I have an important question noone is answering. I'm getting a CS degree and will probably get a Masters, but how do I know what knowledge I need to enter the workforce after the degree? Because I don't believe you can just work with strictly what they teach at university
Of all the skills, communication is the most important one in the age of AI because in order to capitalize on AI, you need to be able to articulate precisely what you want it to build and what problems you're trying to solve.
The advice "Code out loud" is something I would also recommend to everyone! Lately I've been doing that even without noticing but it does actually help me to keep track of what I'm doing and what I want to do at that specific moment. And since English isn't my native language, it also helps me to memorize all of the code related terminology and just improves my general language skills (English and JavaScript, haha).
At the end of the day everyone is self taught. A computer Science or a related degree are very raw. Even if you've a degree you really have to learn a lot by yourself
Being a great developer, for me at least, has always been about being able to learn, adapt, and solve problems. Code and languages are the tools we use to solve problems. Being able to work in a team and communicate, encourage others, and solve problems together is imperative for success.
Thankyou for the reassuring words. Iv lined myself to take a 2 year college course for software development this September, Im 30 now and have no experience at all other than learning some CS fundamentals and other coding basics before it starts. I was worried once I started thinking about what AI can and will do, and how much longer it will take me to not only finish the course, but also build the skill enough to fit into the skill set that many others will have once moving into using ai with their coding. I feel if I had already known code since maybe 5 years ago, it would have been an okay transition. But now, I had fears with time=success. Almost considered switching to maybe cyber security. But, this video helped a bit, and so long as I know what resources to study, maybe I wont have to re-learn a bunch of things that many others will have to.
I love this video. I have been developing across frameworks for 15 years in my career, but I admit there are holes of “fundamental” knowledge. So I am getting an MSCS degree while working full time. I was incredibly torn on “wasting” my time on this. But with this new age of AI, I felt it was needed whether I liked it or not. This video has given me the affirmation I needed to stay motivated. Thank you so much!!
This video is wrong. The author is a great guy, but this time he is giving you the wrong advice which will bring you to the 'incredibly torn on wasting my time' experience again.
@@ucol2068 First, do not try to outsmart the system which is a trillion times smarter than you. Second, find the niche where AI is helpless and excell there.
The reason they don’t want self taught programmers is because they are way less likely to have the debt that getting a degree gives meaning they are way less likely to stay at that job because they are not financially obligated to. It’s beyond obvious to me it’s simple logic when you put it together.
This is incredibly spot on. Thanks for sharing Travis! I love seeing these videos that are more pragmatic. Especially the highlighting of just doing the hard work. Always enjoy your take
"Self-taught"....Many "Self-taught" people that I know actually had degrees in things like Engineering (not computer science or computing), stats, or math etc. A lot of people are not actually "self-taught", they have backgrounds in something technical and the mathematical background to get themselves up to speed. There are some who are self-taught, and there were a lot more before the bootcamps came. However, now that we have bootcamps, I don't think that people who go to them can say they are self-taught anymore. They formally learn logic and the essential technology necessary to be productive. Computer Science majors don't learn technology per se, they learn the logic and structures necessary to design algorithms and build them using computer language.
``` Qs.: What is the Right Answer? Ans.: A Good Questions. ``` This is why learning the Fundamentals is ever so important now. ChatGPT's answers are just as good as your questions are. I often find myself making ChatGPT confused or going around in circles because I mislead it with my ambiguous questions. But when I give it direct hints, in my questions, as to what I really expect, the answers mostly come out very well structured. The soft skills apply to communicating with generative AI as well I feel but the exposure to the Fundamentals is what lets us Devs to get the most out of them. These have just been my observations with ChatGPT. Thanks for putting them into perspective with this video!
It's super weird how people are freaking out about AI. Use common sense! Do you really think any state of any country in the world is going to simply allow their workforce to be completely replaced by AI? That means 30% to 40% unemployment. That means, along with a 10% to 20% aging population collecting a pension, billions and billons of dollars given out as unemployement. That means, sky high inflation. Nobody would be spending any money on anything. How is any state going to collect taxes if half of their population is unemployed? Replacing your workforce would completely destroy the economy of any country and nobody will let that happen. Realistically, many workers will shift careers. New jobs and opportunities will be created and people will simply continue on their path. This has happened so many time in the past that it's hard to believe people simply understand that....
