ChatGPT + Noteable (Jupyter) = Mind-blowing!

preview_player
Показать описание
Do you use Python, Pandas, and Seaborn to collect, analyze, and plot data? Then you'll be amazed by what ChatGPT can do, when using ChatGPT+, GPT-4 model, and the plugin for Noteable's version of Jupyter notebooks.

In this video, I show you how I got things set up for using Noteable (and world headlines), then put together a query, You'll see how it goes well, where it goes wrong, and what sort of code I can create using just English-language descriptions of my plans. And I show you what's happening behind the scenes, as we see the JSON being written.

This is all brand new and exciting, and I hope that you'll post suggestions and ideas in the comments for how we can take this even further!

Рекомендации по теме
Комментарии
Автор

This is mindblowing! Just Imagine what we can do in 5 years from now, this will change everything.

matattz
Автор

You see a lot of YouTube videos with titles these days re: AI using the terms "mind-blowing!" and "crazy!" among other superlatives. And often they really are. Just as often they aren't. This is one that lives up to the hype.

larrysankey
Автор

Ok, my mind is blown!!! 😱 I was having a little trouble keeping up with parts if it, but I'll go back and rewatch it. Thanks.

mauisam
Автор

We're glad you found it useful and we look forward to seeing what folks can do with it!

noteable_io
Автор

This is phenomenal. The exploration and learning are getting shrunk at an amazing pace.
Good news we can work more on the real problems than plumbing.

Thanks for the video absolutely phenomenal

TinyMagiq
Автор

Amazing channel, subbed. You deserve more recognition. I have been debating getting openAI, but after I saw there is a python API, I should just do it. I am a Mechanical Engineer excel guru trying to write scripts to make my workflows make dataframes and end up with visualization dashboards based on sketchy data from our company BOM. After I saw what plotly express and seaborn can do with only a few lines of code, it opened my eyes. I sometimes have a rough time getting excel to make good pivotcharts. My macro recording was a mess trying to make 😅. I was blown away just this week, just trying the Microsoft copilot on one of my code blocks, and it knew exactly what I wanted to do(despite the generalized prompt) and solved my problem. Only thing I don't like as a beginner is that over 50% of my time is wasted doing debugging, but I'll get better thanks to videos like these(and stackoverflow). ☺

superspak
Автор

Great intro to both Notable and GPT+ capabilities! Thanks. Your enthusiastic approach is contagious. We'll try to follow up on this.

joaquingianantonio
Автор

You can in upper right corner of your message go to edit. After changing the content, you press submit again. That way you don't need to copy and paste your prompt to a new message to ChatGPT.

sfsft
Автор

This is unbelievable! It's solves a number of issues in an app I'm building using OpenAI. I hope it works as well via the API. Does anyone have experience using the API with these features? I found this video very useful and hope you do another one featuring how to use the API this way. That could be a true game-changer!

Pthaloskies
Автор

Cool thing you can do -- pass ChatGPT the whole Project URL instead of the Project ID. ChatGPT can figure out the project ID directly from that!

KyleKelleyR
Автор

I like your enthusiasm, this is good content. It gets too complex really quickly though to eventually verify all code lines. You have to trust that there are no mistakes in the handling of data...

Master_of_Chess_Shorts
Автор

Impressive. I'll I've done so far with ChatGPT is write a buisness plan and precipe a couple of books for their introductions. This is a lot more fun to watch.

richpeacock
Автор

I asked GPT to write code to get the data from other sites and create an CSV inside the notebook or access it directly. Works great.

robbe
Автор

Ignore the haters, you were absolutely great, honest & natural in the video, GPT4 with Plugins is like magic wand, you need to know the right spell (prompt) and pronounce (write) it right. Otherwise it can backfire 😢. Thanks for intro to Notable, looks good as alternative Jupyter notebook product, will try right away. Thanks again, can’t wait to watch your next video.

pankymathur
Автор

Hi, this is your first video I've seen, and I just subbed! I've seen videos listed about Notable, but had no idea it was related to Jupyter 😂

bc
Автор

Super fantastic mash-up possibilities! Thanks for the cool vid!

loslosbaby
Автор

Question-01: There is limit to Token size sent to chatGPT, does that also account of size of the file downloaded (or considered) for the Analysis
For example in your DEMO your CSV file was 2GB or 200+ MB, will the size of those files impact in ChapGPT limits set for your account?

Question-02: You mentioned 4000 odd tokens allowed for upto 3 hours, where can I find how-much is available since last use

Question-03: Tokens if NOT used in 3 hours window, will it get carried-forward to next 3 hours window?
I do not think so but still curious

Scooterboy_and_others
Автор

Nice video..🎉 I think what you did is very helpful in literature review / research. I m sure you can do video before anyone else how to use these tools to automate using csv / Excel, then send it to notable, then apply the magic og Gpt by analyzing data . This will revolutionize the field . Please soon

khaledalazab
Автор

As a data science instructor, I find this demo simultaneously fascinating and terrifying.

chyldstudios
Автор

Wow, this is a game changer. Does this mean my python + pandas knowledge is already somewhat irrelevant? I guess it at least helps to be able to verify that there aren't any major mistakes in the code. Ultimately, maybe I should spend more time learning AI prompting rather than coding...

AndrewT