- 2 days ago
Category
🦄
CreativityTranscript
00:00What's going on, everybody? Welcome back to another video. Today, I'm going to tell you
00:04my favorite ways to use ChatGPT for data analysis, and then we're going to actually analyze some data.
00:15Now, ChatGPT has taken the world by storm. It is pretty stinking impressive. I've been using
00:20myself and I've been pretty blown away with some of the results. If you have not tried it yet,
00:25I highly recommend trying it yourself. What we're going to be doing today is I'm going to walk you
00:29through four of the top things that I've been using ChatGPT for, for actual data analysis.
00:34And then we're going to take a small data set and we're just going to ask it some questions and we're
00:38going to prompt it to do a few things and see how it responds. I'm going to tell you right at the
00:42beginning, it's pretty incredible. And this is just ChatGPT 3.5. In future variations, in future
00:49iterations, when they make improvements, I probably will be making more videos on a lot of the new
00:53functionalities and the enhanced abilities that ChatGPT has. So let's take a look at my favorite
00:59ways that I've been using ChatGPT as a data analyst. All right, so I've got my computer in
01:03front of me. If you have not seen this before, if you don't know what this is, this is the interface
01:07for using ChatGPT. So on this left-hand side, we have these discussions. These are previous ones
01:13that I've been doing, just testing things out. I've gotten rid of a lot of them because I've done
01:17this a lot. And this is the actual interface. Now, this is what it shows when it's just prompting you,
01:22hey, here's some ideas that you can do. And then down here is where you're going to actually
01:25write what you're doing. So we're going to be feeding ChatGPT something called a prompt,
01:30basically telling it or asking it to do something. So one of the first use cases for ChatGPT
01:34is that it can explain code really well. So I'm going to ask it to explain some Python code and
01:40we'll see what it says. So let's go ahead and say, explain this Python code. I'm going to come right
01:46up here. This is just my Jupyter notebooks. I'm literally just going to take this. This is from
01:50a project that I created, just a unit of measurement converter. And I haven't given it
01:56any information besides that. It's literally just the code. And it's going to have to take the string
02:03and the code that I've actually written and try to explain it to me. So all we're going to do is
02:07we're going to click enter. And right now I'm working with the completely free version. So everything
02:13that it's about to generate is based off the free version. I'm not paying for any of this.
02:17I don't really believe in paying, but let's actually take a look at what it's explaining.
02:22So it's, I'm not going to read all of it, but it says this code is a simple unit converter
02:26converting between inches, feet, and yards. It says the user is prompted to enter the unit of
02:31measurement to convert. And the code converts the user's input. It says the code uses lower function
02:36to convert the user's input to lowercase. So the inputs are case insensitive. And that's 100% right.
02:42If we go back to the code, it's literally asking the user to input some unit of measurement,
02:47what they want it to convert it to. And then this logic just converts it for it. It's fairly simple.
02:52May be a lot of code if you don't know Python, but it's actually fairly simple. And ChatGPT does
02:58an extraordinary job just explaining it really in layman terms. Now I've been using ChatGPT to explain
03:03code to me because some people are sending me code and I just don't fully understand it a hundred
03:08percent. I'm like, okay, I get the gist of it. But when I ask it to explain it, sometimes it'll go
03:12in even more detail. And it's really, really useful, especially if you're first starting out and you're
03:17not fully understanding some code or a script that you're getting or trying to write. Now that's
03:22just Python code. You could use it for SQL or for R for any other programming language, or even
03:27something like a Excel function. It really could be whatever you want explained. You just put it in
03:32here and it'll explain it to you. Now, the next thing that I can do is actually generate code.