You are 100% correct - but today’s politicians - worldwide - are beyond co-opted and incompetent to avert the inevitable disaster, even if you are only 25% wrong.
Have you heard of Pandora’s box? The moment it’s opened there’s no going back and that’s AI. There’s no stopping the progress it’s like trying to stop aging, there are ways to slow it down but you’re not gonna stop it.
Humble Bundle have HUGE sales for packt books. Keep an eye on HB and grab 20 books for $25. Though a lot of them are garbage but for 25 bucks, even if only 2 are good it's a deal.
In reality though, AI really hasn't advanced all that much and all these LLMs we've come to love are based on the 2017 transformers architecture released by Google in 2017. They just simply threw a bunch of extra compute and training data at the model, then had a eureka momenet, never expected Chat GPT 4 to function this well and still aren't quite sure why it does so well. Then we've had some advancements in image and audio generation, and that's true. There's that Q* leak from last fall that has been confirmed to be a project within OpenAI, but that's it. Open source LLMs have taken off ala HuggingFace which is awesome. Then everything else is basically just a Chat GPT 4 wrapper. That's it though. Really, there's been Chat GPT 4, some image / audio generation, and some video demos of cool shit none of us are allowed to play with. That's literally it so far though, the rest if still just hype atm.
I am 16 years old passionate about ai and ml and robotics, cyber security But i am worried that by the time I graduate college there will not be jobs because of ai and agi. Any tips??
All this doom and gloom is pure marketing from the usual suspects... I think in 5 /10 years there will be even more demand for developers. AI is an accelerator not a replacement
I agree AI is an accelerator The AI evangelists are a little bit too optimistic about this. The next big thing in AI is always right around the corner. But then, years go by, and there is a little progress, but certainly not enough to be a replacement for an actual programmer. Myself and others I know have seen multiple times how AI fails at programming. Just remember how long have self driving cars been right around the corner? Over a decade and they still have issues in many scenarios.
AI is a useful tool, nothing more. To use it, you must know what to ask, how to ask, and how to troubleshoot the code that it provides. To do that, you must understand the fundamentals.
I use GPT4 to write Bash script to manipulate config file. It generates script to solve a simlle use case, but it is very difficult to solve complicated use case. With comprehensive awk and sed knowledge and good problem definition, I can give a good strategy to ChatGPT. However, I did not have, so it had kept generating incomplete solution because the strategy was wrong. With wrong strategy, the solution becomes very complex and easily out of control.
I use chat GPT almost everyday and it is quite helpful. But if you ask questions like "In binary tree data structure, depth of tree is log base 2 (n) -1?", "So If I have 7 nodes?" and "so you mean root node is in depth 1?". More likely you would start getting contradicting answers. This is just one example.
@@tedspens 1.'to know what to ask' you have to be an expert, 2. to know 'how to ask', you have to be an expert in GPT itself, 3. to know 'how to troubleshoot code', you have to leverage a more powerful GPT and it will tell you how. What does it all have to do with 'knowledge of fundamentals'?
Meaning, you need to learn the stuff - which is not guaranteed with a degree, even if people have used it as a proxy for being knowledgeable and qualified.
Please watch Andrew Ng video about Agentic Loop, it will make traditional devs legacy even at GPT 4 True now is the time to learn the fundamentals and most importantly improve your soft skills Within 6 months, as workflow improve, we will see the rise of AI agents affecting many knowledge workers
despite the fact they will ask you about fundamentals during an interview, eventually you will be still doing html through the next super framework :) Hiring is extremely lazy this days...
Well, it looks like it's time for all of us software developers to retire, because the First AI Software Engineer Tool is here to replace us. I guess all those years of debugging and coffee addiction were for nothing. But hey, at least we had a good run, right? And let's not forget about those low code platforms that were supposed to turn everyone into a programmer overnight. I'm still waiting for my non-techy friend to deploy his first app. It's almost like every decade has its 'This will make humans obsolete' trend. Remember when automatic elevators were going to leave elevator operators jobless? Good times. Guess we'll just have to find solace in the fact that we're still here, and venture capital money still needs somewhere to go! 😂
We did everything that the internet tells us to do and more from our learnings, self taught developers are too much work, too much risk and very little reward. They try to nose dive right from the get go even after being directly and indirectly told to ask questions and get clarifications/agreement. They then miscommunicate progress even after being told to be transparent and while others are. They make our other devs and lead devs focus on covering for them when we want them to focus on customers and products.
