Great video! Super interesting how creative you are with these APIs/OpenAI 😁 I had to pull transcripts for KZread videos recently as well, and I went through the exact same pain and also ended up with that library, haha 😅

@AdrianTwarog

6 ай бұрын

Thanks Simon!

@mrtbh

6 ай бұрын

My Comment Copied? lol.

You are an amazing person ! Thanks for sharing your knowledge

yeah spot on narrowing down the specific problem

Awesome idea and implementation !! thanks for sharing

It's amazing you can get so far in a single day. The software world has certainly changed recently! However it is clear that you need to be a rather skilled software engineer to make it happen.

@rd42537

6 ай бұрын

20y developer here - I had grown tired of developing until the last year or so but now have a new thirst for dev with the speed at which ideas can be developed now. I 'avoided' the whole node/reactive dev ecosystem due to preoccupations with backend technologies and apprehensions about how long it would take to 'get up to speed'. Read a few modern full stack react books, some (many) tutorials videos, cloning repos and good ol' trial and error with copilot/gpt by my side I'm simply blown away. It's an incredible community and v exciting time to be involved in the dev world!

@N7Tonik

4 ай бұрын

you absolutely dont need to be a skilled software engineer, anyone who truly wants to build some coding project can do it with the help of ai (at least the MVP)

@coldlyanalytical1351

4 ай бұрын

@@N7Tonik Hmm. I've done a fair bit of ChatGPT4 coding. The code it creates is usually fine .. except for a couple of subtle hard-to-spot but easy to fix errors. You need a fairly experienced developer to do that.

great content Adrian 😍

One day I will make videos loke yours, it takes so much time and i just love them all.

Great Vid thanks

really curious about your vscode theme. Anyone knows what theme that is?

This is so cool

This is cool... I'm looking at the code and it turns out you aren't using the db for anything? Correct me if I'm wrong but you don't need to generate or store an embedding since all you are doing is prompting chatgpt with the "video" json object which has the video details and subtitles. The embedding and astra can be completely removed from this and it would still work the same.

I want to do something similar. I'd like to create an automated workflow that would allow me to create a database of all the videos I give a like, then be able to ask those videos questions using AI. How could I set that up? (I am not a coder but I can usually follow instructions pretty well or just use no-code tools).

@todordonev

6 ай бұрын

I would suggest you self-host n8n(the langchain beta). From there its just like using Zapier(n8n is an opensource zapier alternative). There are videos in youtube on how to pull this off.

@AdrianTwarog

6 ай бұрын

I shared the code in the description if you want to take a look!

One month ago, I tried to the whisper api to get the video captions 😅, but it was taking too much credits

This is GOLDEN man! And love the KZread transcript idea... when can I buy it? :)

@AdrianTwarog

6 ай бұрын

Lool good point, ill give it away free this time!

@user-qg5bo9bd5x

4 ай бұрын

@@AdrianTwarog - where's the url link for it ?

It's freaking cool 😮

This video is inspiring! The process of building an AI startup idea in just 24 hours is impressive.

Hey awesome. Sad that KZread announced that AI feature 😢

It's been long man

Great video. However I had expected a video about building a product/business by almost 100% interacting with AI. In this case you are developing a traditional project which happens to use AI as components.

Why you’re not saying in the video that it’s sponsored by Astra? It’s only in the description. That’s unfair for your viewers.

@v1d300

6 ай бұрын

He does towards the tail end of the video. 9:56

@187Ehrenmann

6 ай бұрын

Cause in the end, it doesn't really matter

@SW-fh7he

6 ай бұрын

Indeed really unprofessional, very bad practice.

@TheHsystem

4 ай бұрын

Another hater, look the video till the end. The content matter not the sponsor.

This is textbook Retrieval Augmented Generation (RAG) use case, at first I wondered why you didnt look to something llamaindex / langchain, then I realized this whole thing was an ad. Come on

🎉🎉🎉🎉🎉🎉🎉

*promo sm* 🌸

Great video! Super interesting how creative you are with these APIs/OpenAI 😁 I had to pull transcripts for KZread videos recently as well, and I went through the exact same pain and also ended up with that library, haha 😅

@AdrianTwarog

6 ай бұрын

Haha yea that's how it is sometimes!

@Thahira-yc2ys

4 ай бұрын

This is the comment of Simon 😮

@mrtbh

4 ай бұрын

I commented before him. @@Thahira-yc2ys