03:36So I'm going to ask it to generate some Python code as well as some SQL code, and we'll see what it
03:41gives us. I'm going to come right down here to the prompt and I'm going to say, write a Python script
03:48to scrape data, oops, data from Twitter without my IP address being blocked. And let's see what it
03:58comes up with. So the generating code is actually really amazing because oftentimes it'll give you
04:05somewhat boilerplate options or boilerplate responses, but it also gives some context here
04:11because what it's explaining right now is that you really shouldn't do that. Or, you know, here's
04:16some reasons why you may not want to do this. And now it's going to actually generate the text. Now
04:21it even has up here, a copy code. So you can just come in here, copy this code, and you're good to go.
04:27Right here, you can see your consumer key, your consumer secret. This is using the Twitter API. So
04:32it's saying you need to use the Twitter API, but I could say not using the Twitter API, and it would
04:38generate an entire other response without using the Twitter API. So it's pretty incredible. And it's
04:44still going. This is a lot of information. And if you don't use any APIs, if you're not used to using
04:50things like that, or if you don't know Python at all, this is fairly straightforward. It's fairly
04:56simple. This isn't, I'm going to scroll up real quickly. This isn't anything like super advanced.
05:01This is something that I could probably Google and get pretty quickly, but it's generating it
05:05based off of my request that I don't want my IP address to be blocked. So it's giving me some
05:10options to actually do that, which I think is just fantastic. That's really, really impressive.
05:15And it even explains it at the bottom, how it helps you avoid this IP address being blocked from
05:20Twitter. So again, just pretty impressive stuff, but let's go ahead and try it with SQL code.
05:25So we're going to come down here and say, write a MySQL stored procedure that automatically
05:34imports data from a CSV at a specific path. And let's see how we write this. Now I've worked a lot
05:45with this stuff, with data collection, importing data automatically. When a CSV gets dropped into a
05:51file, you want to automatically take that data into the database. Super, super common. So let's see
05:57how it writes this. Now I just got a network error. Basically chat GPT is overwhelmed and it couldn't
06:02finish my response. This is happening a lot with chat GPT. I just want to kind of explain this.
06:08They now have a paid version that you can pay for. It's like 20 bucks a month or the more advanced
06:12version, which is $42 a month. This won't happen if you pay for it. But if you don't, which most people
06:18don't want to do, you're going to encounter these types of errors. So what I'm just going to do is
06:22I'm going to refresh. It even says we're experiencing exceptionally high demand. That's
06:27completely true. But I don't want to pay for it. So I'm going to show you and hopefully it'll generate
06:33the entire thing without me having to pay for this because I really don't want to. Let's see what it's
06:37writing. So it's creating the stored procedure and it's using this import CSV. It's giving our file
06:45path and telling it to put it in this, um, in this table right here. And so it's literally looking
06:51for a CSV and then it's going to actually write the code to load that CSV. So that's what it's doing
06:57right here. And you could use this, you know, you have to have certain settings in MySQL. Don't just
07:02copy this and paste it. You have to input some information. You have to have certain settings
07:06set up in MySQL as well, but then it just kind of explains it a little bit for you. It's
07:12pretty great. I've used some variation of this and each time I run this, it gives me a different
07:18version. Sometimes it's more dynamic. Sometimes it's more hard coded. So actually knowing this
07:24skill when it's generating code, actually knowing SQL is really important. I wanted to warn some people
07:30out there because a lot of people are like, okay, you know, I'm just learning SQL. I should just use
07:34this to learn SQL. And I'm like, I don't think that's a good idea because if you do not know this code
07:40already and it tells you to do something, you may get a worse option or a not perfect option
07:46and you need to know it. So just looking at this, this is pretty good. This is pretty similar to
07:51something I've worked with, but in previous versions, it's given me different options that
07:55I didn't like as much. So be careful with this one when it's actually generating the code. Oftentimes
08:00you'll need to tweak it or change it. I have found that ChatGPT is pretty good, but sometimes it just
08:05gets things wrong or it just gives me bad options. I'm like, this is not good code or it's not perfect
08:12code and I have to tweak it quite a bit to get it to what I want it to do. Now, the next thing that I
08:16really like ChatGPT for is writing comments for my code. So let's take this for example. Let's say I
08:23have this code. I'm going to say, can you please, and I like to be polite, can you please write
08:29comments for this SQL code? And I'm just going to paste it in there and I'm going to ask it to
08:37write comments for it. If you haven't written any code, you may not know what that means, but comments
08:41are just, you know, added information for somebody. If they're coming in and looking at your code or for
08:46you in six months to come back and look at your code, you know exactly what it's doing. So it's
08:50going to describe every part of this code for you, which can save you a ton of time. If you're
08:55commenting your own code, that can take a while. And you can just go in here and add a little bit
09:00more information if you want to, but this is like a fantastic starting place for commenting code, which
09:05I personally sometimes forget. And so I've been using this for even more complex code. Like this
09:10is fairly simple code to comment. I've given it even more complex code and it's done a pretty good job.
