I just finished my CSE Bachelors a few weeks ago, and I agree. Chat gpt is wrong a lot. If you don't know the fundamentals, you wouldn't recognize that. Plus, technology is always changing, and LLM's will always be a little behind the latest stuff. I took a class where we used Elixir with the Phoenix framework, and ChatGPT didn't know much about the latest version.
Can't really say it's wrong if it does not care about only giving users accurate info. Anybody who thinks ChatGPT's job is to provide accurate info does not understand nor should use ChatGPT. Of course, it does not only give users accurate info. ChatGPT will make up stuff and deliver that made-up info to users. It just wants to chat and "please". A lazy stupid lawyer found that out the hard way and now is facing potential sanctions.
Exactly, ChatGPT can't read the documentation. Sure one may argue there plugins, and tools to feed it with the latest documentation. However, it read poorly.
I mean chatgpt is designed for large use cases. Due to recent price cut of gpt4 model, u can just put whole documentation of phoenix framework as prompt and ask questions as cheap as toffee
The thing you have to understand is Chatgpt is literally still in its baby stages. The fact that, even in its infancy, it can generate code and summarise topics to some degree of accuracy is crazy. Nobody is scared of chatgpt now, they're scared of what it will be in 3-5 years time.
from the perspective of a self-taught programmer, I think that when faced with a problem, whether it's ChatGPT or StackOverflow, you still need to use the right terminology to (hopefully) get an answer that works. Of course what you also need, is logical thinking, problem solving skills, as well as being able to take an example or concept (from whatever source), and adapt it to what you need. Oh, and I think a Computer Science degree is always going to be a useful and valuable thing to have.
I’m 39 and have a solid job now but I’ll be stuck where I live. I’ve been considering going back to school and finishing finally my bachelor’s but not sure in what field. Seriously have been considering CS but don’t know if my brain is capable anymore! Plus, it’s hard finding good answers and good advice. Your video has helped motivate me a little but the future of the field still seems a bit uncertain.
@@Thuvu5Hello. Please help with your advice. I am interested in programming, I love technology, Linux. And now my father says that they can help me get a degree in Computer Science. But some say that this education is useless. I remember your video in which you said that CS it's a good education but it's been a while since that time. Please tell me what you think about it now. thank you
On the contrary, since AI can write code anyway, having degree will give you a proper edge over bootcamps/self-taughts when it comes to understanding of the system
I was in IT for most of my adult life. I’m now retired. I urge you to complete your degree. I am sure it will be useful. Also brush up on linear algebra, calculus and statistics. All the best.
For actual computer science, sure it is - but a lot of what people would want to do in tech has very little to do with what is covered in CS academia. "This is computer science - why would you be coding? Now, lets talk about circuits..."
I believe that having a fundamental understanding of how stuff works will be valuable in the 21st century. You might learn how stuff works at a fundamental level by taking physics, electrical engineering or computer science. A math or physics degree might be useful for the intimidation factor. Know your applied math (or at least how to get things done on a computer). I would suggest taking intro level courses in the social sciences. I used to think economics was helpful, economics has sort of a barbell utility; the intro (first year) level is useful and econometrics is useful, but everything else is if not a waste of time at least a high opportunity cost when one could be studying other subjects. The underlying human behavior paradigm of economics is too narrow (study some intro psychology). Statistics is very useful, but again the underlying paradigm is too narrow, though Bayesian statistics will continue to grow in importance.
I think "low tech" tech jobs will have to be more technical. Like, I'm surprised that cybersecurity professionals don't even have to learn to code, or how a modern system works. At my job, cybersecurity people help to fill out forms on risk assessment, while the actual security engineers have a cloud infrastructure background and had to teach themselves about the security aspect. Like you said, I think you'll have to have a more technical understanding to effectively use AI in your tech role.
ChatGPT is a long term concern, the rest are short term. Tech layoffs can/will reverse eventually, hiring freezes come and go, chip shortages will change once the new supply chain's are up and running, and competition will always be high as tech advances. ChatGPT is at the beginning and the impact it can/will have is a concern for the whole world whereas everything else are local concerns.
People say that AI is still not very good at writing code, but people dont consider that a GPT model will be trained to code and we know its very close to doing so competently. ChatGPT as an example gets updated every week. This will happen in less time than you all imagine.
