I've discovered the perfect antidote to the ChatGPT Code Interpreter waitlist - the Noteable plugin! It's a game-changer for anyone looking to get ChatGPT Code Interpreter access. 

|| 📝 PROMPT LINKS ||

The Noteable plugin is a fantastic alternative to the Code Interpreter. You can check it out the chat I showed in the video (edit June 17th - that link is not working even though my chat still exists in my ChatGPT space so here is a link to the Noteable Notebook with the Master Plugin Prompt that shows you all it can do: https://chat.openai.com/share/8a2da604-67a1-4406-8705-0fd1395fd629. It's a great option if you're on the ChatGPT Code Interpreter waitlist. 

Here's the link to the Noteable notebook I used as well: https://app.noteable.io/p/35afe982-5023-49b2-abfd-18a2263218dd/Noteable%3A-Code-Interpreter-Alternative

|| 📚 VIDEO SUMMARY ||

I delve into the capabilities of the Noteable plugin, from optical character recognition to data analytics, and even basic video editing. It's a powerful tool that can do everything you can do in Code Interpreter and more. If you're looking for how to enable Code Interpreter ChatGPT or alternatives to the ChatGPT Code Interpreter, this video is a must-watch.

|| 🎥 RELATED VIDEOS ||

🔍 GPT-4 Got Upgraded by @ai-explained-  https://www.youtube.com/watch?v=O8GUH0_htRM

📘 How to get your data into ChatGPT by @noteable_io  https://www.youtube.com/watch?v=VV8dvip6N3s

|| 🙏 CREDITS ||

A big shout-out to:

📚 Noteable for their ChatGPT Integration Info: https://noteable.io/chatgpt-plugin-for-notebook

📖 ChatGPT Plugin Examples: https://app.noteable.io/p/efa0bae2-e7e2-49b1-8b10-6c73aff0bfa5/ChatGPT-Plugin-Examples

|| 📞 CONTACT LINKS ||

🐦 Tweet me your thoughts on the Noteable plugin: https://twitter.com/IversusAI

🎥 Join the AI discussion on my YouTube Channel: https://www.youtube.com/@IversusAI

#ChatGPT #CodeInterpreter #NoteablePlugin #OpenAI

|| 🎬 CHAPTERS - EXPLORING NOTEABLE PLUGIN ||
00:00 - Noteable, the Code Interpreter Alternative
01:26 - Noteable Plugin Installation
04:57 - 01 - Noteable Does QR Codes
06:00 - 02 - Noteable Does 3D Data Visualization
07:48 - 03 - Noteable Does  Optical Character Recognition (OCR)
09:23 - 04 - Noteable Does Data Analytics
10:25 - 05 - Noteable Does Video Editing
13:27 - 06 - Noteable Does Text To Speech
16:21 - 07 - Noteable Does Image Editing
17:57 - 08 - Noteable Does Image Recognition
21:57 - 09 - Noteable Does Text Scanning
24:35 - 10 - Noteable Does ASCII Art
26:49 - Noteable Benefit 01
27:12 - Noteable Benefit 02
27:42 - Noteable Benefit 03
28:54 - Noteable Benefit 04
29:37 - 11 - Bonus
30:29 - Noteable Benefit 05

Everyone is clamoring for the new ChatGPT code interpreter plugin and rightfully so because it promises to be amazing.

But there is 1 teeny tiny little problem.

Most of us plebs don't have it and are stuck on the wait list.

There is a plugin that is available right now that does almost everything I have seen Code Interpreter do.

And some things, arguably, it does even better.

Hey there, it is I vs AI and there is a lot of hubbub going on about Code Interpreter.

In fact, in this video, Philip at AI Explained shows off many of the things that the highly anticipated plugin by OpenAI can do.

Make QR codes, 3D data visualization, wow.

The Notable plugin can do all of that and more.

Like right now, let me show you that you do not have to be a data analyst.

Notable can be used in lots of creative ways.

I'm going to show you 10 prompts that you can use with Notable right now.

And 1 bonus prompt that can't be done with Code Interpreter.

Over here under GPT-4, you're gonna scroll down to plugins.

If you don't have any plugins enabled, you will see that.

Just click here and scroll all the way down to the bottom to the plugins store.

You can now search for plugins, thank God, because scrolling through 38 pages of plugins was doing my head in, and I just put in notebook here to get to Notable.

And you'll wanna click this little button to install.

Now, when you install it, you will be met with a request to set up two-factor authentication for your chat GPT account, which from a security standpoint, you should really do anyway.

Once you have two-factor authentication set up for your chat GPT account, you'll need to log into Notable.

I highly recommend using your Google account here.

And in fact, the team at Notable recommends this.

And inside those projects, there are notebooks.