09:16There are times where it gets it kind of wrong, but this is pretty impressive for actually
09:22generating comments for code. It's saying exactly what each step is doing. The next use case that
09:27I had been using ChatGPT for is for creating data dictionaries. Now, if you don't know what a data
09:32dictionary is, it's basically a table or something that describes what data is in your data set,
09:38whether it's in a CSV or a SQL server or wherever, wherever that data sits, it's good to have that
09:44data dictionary to know exactly what data you have. So I'm going to say, can you give me a data
09:53dictionary or this data set? Now I'm just going to paste this in here. I just copied this from an
10:00Excel. So that's all I did. And I'm going to hit enter. Now it's starting to generate this. It's
10:05going to actually give me a table. So it's creating this table of the column name that's in there,
10:10the type of data that is in there, and then an actual description of it. Now I haven't given it
10:14any previous information. It's doing this purely based off of context. So it's taking this name
10:20state and saying, okay, this description must be the state name. It's doing total population. So
10:26that must be the total population. Then it's even taking in things like comparison operators. And
10:31it's saying these rates dot age is saying a rate for age less than 18. Now this is fairly, fairly
10:39simple. I'm not saying it's giving you anything advanced. It's not giving you a ton of
10:44information beyond what it's reading in, in the column name, as well as reading in, in
10:49the data. So if you give it a more complex data set, sometimes it even takes the data
10:52that's in there and it says, okay, these are state names, but they're labeled something
10:56different. It's labeled just one, two, three for the column name. So then it will tell you
11:01this one, two, three column is state name. So it even takes that into consideration. It's
11:05pretty wild. I mean, it is really, really impressive. Now, again, I'm just giving you like one
11:10or two examples of each of these. Otherwise this video would be like an hour long. But
11:13the next thing that I've been using chat GPT for is optimizing my code. Now let's go right
11:18over here. This is an actual SQL query that I have written. I'm going to copy this. It's
11:24just doing what I would consider a fairly, let's take a look at it, a fairly simple window function
11:31as a sub query. And that's using a case statement to kind of use that information. Let's see how it
11:37would optimize this code. Can you please optimize this MySQL code? And I'm going to give it the
11:47code. Now, what it's going to do is literally take it in and rewrite it to properly run it faster,
11:53to actually optimize that code. Now, if you don't know what query optimization is, it's basically
11:58where you're either changing a query or you're changing the database to run faster. So what it's
12:04actually doing here is creating an index. This is a way that you can optimize code. Although
12:08for this code, that's not exactly what I was wanting. I wanted it to optimize the query
12:14and not create an index, which does speed it up. That is a fact, but I'm going to now say,
12:20I'm just going to say, can you optimize this in just the query itself?
12:27So it's going to remember the previous thing that I asked it to do. And now it's going to
12:33understand by context what I'm asking it because I don't have to actually input my query again.
12:38It's going to know. Let's see exactly what happens. So I've tried this twice and it gave me different
12:45answers each time. Last time it gave me what I hope it's going to give me now. That time it gave me
12:51an index, which it didn't give me before. So it gives you different things. It's just really,
12:57really fascinating how this works. Now, what it's giving me right here is the actual output that I
13:02was hoping it'd give you. Although this is a learning opportunity. That's just how chat GPT works.
