ChatGPT Tutorial - Use ChatGPT for DevOps tasks to 10x Your Productivity

ChatGPT Tutorial for DevOps - What is ChatGPT? How to use ChatGPT? My opinion on whether ChatGPT will replace engineers

#chatgpt #techworldwithnana

► This video is sponsored by Firefly 🙌🏼

I'm sure you have all heard of ChatGPT by now. It has become a buzzword within days of its release and professionals in all fields, especially in high skilled areas like lawyers, doctors, engineers are questioning whether such AI can actually replace them and work.

So in this video I want to talk about what ChatGPT is and how it even popped up, talk a bit about the organization behind GPT called "OpenAI", which has already created many other machine learning models besides Chat GPT and also explain technically about all that.
And then we'll dive in and actually put ChatGPT to use for some DevOps related tasks.
I really want to see how it can help in generating configuration code for building DevOps processes or different parts of those processes and how well it knows different DevOps technologies, but not just some shallow examples or boilerplate code that I can get from official documentation, but instead also try more fine-tuning and small optimizations in that configuration code.

We're also going to check out an open source command line tool that is built on top of ChatGPT and was specifically created for engineers to generate infrastructure as code templates and more and finally we'll talk about the impact of ChatGPT, the quality and usefulness of such a tool for engineers and whether it will really replace the engineers and to what extent you should be concerned.

▬▬▬▬▬▬ T I M E S T A M P S ⏰ ▬▬▬▬▬▬
00:00 - Intro and Overview
01:39 - What is ChatGPT, Who developed ChatGPT
06:45 - Sign Up on ChatGPT
22:13 - Create Kubernetes manifest file using ChatGPT
35:06 - Create CI/CD pipeline code using ChatGPT
50:06 - Convert Jenkinsfile into GitLab CI config file
53:53 - Tools built on top of OpenAI's API
55:01 - AIaC demo - CLI tool for DevOps
01:01:00 - My opinion on ChatGPT & whether ChatGPT will replace engineers

▬▬▬▬▬▬ Want to learn more? 🚀 ▬▬▬▬▬▬

▬▬▬▬▬▬ Connect with me 👋 ▬▬▬▬▬▬

▬▬▬▬▬▬ Courses & Ebooks & Bootcamp 🚀 ▬▬▬▬▬▬

[TechWorldwithNana]

Hope this video was helpful! What are your thoughts on AI and ChatGPT specifically? Did you already use it at work? 🤔

[tirthjoshi]

This month I joined company as junior DevOps Engineer and I'm using this chatGPT to learn and complete the tasks, Thanks for making detailed video on this very informative 😀

[fabianmadueke]

I asked it to solve some critical but simple mathematical problem like the difference between -3^2 and (-3)^2, it got the answer wrong by giving me the answer as 9 in both cases, instead of -9 & 9 respectively, this is a trick question and I told it that the answer was wrong and it apologized and resolved it and got the right answer.

I have also use it in programming to weigh it capabilities, just like you did with devops, it's pretty good in give base knowledge and configuration to work with, which is otherwise, things you would have search on the internet as you work, but more quickly and tailored to your need, unlike an ordinary search engine.

Here is my observations:
1) ChatGPT is a pretty awesome tool, that will be added to our range of tools to work faster and better.
2) It can help you learn more quickly, by fine tuning your request within the given context, better than an ordinary search engine like Google.
3) But the result is not entirely or 100% reliable to use at a go, it will require verification and assurance to accept that.
4) for an effective verification and acceptance, you need to know what you want and what you intend to expect, to know when it's going wrong.
5) For knowing the expected outcome, you need to be very knowledgeable in the field that you want to apply it on, so in a not shell it does not replace the expert, especially when the field gets very mathematical or technical, but can make you work better and faster.
6) it's an excellent broiler plate generator and provide a single communication point, within the context, without digressing very much.

In conclusion, I will say that it's a good starter tool for learning or for expert working on something, but not a replacement.

And also being that it's core machine learning algorithm, is mining on existing pre-trained information source like the web or any info source, the ability to generate a new idea is entirely is limited, so it's a good helper or assistance.

[BrunoBuenoLinux]

Hi Nana,
I'm an SRE from Brazil, and I use ChatGPT many, many times in my daily work.
Actually, I use AI more than Google search nowadays. It helps me work faster and more dynamically, whether solving problems or creating solutions like scripts or microservices.
AI has become incredibly useful for me as an SRE.

Your vision about AI and its future use is very interesting.
I share a similar perspective, and I’m curious about the future of this tool—it’s truly disruptive.

Thank you for your video and for sharing your insights about AI.

I love your videos, Nana! I believe you have a big audience here in Brazil.

Thanks, and big hugs from Brazil!

[Sankaritarina]

We are currently missing DevOps on our team so we fed chatGPT our docker-compose, nginx config etc. He understood that we have a rails backend and a Nextjs frontend. I then asked him how to setup the proxy pass for websockets for Nextjs. He gave the precise code. Saved me a lot of time, especially when DevOps is not my field of expertise.

[AdmVrln]

Nana you are an amazing creator and engineer.

For myself, I started in the 1990s in IT. I've been an AS/400 system admin, PC / AS/400 PC support, NT admin, Windows and Novell server admin, IT manager, etc. In all cases I always knew I wasn't adept, or even having the aptitude in the code, or scripting areas. I watched powershell land and I never took to it as some do, in the same way I didn't take to bash. Not that I can't hack basic things together - but simply know its not an area of strength *or interest*. None of this stopped me being a reasonble tech/admin/manager, but its certainly a road block to being near or in that upper 5% peer area which I clearly understand.