These notebooks are what ChatGPT creates when you ask it to do a task.

And this notebook right here is the 1 in which I'm going to be running all of the prompts that I'm gonna be showing you today.

This notebook is actually set to view anyone.

So you'll be able to go into the notebook, click on anything in these items here and more things will be in here by the time I am done.

This is basically what is happening when ChatGPT is talking to Notable.

It's really fun to open up a notebook and watch ChatGPT fill it out.

It'll start out empty and then it will actually fill out and run the code right before your eyes.

So watching what ChatGPT does here in the notebook is really fascinating.

But 1 of the things that is getting people so excited about Code Interpreter is the ability to upload your own documents.

And as you can see here on the side, you can do that.

In fact, Down here under add and create, you can add files.

You can import up to 25 files at 1 time with a file size limit of 100 megabytes per file.

Now that you have installed the Notable plugin, you simply need to enable it from the plugin list.

This is just an overview of the Notable plugin with the available functions and parameters and the important thing I want you to note is this get data sources.

This is how you can upload documents directly from the chat GPT window as long as the link is to a document type that Notable can read like an image or a .csv or a Google Sheets then Notable can read the information in that data source that you're linking it to, and then install that data source right directly into your plugin.

The first step is to get or set your default project.

If you don't do this, you will waste 1 of your precious 25 per every 3 hour prompts and I don't want you to do that.

So the first thing that you do before you ask it to do anything is to simply ask it to set your default project like this.

Set default project to and whatever the link is.

Now let's move on to the first thing that Notable can do that Code Interpreter can do as well and that is QR code.

Did you know for example it can generate QR codes?

I said create a QR code that I can scan with my phone to reach the following URL and lo and behold it creates it and yes it does work.

Maybe I'm easily impressed but I think that's pretty amazing.

The prompt for the QR code is create a new notebook called QR code YouTube install the necessary QR code libraries and create a 500 by 500 pixel QR code on a white background with the data matrix, the part that's usually set to black, set to a different color, and linking to YouTube.

And I ask it to do that again in a new notebook and link it to Twitter for iVersusAI and YouTube channel for iVersusAI.

And now it's gonna create the QR code for Twitter.

And you can see it beginning to run the code and producing the output which will then show up right here in the chat gpt window.

And look just for fun I'm going to go into the data.

I don't know how helpful it is, but I think that's just beautiful and crazy.

I'm asking chat GPT to analyze the following Game of Thrones data set, which predicts which character will die.

I'm asking it to give me some insights about the complicated political landscape and I am requesting it to ensure that it suggests diagrams that the plotly Python library can create and then I'm telling it where the data set is.

It will tell you about the data set and then suggest some 3D diagrams that can be done.

It's correcting its mistakes as it goes along and gives you some ideas for other types of 3D diagrams that you can use.

That leads us to the second step of this 3D diagram prompt.

I am now asking it to take the same data set and asking it to create a 3D scatter plot of age, popularity, and number of dead relations on 3 axes to see if there's a relationship between those variables.

Once you get a 3D scatterplot, Chatshubpt will not be able to show that in its window.

So head on over here to the cell menu, click on Embed output.

I just threw it here into this HTML viewer and clicked on Run.

Here, you can see that it has given us indeed a 3D scatterplot.

Yeah, dead relations 6 and dead relations 0 based on age and popularity.

So this is a character here and another character here, outliers, but that is just beautiful and is done with Notable.

Next, did you know that Code Interpreter can do optical character recognition?

I screenshotted this text from a New York Times article, I think it was, and I asked OCR the text in this image and write a poem in Danish about it.

For this 1, I grabbed this screenshot from Reddit with thanks to Redditor Jagonew.

This is the system prompt for Code Interpreter.

And this is really important because it shows you exactly the instructions that are being sent to Code interpreter.

The tool that Code Interpreter is using is simply Python, which is exactly what we are doing using Notable.

Therefore, in theory, you should be able to do everything you can do in Code Interpreter in Notable.

Create a new notebook called OCR image and install EasyOCR and other necessary libraries and modules and use it on this image.

1 thing to note here is that I had a lot of trouble with this 1 because chat.gpt kept wanting to use the tesseract library which cannot be used in the notebook environment that notable have.

But I did find that easy OCR does work, though it is a bit slow.

It took a few steps, but we did get the text from our image.

And if you weren't that impressed already, here's where it gets fairly game changing.

You can get it to do the data analytics, not just the visualizations.

find 5 unexpected non-obvious insights from this data source about teen actors.

After some considerable time, it begins to output some really nice unexpected insights.

Age difference distribution, age difference by gender, age difference versus actor popularity, age difference by number of love interests.

And then it concludes with a maximum age difference.