Back in 70s and &0s, lots of people believed that coding or application development would be done by high school kids, not college graduates. 40 years later, college graduates are still dominate the tech industry. Now people believe IA will soon replace college graduates in CS. I would think that AI will replace other jobs first before CS jobs be replaced. Who will be continuing AI development? It may still have a long way to go for AI to be able to develop AI without human involvement. That human involvements are CS/tech jobs.
All the people who fearmonger about AI don't know what AI is at this point. Custom solutions are still necessary. Analysis is till necessary, and the design choices for the analysis as well. Basically computer scientists will be to AI later what economists are to companies. The only problem right now is that people don't realize this, and so everyone is saying "What do I need an economist for?". A tectonic event has to happen for everyone to become scared, and then they'll all start hiring computer scientists. Tectonic event like when GDPR came up for Facebook in the U.S. senate, and suddenly every single company reacted to it.
Get a masters degree that focuses on machine learning, that way you can create the machines, therefore the bachelor degree is not useless as you need it for machine learning.
I completed CS at Open Uni. It is a worthless degree, unless you have the tone of experience or a big talent. Even unpaid internships are extremely competitive, at least in the UK. Better go to Medicine, or Engineering, or any Civil Service job. AI will remove any need for tech jobs at entry level in the next few years. Anyone who tells you otherwise is lying. There is literally zero demand for Computing Science graduates.
A computer science degree is not worthless. It’s oversaturated and tech workers didn’t unionize therefore they dealt a significant blow with layoffs. You just have pick fields that are viable which are cybersecurity, game development, ai technology, and robotics.
You need a formal education to be able to articulate yourself in a clear and concise manner. Sure, ChatGPT builds great prototypes, but without a formal background in the topic, or proper understanding of the framework of the libraries that you're working with, you'll be in for an endless loop or garbage-in and garbage-out. Besides, you'll be in a bigger disaster if you're ever asked to present the code that you have "written" by walking through the implementation process. In the question of whether "formal education" must be strictly a computer science degree; in my opinion, the answer is no. In any undergraduate discipline, you are expected to have a grasp of critical thinking. This means that you should know what questions to ask and what facts or observations to ignore (right, the BS detector). It would definitely be beneficial if you have taken at least some CS101 courses, ie knowing how the OOP paradigm works by heart is a good place to start, if not, then this is where continuous learning, either through self-taught or taking relevant courses, comes in. On a related note, whenever you are self-learning, you accept the risk that you might follow a suboptimal course structure in exchange for the benefit of paying lower tuition costs. I think the best way to mitigate this risk is to critically examine the course syllabuses before starting the studies (and know that you have the prerequisites). Don't start a course if you think you will get lost when you're halfway through it. TH-cam channels (like this one) can be a great resource for learning about big ideas such as what latest development toolkits you should consider, and sharing strategies for learning in general.
The ONLY reason that corporations require Computer Science Degrees is because they are getting kickbacks from the Federally backed University system. On top of that, algorithms are constantly searching for career potential largely based on a lineage of 'degrees' in that area. Lastly, you have the ability for students to constantly undermine existing parameters to get leverage over other students and job applicants. In a world of AI, this will only become more apparent as the kinks in ChatGPT and other applications get sorted out. If anybody and their brother can get an AI written assignment turned in. Then a degree means nothing, aside from sorting out 'honest' people from 'con-artists'.
Im really eager to learn. I want to obtain a associates in AI engineer or cs. So far cs seems important and ai engineer seems to be very limited to ai and not things outside of it
Yeah it’s been over a year, programming/coding will 100% be USELESS. As A.I. advances, it will be the end of the human intelligence. With high EQ, not anything lower, will be the main asset. If you want better digital security, shouldn’t you just get a more advanced A.I ??? Just for an analogy, thousands of years ago hunting was for the majority. Now we just pay a better hunter if it means to want more meat. Thus, no need to learn to hunt. Humans have 3 main factors in the workspace. Body, mind, emotion. We’ve given up the body since the Industrial Revolutionary. And when AGI is here, we’ll then start giving up the mind. Emotion is what’s left, but a high EQ will be what’s needed.
There’s never a scenario where 2 equals candidates and only one has a degree. The reality is that it’s over saturated with people with degrees. Way too saturated with entry level degree holders.
i am an electrical engineer and don't know much about computers and IT. would you please tell me that learning data science is easy to learn or you recommend something else to earn money with?