A lot of the folks might've done 1 hour of practice and will consider themselves a "self-taught" dev, that's why there so much negative connotation around it lol
Cancelled subscription. This video will not age well... We are doomed, because these un-trained, un-qualified people got employed and did "their best". It was simply not good enough. Like the LLM now. Compare their output to mine in case of code quality and functionality: seems to be missing A LOT. Because they learned from these unqualified people, from stackoverflow, from all these sources... So, waiting for the collapse of this bubble.
Use my link sponsr.is/tld_travismedia and APPDEVFOO5 and get a .dev, .app, or .foo domain name for only $5!
I have an important question noone is answering.
I'm getting a CS degree and will probably get a Masters, but how do I know what knowledge I need to enter the workforce after the degree? Because I don't believe you can just work with strictly what they teach at university
Of all the skills, communication is the most important one in the age of AI because in order to capitalize on AI, you need to be able to articulate precisely what you want it to build and what problems you're trying to solve.
Yeah, we're also gonna need to be able to inform potential clients that their AI-produced code base is a hot mess in a way that doesn't offend them.
The advice "Code out loud" is something I would also recommend to everyone! Lately I've been doing that even without noticing but it does actually help me to keep track of what I'm doing and what I want to do at that specific moment. And since English isn't my native language, it also helps me to memorize all of the code related terminology and just improves my general language skills (English and JavaScript, haha).
At the end of the day everyone is self taught. A computer Science or a related degree are very raw. Even if you've a degree you really have to learn a lot by yourself
Being a great developer, for me at least, has always been about being able to learn, adapt, and solve problems. Code and languages are the tools we use to solve problems. Being able to work in a team and communicate, encourage others, and solve problems together is imperative for success.
Thankyou for the reassuring words. Iv lined myself to take a 2 year college course for software development this September, Im 30 now and have no experience at all other than learning some CS fundamentals and other coding basics before it starts. I was worried once I started thinking about what AI can and will do, and how much longer it will take me to not only finish the course, but also build the skill enough to fit into the skill set that many others will have once moving into using ai with their coding. I feel if I had already known code since maybe 5 years ago, it would have been an okay transition. But now, I had fears with time=success. Almost considered switching to maybe cyber security. But, this video helped a bit, and so long as I know what resources to study, maybe I wont have to re-learn a bunch of things that many others will have to.
I love this video. I have been developing across frameworks for 15 years in my career, but I admit there are holes of “fundamental” knowledge. So I am getting an MSCS degree while working full time. I was incredibly torn on “wasting” my time on this. But with this new age of AI, I felt it was needed whether I liked it or not. This video has given me the affirmation I needed to stay motivated. Thank you so much!!
This video is wrong. The author is a great guy, but this time he is giving you the wrong advice which will bring you to the 'incredibly torn on wasting my time' experience again.
Which part is wrong ?@@maksymdudyk1718
@@maksymdudyk1718and then what is right?
@@ucol2068 First, do not try to outsmart the system which is a trillion times smarter than you. Second, find the niche where AI is helpless and excell there.
The reason they don’t want self taught programmers is because they are way less likely to have the debt that getting a degree gives meaning they are way less likely to stay at that job because they are not financially obligated to. It’s beyond obvious to me it’s simple logic when you put it together.
This is incredibly spot on. Thanks for sharing Travis! I love seeing these videos that are more pragmatic. Especially the highlighting of just doing the hard work. Always enjoy your take
"Self-taught"....Many "Self-taught" people that I know actually had degrees in things like Engineering (not computer science or computing), stats, or math etc. A lot of people are not actually "self-taught", they have backgrounds in something technical and the mathematical background to get themselves up to speed. There are some who are self-taught, and there were a lot more before the bootcamps came. However, now that we have bootcamps, I don't think that people who go to them can say they are self-taught anymore. They formally learn logic and the essential technology necessary to be productive. Computer Science majors don't learn technology per se, they learn the logic and structures necessary to design algorithms and build them using computer language.
```
Qs.: What is the Right Answer?
Ans.: A Good Questions.
```
This is why learning the Fundamentals is ever so important now. ChatGPT's answers are just as good as your questions are. I often find myself making ChatGPT confused or going around in circles because I mislead it with my ambiguous questions. But when I give it direct hints, in my questions, as to what I really expect, the answers mostly come out very well structured.
