup, and reading research papers to get into the research space).
I also remember seeing only 2 k subs but now bro has reached soo much but i have a question what should i do after completing ibm data science course can you help me a little brother? And guide me please
Crazy! Time flies haha Courses are really great to learn the fundamentals! It's difficult to give general advice in the comment section and without knowing more about your situation haha but I think the best thing to do is get to working on projects. In the video, you saw how many projects I did and how many courses I took at uni. You need both, but courses are easier, because you there just follow what the instructor teaches you and tells you. With a project, it's up to you what you do. Kaggle might be a good place to start! Otherwise, you could try to get creative and work on a passion project of yours that involves ML. I hope this helps :)
bro kindly tell me how to work on project. i mean i am good at theoretical things. when coming to projects, just mindstruck where and how to start. i tried kaggle. but it is not working for me. kindly enlighten me please..
hey iam currently final year of my post graduation and i know AVG level of python programming. i like to set my carrier in machine learning engineer so in which role i need to join as an beginner
The courses are from my university. If you are looking for some that are more easily publicly available, I can recommend Andrew Ng on Coursera, or Jeffrey Hinton on RU-vid :)
I would like to conduct research in DL or ML. I studied mathematics, and due to the "theoretical" nature of my courses (theorem-proof), I would like to know what it's like to do research in DL or ML. I see you have a lot of experience, and if it's possible for you to help me by answering, I would be grateful. I am learning ML and started with the mathematical part, but I feel a bit confused about how deep I should go. If you have any advice, I would greatly appreciate it.
I general, I would argue you really don't need that deep math skills. Most of ML boils down to derivatives and matrix multiplications, and knowing the lingo for how to talk about probabilities. Of course, going deeper will at some point always help. But in general, you will rarely really need to do complex theorem proofs. At least not in the beginning. If you get a to a point where you feel very comfortable with developing and understanding ML models/ papers, you will naturally start to read more and more about the potentially more complex math. But again, even "complex" ML math is nothing compared to the "real math" with pages long proofs. I hope this somewhat helps :) Happy learning and keep it up ❤️
I want to take a machine learning course; I graduate in mathematics. Do you think that I can understand ML and get a job in machine learning by just taking one course from Coursera, brilliant, or Udemy? Please answer my question, I realy need your answer
If you graduate in Math and take such a course you should have decent chance (depending on where you are applying to). But what I would encourage you to do, is to code a lot. Try to build actual ML projects (that doesn’t mean just train a model, but actually try to solve a problem and build some app or demos to present your model). The more time you put in, the better you become and the more impressive projects you can showcase. The course is important. But what will make you stand out are the projects. Keep it up!
i am new to this field . can anyone suggest me a laptop . I have two preferences first is ultra 7 processor with NVidia (4060,4050) and other is apple m3 MacBook(16gb ram , 512rom) or suggest any other
hey man, so you studied in TU Berlin. I heard that tu berlin as well as other german universities bachelor or ug course is only for 3 years. Then why is you timeline is different. can you share it? Im also planning to apply in tu berlin.
After six years of studying machine learning, you gain deep expertise in both the theoretical and practical aspects of the field. This extensive experience equips you with advanced skills in building, deploying, and optimizing machine learning models. Smythos offers resources and tools to support your ongoing learning and help you stay at the cutting edge of machine learning advancements.
Making an entire in depth video on reimplementing a paper is tough haha But what I can highly highly recommend, which is essentially exactly what you are looking for is Andrey Karpathy's video where he builds GPT-2 from scratch! ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-l8pRSuU81PU.html It's an insanely amazing video by one of the greatest ML engineers of our generation!
This course could be a great starting point: huggingface.co/learn/nlp-course/chapter1/1 After that, try to build cool projects (keep in mind, that an ML project does not restrict itself to just training a model, but for example also building an app or demo to really show your work :) Keep it up!
Excuse me , i just got my engineering degree in artificial intelligence and i'm interested in theoretical machine learning , can anyone publish a research paper or do i need a phd ? Thank you.
Hey it's fantastic to know how was your interest in ML I'm also introducing myself in ML for the same reason, I don't know if you heard something about Issac Sim I'm looking to understand how to implement RL to robots if you know something
Sleep is a very very important part of learning anything. When we sleep, we consolidate our knowledge we learned throughout the day and stored in short term memory into our long term memory. There are many things you can try out to get more done in the hours you are awake, and that is my definition of productivity. Getting a lot done in the hours you assign to your work. Of course, sometimes you have to hustle, I don‘t deny that, but the goal should be to somehow have a healthy balance. Again, don’t underestimate the importance of sleep when learning. Keep it up!
Absolutely! If you put in the time, you will become great at ML! And if you start putting in the time this young, you will also become a great ML guy at a young age!
Try to have a good mix of theory and coding :) You will need both, but will be working on coding much more than on math, so keep that in mind 😊 Happy learning and keep it up! ❤️
Oh, really sorry for the confusion!! Saving 6 years is a bit challenging 🫣 haha But I have thought of doing a crash course style video on all the math you need as a ML beginner. But that‘s a bigger and more challenging project to do right. need to think that one through a bit more :/ I hope this video was still valuable and could show you what it takes/ what it took me to get where I am so that you can know what to potentially expect :)
00:01 6-year journey from studying Computer Engineering to working at favorite AI startup 02:35 Gained diverse experience beyond standard computer science courses 04:58 Learning curve plateaued after 8-9 months 07:36 Transition from bachelor to master's focused on machine learning and reinforcement learning. 10:03 Transition to applied scientist student researcher in ML Department 12:40 Steep learning curve in first computer vision project 15:19 Striving for top ML research opportunities after learning from project experiences. 17:46 Struggles with ML internships and shifting focus to research 19:56 Continuing education in Machine Learning with focus on computer vision and deep learning. 22:14 Completed Master's thesis in final year of university, focused on multimodal learning and research. 24:32 Overcame challenges to secure dream AI startup job