Firstly, thank you for this video. Secondly, "I don't agree with your point. Specifically, the section about AI in my piece didn't address hacking issues because you may not have inquired about all facets of the particular problem. Building on the discussion in this video, I believe we need you to create another video focused on prompt engineering, particularly CoT and ToT. This will assist us in fully leveraging AI."
Total nonsense. A cs degree while it may seem long and expensive, the opportunities it opens up and how much more competitive it makes you will pay off greatly in the long run. Sure you can learn everything without a cs degree but the thing is that hr departments need to weed out candidates so most require a degree to thin out the pool of applicants.
Computer science degree will be absolutely irrelevant. For your information, 80 % of software development in 2022 was low code no code. It is just the beginning. University degree as a whole will still be relevant, agree on that. But not computer science as a degree.
I just finished my CSE Bachelors a few weeks ago, and I agree. Chat gpt is wrong a lot. If you don't know the fundamentals, you wouldn't recognize that. Plus, technology is always changing, and LLM's will always be a little behind the latest stuff. I took a class where we used Elixir with the Phoenix framework, and ChatGPT didn't know much about the latest version.
Can't really say it's wrong if it does not care about only giving users accurate info. Anybody who thinks ChatGPT's job is to provide accurate info does not understand nor should use ChatGPT. Of course, it does not only give users accurate info. ChatGPT will make up stuff and deliver that made-up info to users. It just wants to chat and "please". A lazy stupid lawyer found that out the hard way and now is facing potential sanctions.
Exactly, ChatGPT can't read the documentation. Sure one may argue there plugins, and tools to feed it with the latest documentation. However, it read poorly.
I mean chatgpt is designed for large use cases. Due to recent price cut of gpt4 model, u can just put whole documentation of phoenix framework as prompt and ask questions as cheap as toffee
Yall just wait
The thing you have to understand is Chatgpt is literally still in its baby stages. The fact that, even in its infancy, it can generate code and summarise topics to some degree of accuracy is crazy.
Nobody is scared of chatgpt now, they're scared of what it will be in 3-5 years time.
from the perspective of a self-taught programmer, I think that when faced with a problem, whether it's ChatGPT or StackOverflow, you still need to use the right terminology to (hopefully) get an answer that works.
Of course what you also need, is logical thinking, problem solving skills, as well as being able to take an example or concept (from whatever source), and adapt it to what you need.
Oh, and I think a Computer Science degree is always going to be a useful and valuable thing to have.
I’m 39 and have a solid job now but I’ll be stuck where I live. I’ve been considering going back to school and finishing finally my bachelor’s but not sure in what field. Seriously have been considering CS but don’t know if my brain is capable anymore! Plus, it’s hard finding good answers and good advice. Your video has helped motivate me a little but the future of the field still seems a bit uncertain.
AI is crazy RN! Such a valid question to answer! 🙌🏼
True! So curious how it will turn out in the next few months and years 🤔
@@Thuvu5Hello. Please help with your advice. I am interested in programming, I love technology, Linux. And now my father says that they can help me get a degree in Computer Science. But some say that this education is useless. I remember your video in which you said that CS it's a good education but it's been a while since that time. Please tell me what you think about it now. thank you
On the contrary, since AI can write code anyway, having degree will give you a proper edge over bootcamps/self-taughts when it comes to understanding of the system
I was in IT for most of my adult life. I’m now retired. I urge you to complete your degree. I am sure it will be useful. Also brush up on linear algebra, calculus and statistics. All the best.
congratulations
you dont need maths
@@lemonstrangler Depends what your doing. There are many sub-fields in computer science that require advance math.
@@lemonstrangler React developer? 🤣
For actual computer science, sure it is - but a lot of what people would want to do in tech has very little to do with what is covered in CS academia. "This is computer science - why would you be coding? Now, lets talk about circuits..."
I believe that having a fundamental understanding of how stuff works will be valuable in the 21st century.
You might learn how stuff works at a fundamental level by taking physics, electrical engineering or computer science. A math or physics degree might be useful for the intimidation factor. Know your applied math (or at least how to get things done on a computer). I would suggest taking intro level courses in the social sciences. I used to think economics was helpful, economics has sort of a barbell utility; the intro (first year) level is useful and econometrics is useful, but everything else is if not a waste of time at least a high opportunity cost when one could be studying other subjects. The underlying human behavior paradigm of economics is too narrow (study some intro psychology). Statistics is very useful, but again the underlying paradigm is too narrow, though Bayesian statistics will continue to grow in importance.