ChatGBT has been doing pretty well so far, but this next task really caused it to fall on its face.

And to be honest, it's hard to tell whose fault it was.

Now, there is a limit to what it can do, but it can do some basic video editing if you ask it.

For example, I uploaded a short file and asked it to rotate the file 180 degrees and it was able to do it.

It is possible to do simple video editing with Python But this 1 really gave me some problems.

I first asked it to trim 1 minute from the beginning of a video, to rotate that video 180 degrees, and change its tone to grayscale.

This was a lot harder, and I wasted a lot of prompts.

Eventually, I realized that the task might need to be taken into smaller bytes, so I simply asked it to trim 1 minute from the beginning of a small three-minute clip.

To complete this task, it does install MoviePi, And I have found that whenever there is a lot to install, chat-gpt can often get hung up or time out.

But then it again tries to create a file or find the file in a path that does not exist.

And it just made this path up despite me telling it that it is located in the default project.

So I tell it it's in the default project and that there is no temp folder and it gets back to work.

Now I've only seen code interpreter videos where the prompt has already been run and the output has been made.

And I don't see all of the back and forth that may have happened behind the scenes.

So I want to be really honest with you and show you that there is a lot of back and forth when working with chat GPT and notable, even if you are extremely clear in your prompt, it will frequently just make up a folder and then tell you there's nothing in it because that folder doesn't even exist.

And it now tells me that the video has been successfully trimmed and is named Trim Video MP4.

But when I look over here in the project folder, there is no trimmed video despite saying that the command is successful.

So this is 1 where I'm going to have to give it to Code Interpreter, because for some reason, this was simply not working, at least with my limited Python skills, in the notable environment.

Just to be cheeky, I also asked, can you make it black and white?

Oh, and I also asked to add music, but it couldn't add music.

In AI Explains video, he said that Code Interpreter could not add a music track to a video.

And having had the beating that I've just gone through, I didn't feel brave enough to try and get ChatGPT to add audio or a music track to a video that it can't even render or doesn't even show up in the notebook as I just showed you.

But according to this website, thepythoncode.com, it can be done.

So if you're feeling brave, go ahead and try

Next, did you know that GPT-4 with Code Interpreter can do text to speech?

As you can see, I asked turn this entire prompt starting from the beginning into a text to speech file.

Now quite a few times it denied it had the ability to do this, but eventually I got it to work.

This 1 I had a lot of problems with, including it not doing the complete text, for example.

It would stop at random places and stop like here, for example, and only do this part.

Sometimes the same problem that I had with the video where it would tell me that the TTS MP3 file existed and that it had been created and then it wasn't in the notebook.

The text has been successfully converted to speech and saved as ttsmp3.

And here it is and we can just give it a listen.

You'll just click download and it'll open up and immediately start playing in this tab.

Your chat gpt a large language model trained by OpenAI Knowledge Cutoff

current date, the 30th of May 2023 math rendering.

ChatGPT should render math expressions using latex within.

As you can see, it does work and it sounds OK, but there is a possibly better option and certainly easier.

I have featured the speech key plugin before on this channel and it's very handy.

You just simply ask it to convert the following to speech.

And I asked it to use a male British voice.

You do need to create an account, but once you do so, you'll simply be able to click on it.

You are Chet GPT, a large language model trained by OpenAI Knowledge Cutoff, 2021 to

ChatGPT should render math expressions using latex.

I'm going to go out on a limb and say that that is absolutely not a British voice at all.

SpeechKey does have a parameter where you can get a list of available artificial voices and it's quite a long list.

But in my testing it seems to pay absolutely no attention to any kind of instruction.

Maybe I could try using a unique identifier to get the voice, but just telling it to use a male British voice seems to make no difference or female.

However, this method is much faster than working with Notable and I didn't even try to get Notable to give me different voices for the text to speech.

I was just grateful when I got the complete text converted.

I created this image in Midjourney version 5 and then here's what I asked.

I said use OpenCV to select the foreground of this image And look what it did.

Now I know it's not perfect, but it's nevertheless impressive all within the window of ChachiBT.

Create a Python notebook and install REMBG for remove background.

In order to remove the completely white background of sticker.jpg which is this adorable little kitty here and I chose this from mid journey because it had such a simple background and having been so thoroughly defeated from video editing and text to speech, I wanted to give Python something very easy to complete, at least Hopefully.

The result I asked for is to have a chat GPT output it as a transparent PNG.

The task took long enough that it did time out.

But that's 1 of the cool things is that if it times out in chat GPT, that doesn't mean that it isn't completed and you could just head on over to your project and see if it's here in the notebook.

Now Python is perhaps not the greatest with background removal as you can see.