The soft skills apply to communicating with generative AI as well I feel but the exposure to the Fundamentals is what lets us Devs to get the most out of them.
These have just been my observations with ChatGPT. Thanks for putting them into perspective with this video!
I have an undergraduate and master degree in CS and also have worked a few years in Big tech. There is always something to learn from your videos.
Good level-headed advice. I like the references to information resources too.
It's super weird how people are freaking out about AI. Use common sense! Do you really think any state of any country in the world is going to simply allow their workforce to be completely replaced by AI? That means 30% to 40% unemployment. That means, along with a 10% to 20% aging population collecting a pension, billions and billons of dollars given out as unemployement. That means, sky high inflation. Nobody would be spending any money on anything. How is any state going to collect taxes if half of their population is unemployed?
Replacing your workforce would completely destroy the economy of any country and nobody will let that happen. Realistically, many workers will shift careers. New jobs and opportunities will be created and people will simply continue on their path. This has happened so many time in the past that it's hard to believe people simply understand that....
You are 100% correct - but today’s politicians - worldwide - are beyond co-opted and incompetent to avert the inevitable disaster, even if you are only 25% wrong.
Well said
That's not really how capitalism works, it will rather destroy high skill roles and force people into service role.
Have you heard of Pandora’s box? The moment it’s opened there’s no going back and that’s AI. There’s no stopping the progress it’s like trying to stop aging, there are ways to slow it down but you’re not gonna stop it.
Are you stupid?its the company that doesnt wanna hire bro thats how anybody is replaced
How do you tell the difference between good and bad books on packt?
Reading amazon reviews I guess
Thanx Travis. I appreciate the content, suggestions, and links.
Mos' Def worth a subscription ;D
Self taught dev here. Still diplomed engineer, but not in IT domain. Does it count?
what a great mate!!! thanks so much for this informative vid
Excellent delivery!!! Great work! Thank you 🕊️🙏❤
how to take the first job
Great advice... You hit a home run!
Do we need to learn this if we are frontend devs
Humble Bundle have HUGE sales for packt books. Keep an eye on HB and grab 20 books for $25. Though a lot of them are garbage but for 25 bucks, even if only 2 are good it's a deal.
Thanks the video - do you have any previews of the Imposter Devs material; I'd like to get a taste of it before signing up.
In reality though, AI really hasn't advanced all that much and all these LLMs we've come to love are based on the 2017 transformers architecture released by Google in 2017. They just simply threw a bunch of extra compute and training data at the model, then had a eureka momenet, never expected Chat GPT 4 to function this well and still aren't quite sure why it does so well.
Then we've had some advancements in image and audio generation, and that's true. There's that Q* leak from last fall that has been confirmed to be a project within OpenAI, but that's it. Open source LLMs have taken off ala HuggingFace which is awesome. Then everything else is basically just a Chat GPT 4 wrapper.
That's it though. Really, there's been Chat GPT 4, some image / audio generation, and some video demos of cool shit none of us are allowed to play with. That's literally it so far though, the rest if still just hype atm.
Kindly elaborate further..any book recommendations on transformers and other essential things to learn..would be helpful.. thanks🙏🙏
What happens when the increasingly wild promises for AI fail to materialise and the VC capital stops?
I am 16 years old passionate about ai and ml and robotics, cyber security
But i am worried that by the time I graduate college there will not be jobs because of ai and agi.
Any tips??
stop worrying about things you can't control and just be better than everyone else
All this doom and gloom is pure marketing from the usual suspects... I think in 5 /10 years there will be even more demand for developers. AI is an accelerator not a replacement
I agree AI is an accelerator The AI evangelists are a little bit too optimistic about this. The next big thing in AI is always right around the corner. But then, years go by, and there is a little progress, but certainly not enough to be a replacement for an actual programmer. Myself and others I know have seen multiple times how AI fails at programming. Just remember how long have self driving cars been right around the corner? Over a decade and they still have issues in many scenarios.
Thanks for share this good advices with such an optimistic view
Thanks Travis. your every video is Eyeopener for self taughts.
Problem solving skills will always be valued no matter what
I mean that is soooo vague
Amazing video and helpful, thanks
Certainly saving this book.
Great vid. You deserve every dollar you get off TH-cam
This was really helpful, I subscribed ❤
This is very useful!
Solid speaker, thank you for the content!
Thank you. But how knowledge of fundamentals helps us contradict the upcoming total dominance of AI?