13:07This right here is what I thought it would give you instead of the index. But what it did here is
13:13instead of in this part of the code, having this select everything, what it is now doing is just
13:19selecting the actual columns that we need. So now it's just selecting employee ID department and
13:24salary instead of all of the columns, as well as this window function right here. Now it's going to
13:29even tell you why it does that or how it does that. So by using a sub query to only retrieve the columns
13:35that are actually needed in the final result, you reduce the amount of data that is needed to be
13:40processed, which speeds up the query. So those are all ways that I've been using chat GPT. It's been
13:44pretty amazing. I will give the disclaimer though, that not every time do I get great results. There
13:49are oftentimes where I have to reword things or I have to ask it to generate another response.
13:55And that happens a lot. But when I actually do get what I'm looking for, it's pretty impressive.
14:00Like I usually am like, wow, how did it do that? It's pretty cool. Now what I want to do is just take
14:05a small data set and do some data analysis, kind of show you some of the things that it can do.
14:09Some of the ways that I would use it, I'm just going to pull a data set from Kaggle and use that,
14:15keep it pretty simple, not use any real data with any information I shouldn't be putting into chat
14:20GPT on it. So let's go down here and in this Excel file, and we'll be using this one later in this Excel
14:27file, have this data set right here. Now this is going to run really slowly because this is a lot
14:32of data, but I'm just, and I say a lot of data, it's only 30 rows, but chat GPT does not take in this
14:39data well. So let's get rid of this. And all I'm going to say is you are a data analyst.
14:46So I'm giving it some context. That's who you are. That's what you do. Here's what I want you to do.
14:51Can you please give me some insights and recommendations on this data set? And I'm going
15:01to input this data in here. Now, when I hit enter, it's going to analyze this really quickly. Now,
15:07if you ask it to do other things, which we're going to do in just a little bit, it may take
15:11quite a while. So I want to give it a second to actually write everything. Basically what it's
15:16saying is to actually provide insights and recommendations, we want to take some initial
15:20steps. And these are some pretty common things that you'll want to do with data, like data
15:23cleaning, exploratory data analysis, product analysis, geographical analysis, and time series
15:29analysis. So let's take a look at this data really quickly. We do have a product. We do have
15:36an order ID, the actual product, how many we have, the price, the order date, and where it was purchased
15:42from. So what it can say based off of just giving it the data set randomly, it says you can offer
15:49discounts or promotions on the least frequently ordered products to increase their sales. Now,
15:54what we could do is literally take this and say, can you tell us more about this?
16:03And I'm just going to ask it to explain it a little more. So what it told us is you can offer
16:08discounts and here's why you'd want to do it because you'll attract more customers. And it even
16:12tells us how to identify these products. And I'm going to ask it, can you identify these products
16:20for us? And I'll say, and output the SQL code as well. So if we go back up, so it was generating
16:29the code. It says to identify the least frequently ordered products, you can use the following SQL
16:34code. And it literally writes the code out for you with the actual column names that it says you
16:39should be using. That's the part that always kind of like, I just think is really impressive because
16:44this kind of code is actually fairly simple, fairly generalized, but it makes it not so generalized
16:51because it's actually using your data. That's the part that always like blows my mind. I'm like,
16:55that's just really impressive. That's using your actual column names or your actual code
17:00or generating your code, using your column names. That is the part that's usually pretty impressive to
17:05me. So all this is doing, and it even says we're using the product count. So we're using an aggregate
17:11function in the orders table and counts the numbers of times each product appears and selects the top
17:1710 least frequently ordered products. Again, pretty simple, but you know, we've asked it to do a lot
17:24of contextual things. Take a look at this data set, tell us how to offer discounts and actually write
17:30the code to identify what we should be offering discounts on. It's pretty, pretty cool. Now let's say
17:36we didn't want to use SQL because that's what we're using right now. I'm going to say, can you take
17:41our data set and put it into a pandas data frame in Python? So I'm going to see if it remembers.