There's some sections here around the whiskers but I'd count that as a win though I'd be very curious to see if code interpreters Python can do a better job.

For the moment though before anyone gets too carried away it does still hallucinate quite a lot.

So I uploaded this image and I asked it questions about it and it answered and I was like, wow, it can do image recognition.

It said, this image appears to be a digital painting of a humanoid figure at a desk with a rather complex background.

I was initially amazed until I realized that it probably got that from the file name.

This is 1 that does indeed work with Python in Notable, but there may be an easier, faster way for this 1 as well too, at least until Code Interpreter comes along.

This image recognition task took about 5 to 7 minutes of going back and forth with this.

You can see that it ran quite a lot here.

And it told me all the libraries it needed.

And the problem with this task and notable is that image AI here is a library that's required that also has dependencies that have to be installed.

And this is where chat GPT was getting stuck.

Now, I have done this before in my testing before recording this video and I'm not going to run this again because I'll let you know that the outcome was not that interesting.

Basically what you get is after installing all of this and then you get a module error down here at the bottom.

It will give you image recognition based on probabilities.

It will predict what's in the image and it will give you 5 results.

But they are not full sentences or complete guesses.

This is the image that I was inputting to use for this.

And the number 1 word that I continued to get back in my testing was castle.

That's really the only recognizable word that I got.

Every once in a while I would get an odd word that had nothing to do with the image at all.

Again, is this a Python problem or a Notable problem or a ChatGPT problem?

But I do know another way to give you image recognition that works quite a bit better than this.

This is the Scene Explain plugin and it does a better job, But it's still not perfect.

This is the explanation it gave us, which is this is a watercolor painting of a picturesque Cotswolds village scene, which is excellent.

The idyllic country road winds through lush green fields with charming cottages and houses interspersed.

That all sounds correct, but let's try 1 more.

This is an image I did of a wizard just sitting at a table reading a book by the light of his wand.

A white bearded wizard sits at a wooden table in a dimly lit candle-filled room engrossed in a thick book.

A glowing purple flamed pot sits on the table.

And here's where it goes completely off the rails.

A young woman stands behind a bookshelf holding a potion bottle.

The bookshelf contains various volumes, including Harry Potter and the Goblet of Fire, a painting of a candle hangs in the foreground, and a shiny blue glass ball hangs from a chain.

Twisted trees and a nebulous sky in the background suggest a world of magic and witchcraft.

So basically from the second sentence on, it made that entire thing up more or less.

And what I wondered was what is the biggest document I could upload to get it to analyze.

The longest book that I've ever read is Anna Karenina.

I think it's about a thousand pages long and I pasted it into a word doc and it's about 340, 000 words.

I uploaded it And then I asked, as you can see, find all mentions of England, analyze them to discover the tone in which the country is perceived in the book.

I decided to use the same text, Anna Karenina, as Philip used in his AI Explained video.

But I planted 2 sentences of a well-known tongue twister.

How much wood could a woodchuck chuck if a woodchuck could chuck wood?

As much wood as a woodchuck could chuck if a woodchuck could chuck wood.

And I placed them at different places in the text which is indeed massive at 39, 831 lines here and notable.

I asked ChatGPT to discover the 2 tongue twister sentences of Outwood that I have inserted into 2 different places in the text that are in Congress and do not belong.

And once those 2 sentences are found, print out the sentence that appears directly before them in the text so that I know you have actually scanned the text.

I had to tell it again that there's no temp directory, that the file is in the default folder because it kept insisting on putting the file in a directory that did not exist.

It gave me this output here and both of them are incorrect.

But in the text scanning notebook, it does find the sentence.

And this is indeed the sentence that was before.

But the I believe you, I believe you, what we were saying.

And but if only you knew how wretched I am was correct.

And the second sentence was as much wood as a woodchuck could chuck.

I feel like the reason why ChatGPT could not find these sentences is because of a problem that occurs with all plugins and that is the length of context returned.

This is a lot of context And I've noticed that it causes chat GPT to be unable to complete the task, whether that be a transcript for a long YouTube video or something very long like Anna Karenina.

I was surprised by the number of comments asking me to get it to do ASCII art.

Now you may remember from the last video that I got it to analyze this image.

And yes, I asked it to turn it into ASCII art and here is what it came up with.

This 1 I had a lot of trouble but it's actually good that I did.

I did get an ASCII image but as you can see I kept getting problems where it was trying over and over again to send the same exact code and it would return an index error string index out of range.

But this leads to 1 important difference between notable and code interpreter and that is that notable's code can be fixed.

Here is the code and this area right here, this code block, was what was causing the error.

See all I versus AI transcripts on Youtube

This Plugin Is ChatGPT's Ultimate Power Tool (Code Interpreter Can't Do This)