AI is a useful tool, nothing more. To use it, you must know what to ask, how to ask, and how to troubleshoot the code that it provides. To do that, you must understand the fundamentals.
I use GPT4 to write Bash script to manipulate config file. It generates script to solve a simlle use case, but it is very difficult to solve complicated use case. With comprehensive awk and sed knowledge and good problem definition, I can give a good strategy to ChatGPT. However, I did not have, so it had kept generating incomplete solution because the strategy was wrong. With wrong strategy, the solution becomes very complex and easily out of control.
I use chat GPT almost everyday and it is quite helpful. But if you ask questions like "In binary tree data structure, depth of tree is log base 2 (n) -1?", "So If I have 7 nodes?" and "so you mean root node is in depth 1?". More likely you would start getting contradicting answers. This is just one example.
@@user-jy4lk4uj5r I agree, but how knowledge of fundamentals could help you outrun the AI?
@@tedspens 1.'to know what to ask' you have to be an expert, 2. to know 'how to ask', you have to be an expert in GPT itself, 3. to know 'how to troubleshoot code', you have to leverage a more powerful GPT and it will tell you how.
What does it all have to do with 'knowledge of fundamentals'?
Great information.
Not to discredit the excellent advice and resources you've provided here but I think "talk out loud" may be the best single bit of all.
Hi,
Please make a straightforward course on coursera,
Thanksq
great one
Meaning, you need to get a CS degree.
Only if you're looking for a data entry position.
Or, even better, a PhD in mathematics.
Not at all
Meaning, you need to learn the stuff - which is not guaranteed with a degree, even if people have used it as a proxy for being knowledgeable and qualified.
Either do CS from reputable University or just don't do
Please watch Andrew Ng video about Agentic Loop, it will make traditional devs legacy even at GPT 4
True now is the time to learn the fundamentals and most importantly improve your soft skills
Within 6 months, as workflow improve, we will see the rise of AI agents affecting many knowledge workers
Subscribed 🎉
Exactly what I was thinking 🤐🤓🧐
Lol clever marketing. I keep reading this as Traversy Media
Ikrrr, I was thinking the same. Both are great though.
The CEO of Nvidia says devs should learn prompt engineering instead of coding. This made me sad but good thing I can write pseudocode.
despite the fact they will ask you about fundamentals during an interview, eventually you will be still doing html through the next super framework :) Hiring is extremely lazy this days...
now you're going to feel how concept artists felt when mid journey came out, there's no keeping up with AI
I like Packt too.
Traves thank ❤
Well, it looks like it's time for all of us software developers to retire, because the First AI Software Engineer Tool is here to replace us. I guess all those years of debugging and coffee addiction were for nothing. But hey, at least we had a good run, right? And let's not forget about those low code platforms that were supposed to turn everyone into a programmer overnight. I'm still waiting for my non-techy friend to deploy his first app. It's almost like every decade has its 'This will make humans obsolete' trend. Remember when automatic elevators were going to leave elevator operators jobless? Good times. Guess we'll just have to find solace in the fact that we're still here, and venture capital money still needs somewhere to go! 😂
Or just open a TH-cam channel like this guy and skip the rest of it.
Loooool he really referred to these ai tools as Terminators 😂😂😂
Prompt Engineering won't be a bad skills to have as well.
I simply wait for the VC funding for AI to dry out and the generative AIs to fail.
We did everything that the internet tells us to do and more from our learnings, self taught developers are too much work, too much risk and very little reward. They try to nose dive right from the get go even after being directly and indirectly told to ask questions and get clarifications/agreement. They then miscommunicate progress even after being told to be transparent and while others are. They make our other devs and lead devs focus on covering for them when we want them to focus on customers and products.
A lot of the folks might've done 1 hour of practice and will consider themselves a "self-taught" dev, that's why there so much negative connotation around it lol
AI fan boyz = unsubscribe. Good luck with your future endeavours.
Bye 😂👋🏼
Garbage video made with AI script + sponsored products and no real content.
There's no AI involved in this video.
Cancelled subscription.
This video will not age well...
We are doomed, because these un-trained, un-qualified people got employed and did "their best".
It was simply not good enough.
Like the LLM now.
Compare their output to mine in case of code quality and functionality: seems to be missing A LOT.
Because they learned from these unqualified people, from stackoverflow, from all these sources...
So, waiting for the collapse of this bubble.
The AI bubble?
true