ChatGPT for me, closes a gap that exists. Brutally, I now use it akin to having my own Powersheller (Its an example) on staff. Same applies for delving into areas where I know what I want, but don't actually have the depth background skill. Need to knock up a docker config file for building a docker on debian, for a specific docker - I can get ChatGPT to gen the config. In all cases, its not that I could not put my head in a book for hours or days and get it done, but more that I don't want to do that. I've never really wanted to do that.

What this is changing - and this is only my opinion - is its taking computers in the old style - where a human has to gather up all the variables for the computation, and the computer carries out the last step, to converting the system into - 'I want these variables gathered, I have these goals, and need this output' and hit go - and the AI generates the rounded answer.

I feel that if we had many of the historical computing greats sat here today, this would be akin to a eurika! moment where their original dreams are made manifest into computers they dreamed of.

There has always been a very high barrier - sadly not realised by people who have the greater aptitude and background and talent where use of computers was limited in various ways - this tech feels like an empowering moment where computers take on an entirely new level of usable capability. I was around during the original shift when google landed. It wasn't overnight that we gave up on the huge chunky books, but it was a very fast shift to the internet once the tech hit a certain point. I feel that now with AI. GPT3 >>> events are I think going to change everything. And it may I think be an equal change to that of the internet in size and scale.

Thank you for your content and work.

[janismeyer]

Thank you very much for this video. One thing, I really liked was that you not only got impressed by the results but also rigorously assessed them pointing to its strength and weaknesses. Unfortunately, I too often miss this crucial step: People just start getting impressed or even intimated by one or two responses to their prompts but do not really reflect the returned result. Assessing AI this way will leave humans being „overfitted“ to AI capabilities: Thinking AI is able to do what it actually cannot do. Now, because the model might and will return wrong answers, or at least results that have to be revised, the user must have to some extent a deeper understanding of the topic. The user does not have to know every detail but the user will have to know, what she/he is looking for.
The way you taught people using this service is actually the one a consider myself to be the right way. Thank you very much again!

[chandywerks]

This is great. This can take the chore out of the parts of coding where you need to dig through the documentation to figure out all the the boiler plate and give us more time to spend doing the more interesting things.

[attilapinter]

ChatGPT is nice and all, but can't really replace the actual knowledge one requires to do things with confidence. Or at least not yet. It is sure helpful, but it can get a lot of things wrong so you need to understand what is happening before implementing what it suggests. Feels like the classic copy from Stackoverflow to prod mechanism, but on steroids. Some of the comments are terrifying.

[ncoles]

Hi Nana. Before I watched your video I thought about the possible impact AI (and ChatGPT) would have on software engineers. I came to the conclusion that far from reducing engineer numbers, it could well increase them. Using AI, engineers will become far more productive which will lead to more products and services being produced growing the software and tech market even more. I see AI as a real growth factor for engineers as long as they continue to learn and use AI tools effectively.

[VasLuPo]

Words that convinced me to not avoid AI stuff:
"One thing that hasn't changed (yet) is that AI still needs to be used by humans. And that's the whole point: using AI, humans can be more productive.

So, the fact that AI can do certain tasks better than humans, does not necessarily mean that you need less engineers.
It means that
- the engineers will be more efficient,
- so, the same number of engineers can do more stuff faster,
- so, it accelerates the growth and speed of development,
which obviously every company wants to have."

Thank you very much for the explanation!

[j]

thank you nana. as a junior engineer, yes chatgpt helps me in researching and referencing code to my client's needs. however, not all designs are the same and every client requirement is unique, and my job is to come up with designs specific and tailor-fit for the client. you are right, chatgpt can help but it still needs info to better its answers, and dev engrs continuously create and develop, where this can be used as a tool to better how we code.

[pbdinesh]

The first thing i searched when I'm using chatgpt was reg devops pipelines docker file etc..finally people are on it now.
Thanks Nana. keep making it as a series with these kinda videos.

[kjmrch]

When ChatGPT appeared, I immediately thought of whether Nana had a video tutorial.
Sure enough, the first time to update, your video is great, thank you

[thegreenpickel]

I used chatGPT to help me understand writing for loops and output data to a max7219 multiplexed 8 digit 7 segment display. The speed chatGPT operates is amazing.
Edit:
Finished watching your great tutorial and installed the API on a headless pi4 I use to program the pi pico. AIaC currently requires go version 1.19 while APT repositories use version 1.15. Really neat, thanks.

[Claire-defs]

I find ChatGPT very useful. I heard somewhere that ChatGPT will replace StackOverflow in 2023. After using it for a bit it seems like that might actually be the case.

[ceser]

Hello Nana. In the last two days I watched many of your series and I have to thank you for providing free ultra high quality content! your video series changed my life (at my computer and servers lab at home) 🙂THANK YOU

[servicekingbd]

Thanks for promoting me.I'm stronger day by day.stay with chatGPT.

[Tony-dprl]

Everyone knew this topic was coming soon to DevOps channels all over YouTube. :)

[striker]

Over time, the accuracy of output from ChatGPT will increase as it crowdsources and usurps the knowledge base. This will relieve the engineers from mundane work and allow time for engineers to come up with new solutions. As a hardware engineer, I tried verilog/vhdl code of simple logic to DSP filter convolution (which it did well) to systems designs (it gave up but has potential). Certainly, it will have a great positive impact on many forms of engineering work.