What is the underlying paradigm in statistics?
such as enlightening you tell about software engineering 💥💫
I think "low tech" tech jobs will have to be more technical. Like, I'm surprised that cybersecurity professionals don't even have to learn to code, or how a modern system works. At my job, cybersecurity people help to fill out forms on risk assessment, while the actual security engineers have a cloud infrastructure background and had to teach themselves about the security aspect.
Like you said, I think you'll have to have a more technical understanding to effectively use AI in your tech role.
What type of job is that? The title or company. I'm looking for technical work that I can get experience while studying.
Thankss!! This video helped me select my major in school...❤
I like how the concern is ChatGPT and Not the tech layoffs, hiring freezes, chip shortage, or the fact the competition is higher than ever
ChatGPT is a long term concern, the rest are short term. Tech layoffs can/will reverse eventually, hiring freezes come and go, chip shortages will change once the new supply chain's are up and running, and competition will always be high as tech advances. ChatGPT is at the beginning and the impact it can/will have is a concern for the whole world whereas everything else are local concerns.
People say that AI is still not very good at writing code, but people dont consider that a GPT model will be trained to code and we know its very close to doing so competently. ChatGPT as an example gets updated every week. This will happen in less time than you all imagine.
I have subscribed your channel when you had almost 1k subscribed. So nice to see you grow so fast..
Hey Tanya, thank you for being one of my first subscribers, and following me for such a long time! 💖
Back in 70s and &0s, lots of people believed that coding or application development would be done by high school kids, not college graduates. 40 years later, college graduates are still dominate the tech industry. Now people believe IA will soon replace college graduates in CS. I would think that AI will replace other jobs first before CS jobs be replaced. Who will be continuing AI development? It may still have a long way to go for AI to be able to develop AI without human involvement. That human involvements are CS/tech jobs.
First Comment : Great video as usual thu vu, wish you all the best
All the people who fearmonger about AI don't know what AI is at this point. Custom solutions are still necessary. Analysis is till necessary, and the design choices for the analysis as well. Basically computer scientists will be to AI later what economists are to companies.
The only problem right now is that people don't realize this, and so everyone is saying "What do I need an economist for?". A tectonic event has to happen for everyone to become scared, and then they'll all start hiring computer scientists. Tectonic event like when GDPR came up for Facebook in the U.S. senate, and suddenly every single company reacted to it.
Get a masters degree that focuses on machine learning, that way you can create the machines, therefore the bachelor degree is not useless as you need it for machine learning.
I completed CS at Open Uni. It is a worthless degree, unless you have the tone of experience or a big talent. Even unpaid internships are extremely competitive, at least in the UK. Better go to Medicine, or Engineering, or any Civil Service job.
AI will remove any need for tech jobs at entry level in the next few years. Anyone who tells you otherwise is lying.
There is literally zero demand for Computing Science graduates.
A computer science degree is not worthless. It’s oversaturated and tech workers didn’t unionize therefore they dealt a significant blow with layoffs. You just have pick fields that are viable which are cybersecurity, game development, ai technology, and robotics.
You need a formal education to be able to articulate yourself in a clear and concise manner. Sure, ChatGPT builds great prototypes, but without a formal background in the topic, or proper understanding of the framework of the libraries that you're working with, you'll be in for an endless loop or garbage-in and garbage-out. Besides, you'll be in a bigger disaster if you're ever asked to present the code that you have "written" by walking through the implementation process.
In the question of whether "formal education" must be strictly a computer science degree; in my opinion, the answer is no. In any undergraduate discipline, you are expected to have a grasp of critical thinking. This means that you should know what questions to ask and what facts or observations to ignore (right, the BS detector). It would definitely be beneficial if you have taken at least some CS101 courses, ie knowing how the OOP paradigm works by heart is a good place to start, if not, then this is where continuous learning, either through self-taught or taking relevant courses, comes in.
On a related note, whenever you are self-learning, you accept the risk that you might follow a suboptimal course structure in exchange for the benefit of paying lower tuition costs. I think the best way to mitigate this risk is to critically examine the course syllabuses before starting the studies (and know that you have the prerequisites). Don't start a course if you think you will get lost when you're halfway through it. TH-cam channels (like this one) can be a great resource for learning about big ideas such as what latest development toolkits you should consider, and sharing strategies for learning in general.