17:51Sometimes it does. Sometimes it doesn't. It looks like this time it does. Depending on how far back
17:56the conversation is, or if it remembers what our data set is, sometimes it remembers, sometimes it
18:00doesn't. This time it looks like it is working. Now what it's actually doing is creating each row
18:07of that data and inputting it as data. Now, sometimes it's going to error out. It takes
18:12too much processing power and we have like 30 rows. It's only on row three. So I'm going to let this
18:17run for a little bit. We'll see if it actually works. So it's looking like it aired out. It just
18:22stopped. It just stopped. This is the, uh, I would say somewhat downfall of the chat GPT right now is
18:29that the servers are just overloaded. There's like millions and millions of people trying to use it
18:33at this very moment. So oftentimes when it's using these large data sets, it just doesn't work in
18:38the future. I think there'll be add-ons to things like Excel and MySQL and all these different things
18:42where it'll be integrated with it and process it much faster. Um, but now it doesn't. So we're just
18:47going to refresh and keep going from here and we're just going to go right back up here. And this is our
18:53previous data sets and we'll scroll all the way to the bottom. So it saved our conversation.
18:57Thank goodness. Um, but it just, you know, it just blanked out and luckily it saved us. That was a
19:04good thing. Now, because I refreshed it, let's see if it still remembers. Um, again, we're trying to
19:09analyze data. We're trying to get some information out of here, but you know, these things are happening.
19:13So I'm just kind of going with the flow. So what I'm going to say is what products
19:17in our data set made us the most profit. Let's see if it remembers.
19:27So now it's going to say to identify these products, you need to calculate the total revenue.
19:33So this part is really cool because it's going to kind of based off of our prompts,
19:37based off of us saying we want profit, it's going to say, we need to calculate the profit first
19:43to determine the profit. So it's going to write it using pandas, but we can ask it to write this
19:49in MySQL, but it's going to take it and say, okay, take the quantity ordered times the price of each
19:55one. And this is our like potential profit. And it's even giving us comments. I mean,
19:59it's really going above and beyond. I mean, if you were to integrate this in something like Python or,
20:04or, you know, any other thing, this could be a fantastic tool, but this is pretty great. Um,
20:10this looks like pretty standard code. It's even grouping it and giving us the sum. So if we ran this,
20:15if we took this and put it in Python and we had our data set in there, then it would probably give us
20:20pretty close to the right answer. Now I'm going to ask it to write this in MySQL as well. So I'm
20:25going to say, can you write this in MySQL? And there we go. So it wrote it in MySQL as well.
20:33And this looks pretty simple and very straightforward. It looks correct to me. Now,
20:37one thing I want to say is we're analyzing this data. We're kind of prompting it. I'm just giving
20:40some examples, right? We're not going to do like a full analysis that I could do that in a whole
20:44another video, but, um, what we're doing is just, I'm giving examples of how you can actually ask it
20:49to analyze this data for you. So what we're going to do next is we're going to ask it to categorize
20:54some information for us. So I'm going to say, can you categorize, let me say the products that made us
21:01a lot of money versus a little bit of money in MySQL and do MySQL. So we're asking it now to
21:13categorize this data for us. If it made us a lot of money, we're going to want it to say it made us
21:17a lot of money. If it's a little money, we want it to tell us it made us a little bit of money.
21:21Let's see how it does this. And again, this is all pretty, you know, generalized questions,
21:28kind of open-ended questions. We're going to see how it actually interprets that open-ended question.
21:34Um, and so it's literally writing a case statement saying it's a, if it's over a thousand,
21:39it's a lot of money. If it's less than 500, it's a little bit of money. So this, um, this is something
21:46that's, you know, it's really interesting. You know, this is not something that I'd ever write
21:50in like a real SQL query, um, because it's leaving that 500 to a thousand somewhere in there.
