Noteable Brings the Power of Computational Notebooks to ChatGPT

hi I'm Elijah notable I'm excited to introduce you to our new plugin for chat GPT that allows you to create a computational notebook using just a prompt so to get access to it you'll have to log into chat GPT of course this should look familiar to you if you've used it before you'll need chat GPT Plus for plug-in access plugins will be up there in your models you'll see plugins as an option and if you've got plugins then you'll install notable as the plugin and once you've got that in place then we can get to the meat of this which is right in a prompt now the first thing I do is I tell it which project at notable I'm using and then I'm going to tell it what I want it to do in this prompt which is I want it to load some data that I found this is from the Museum of Modern Art it's about the artists that they have in their collection and I'm going to create this Persona that says hey I'm a coordinator for a major Art Museum I want to think about what we're going to feature in the next year take this data analyze it for Trends and anomalies based on gender nationality historical period and remember that I'm not experienced with data analysis so explain to me what's going on while you're doing it let's see what happens when we run this we'll notice that it starts to use the notable plugin and if we go over and take a look at that notebook that it's created in this project that I told it to use we can see that it's already started to populate that notebook with a text cell and a code cell so for those of you who aren't familiar with notebooks a computational notebook is a collection of code and text along with other functionality allow you to connect to data sources leave comments visualize data do all sorts of great stuff and chatgpt is using our notebook to answer these questions that we've posed as this coordinator at a major Museum so what it's done is it's created what's known what's called a markdown cell to explain the analysis it's going to do it's loaded some of the libraries that are necessary to do this analysis and now it started to load the data and we can see the shape of the data it's going to describe that data and then when we come back to it after a while it's going to continue to build out this notebook it'll let you know where the notebook is it'll give you some sense of the work it's been doing and the amazing thing about this is rather than just seeing all of this in the chat GPT interface is at the end of this you have this artifact this computational notebook that's Interactive that you can develop on your own or with a team but once it was done with its analysis I thought it would be interesting to add a little bit more so I asked it to add a cell analyzing artists from countries that are majority Muslim if we look at the results of that it added some more code into the notebook you can see that we've got this fully fledged notebook that's describing gender describing nationality and it's describing the the periods everything that I asked to do and in fact it's showing the distribution of artists in a couple of different ways to answer that periodicity question but when I asked it to narrow down and only focus on artists who were from majority Muslim countries what it ended up doing was it ended up adding code to analyze the data based on a list of countries and that doesn't match up with the code because what's in the data is actually nationality so you'll see that it came back with the result of zero if we have to but we want it to look at nationality and not the actual country of origin so I simply asked it I said hey it looks like used countries instead of nationalities can you update the cell to use nationality and after I asked it to do that sure enough it updated the code it uh plotted the results and it gave me exactly what I was asking for and once again entire computational notebook full of analysis data visualization explanation and I haven't written a single line of text in that notebook so at the end to add a cell to give suggestions data driven suggestions for how a museum should approach its exhibitions in the following year and I also something I find really exciting I said add a bit of marketing Flair at the end I wanted to not just give me this try text but also um give me some sort of uh exciting call to action and so sure enough it gives me four data-driven suggestions for how to design a an exhibition and at the end of it it says the world apart is full of endless possibilities and at the moment we're just getting started it's hard to overstate how exciting this is we've created an entire data-driven document this computational notebook with text with code with analysis and visualization all from a single prompt when we've seen some errors we've returned to chat GPT to ask it to correct those errors or adjust it or even to focus in on a particular part of the data and it's responded by editing and appending to that computational notebook and now I'm at the point where if I want to share this with stakeholders if I wanted to share this with my peers to collaborate on it to develop it further I have it available to me to continue to edit and work on it throughout its lifetime and while that application is exciting I want to emphasize that this plugin isn't just for people who don't have experience creating notebooks or coding for folks who are Advanced Notebook users this will give you options that you might not have expected this little smooth the process of creating that initial project so that you can focus more on the parts of the notebook and the parts of the analysis and the parts of the data science that really need your skills example here a second example that focuses on learning how to use a python Library there's this word Cloud Library that I've always been very jealous of because I have a data visualization background but my background's in JavaScript and so I wanted to understand how could I use this library and what I love about it is that it allows you to use masking so for instance in their documentation it shows Alice in Wonderland and then the text from Alice in Wonderland has a word cloud within that so I asked it how do I use this Library speak to me as somebody who uses who understands JavaScript and doesn't understand python and sure enough it created this whole notebook The Notebook goes through the individual steps the interface in chat GPT explains to me uh what it's doing in the same language it puts it into the notebook and now I have this artifact from which I can work with so that I can use this library or Library similar to this or workflows similar to this I hope these two small examples have you as excited about this plugin as I am and I'd encourage you if you have access to chat GPT plugins and you've never worked with notebooks before to go and give the notable plug-in a spin give it a try to create a data-driven document based on the problems that you're solving in your everyday life and if you do use notebooks I think this is a great tool to make your workflow more efficient and help to generate more ideas we put a lot of work into building this plug-in and we're excited to see how you use it

Channel: Noteable

Views: 13,826

Keywords:

Id: lpJ6kQH2BgM

Length: 7min 59sec (479 seconds)

Published: Thu May 11 2023