Absolutely agreed! Thank you for such a thoughtful comment 🙌🤩
The ONLY reason that corporations require Computer Science Degrees is because they are getting kickbacks from the Federally backed University system. On top of that, algorithms are constantly searching for career potential largely based on a lineage of 'degrees' in that area. Lastly, you have the ability for students to constantly undermine existing parameters to get leverage over other students and job applicants.
In a world of AI, this will only become more apparent as the kinks in ChatGPT and other applications get sorted out. If anybody and their brother can get an AI written assignment turned in. Then a degree means nothing, aside from sorting out 'honest' people from 'con-artists'.
I have just one thing to say :
"...we should ride the waves..."
Im really eager to learn. I want to obtain a associates in AI engineer or cs. So far cs seems important and ai engineer seems to be very limited to ai and not things outside of it
Love from Vietnam ❤
I’m thinking of making my major a double, business and Comp sci with a focus on AI. would that be a good idea?
I personally think it would be. Dual degrees that merge tech are becoming increasingly common.
Yeah it’s been over a year,
programming/coding will 100% be USELESS. As A.I. advances, it will be the end of the human intelligence. With high EQ, not anything lower, will be the main asset. If you want better digital security, shouldn’t you just get a more advanced A.I ??? Just for an analogy, thousands of years ago hunting was for the majority. Now we just pay a better hunter if it means to want more meat. Thus, no need to learn to hunt.
Humans have 3 main factors in the workspace. Body, mind, emotion. We’ve given up the body since the Industrial Revolutionary. And when AGI is here, we’ll then start giving up the mind. Emotion is what’s left, but a high EQ will be what’s needed.
Can someone help me.. Master's of Computer science or Master's of Artificial intelligence which is more beneficial in 2024 😊
There’s never a scenario where 2 equals candidates and only one has a degree. The reality is that it’s over saturated with people with degrees. Way too saturated with entry level degree holders.
❤❤
i am an electrical engineer and don't know much about computers and IT. would you please tell me that learning data science is easy to learn or you recommend something else to earn money with?
Which university are you studying in dear?
University of London. Got it.
Wait. Hold on! There are four universities of London?!
I agree with you but I have a question if I come from the business field, Is a degree necessary for becoming a data analyst? thanks anyway.
I don't think a CS degree is necessary to become a data analyst :)
@@Thuvu5 thanks ,look forward for your next video.
would you recommend a master's in CS or just AI?
Computer Science definitely with a focus on AI. Do a CS master in AI
Get use to working in a different field for those who got cs degrees, Ai is coming for your jobs
Like working with/on AI? You are confusing CS with self-taught web developers...AI is one of many fields of CS.
Firstly, thank you for this video. Secondly, "I don't agree with your point. Specifically, the section about AI in my piece didn't address hacking issues because you may not have inquired about all facets of the particular problem. Building on the discussion in this video, I believe we need you to create another video focused on prompt engineering, particularly CoT and ToT. This will assist us in fully leveraging AI."
Hey can you reçommènd me a online degrees exam , please..
Note; financial problem
Go get your CS degree. A Computer science degree is not a programming degree.
Is a Math Science Degree worthless since we have calculators, of course not. We study to innovate, create, not repeat like a parrot machine aka LLMs
even without ai cs is useless lol. study in a bootcamp or maybe study i.t
Total nonsense. A cs degree while it may seem long and expensive, the opportunities it opens up and how much more competitive it makes you will pay off greatly in the long run. Sure you can learn everything without a cs degree but the thing is that hr departments need to weed out candidates so most require a degree to thin out the pool of applicants.
My cs degree is useless. I'm tempted to throw it in a fire pit someday.
Why do you think it’s useless?
Computer science degree will be absolutely irrelevant. For your information, 80 % of software development in 2022 was low code no code. It is just the beginning. University degree as a whole will still be relevant, agree on that. But not computer science as a degree.
what degrees do you think will be relevant?
Source that shows your claim of 80% of software dev in 2022 being low/no code please.
@@joshuaolds4747 👍
I disagree, a lot of jobs nowadays are requiring you have a degree in something technical.
This doesn't make sense because the creation ,maintenance and enhancement of AI itself requires computer science background.
s b