21:57And it's saying it's somewhere in between. Um, I would never actually do this in real code,
22:02but that's how it's interpreting it and writing it. So this would be a mistake to me. This is just
22:07like a small thing that I would go back and change and, you know, customize. Um, but we could give
22:12it more specific prompts. We could say, if they made us a profit of more than a thousand dollars,
22:17give it this. If it was less profit than this number, give it less. So we could even give it
22:20more specific and it would do it. Now, if you remember from our Excel right here, we have this,
22:26a column, and what we want to do is break it out by the street, the city, and the state,
22:32as well as zip code. So I'm going to ask it to do that because this is something that I've actually
22:37done a ton in SQL Python and any programming language is breaking it out. So you can group
22:42on it and, and, you know, um, clean up that data better because in its current state, it's not
22:47very usable. So I'm going to say, can you break out the last column into street, city, state,
22:58and zip code columns. And we'll say in my SQL. Now you can do the exact same thing for Python
23:06or whatever, but let's see how it writes this. Now I post this on Instagram the other day and
23:12before it was using substring index. Let's see if that's how it does it. Um, again, and it looks
23:18like it is, and this is almost exactly how I would write this. So this one really impressed me because
23:23I'm, I'm like writing this code to test this video or like write, make this video. This one really
23:28impressed me because, you know, substring index, um, using substring and locates like this is,
23:33you know, it's not super straightforward. It's not super easy, but it does it. Um, just very,
23:39very impressive in my opinion, very impressive how it's able to write this. And this code,
23:44if I'm just glancing at this, this code looks correct. Like I could run this and it would take
23:49that column, break it out into those four columns for us. Now, what we would want to do, cause this is
23:54just in a select statement, but I could literally, and let's try it. I could literally just tell it,
23:58can you create new columns for each of those? So again, I'm being vague. Each of those, I'm just
24:08trying to test it a little bit while we're in this video. So you can see how impressive this is. So now
24:14it's going to create a new column, city, state, street, and zip code. And then it's going to tell us
24:20how we can do that. Um, I don't know. It's just really, really cool. And, uh, now it's literally
24:26writing the code to use that substring index, which we generated up here, which looked correct.
24:31And now it's telling us in these, um, you know, update statements, how to actually apply that to
24:38those columns that we created with this code. All I have to do is copy and paste this again. It's
24:44blowing me away a little bit. Every time I run it, it's just kind of like, how does that work? Like
24:48on the backend, what code are they writing? Uh, or what code are they using to, to generate this?
24:53It's just really impressive. Now you can ask it to convert it to a different programming language
24:57or whatever you want, and it would do it. Um, but what we're going to do now is do something a
25:01little bit different. I'm going to ask it and I'm going to actually, I forgot to create this
25:04beforehand. I'm just going to say first underscore name. And what I'm going to do is I'm going to give
25:10it all types of different inputs. So I'm going to say, Sam, and then I'm going to do, um, Kelly,
25:17and then I'm going to do Josh one, two, three. So I'm going to take this. I'm going to ask it to
25:23clean this data for us just as like a small example. Can you clean this new data set?
25:31And I'm going to give it to it and just see what it does. Um, if I need to be more specific, I will,
25:36but let's see how it takes in that data and standardize it. It even says right here,
25:39we're going to use upper to standardize the text and to remove numbers from the characters.
25:45You can use regex replace, which is a hundred percent accurate. That's probably what I would do
25:51if I had a large data set with all data like this in there, I would be using those things to clean
25:56this up. So this code right here is going to take regex replace and replace all those, uh, numbers
26:01and just keep the letters basically is actually what that's doing. Um, and that's going to put it all
26:07upper. So I don't want it in uppercase. I'm going to say, I don't want it in uppercase.
26:13I want proper case. Could you please write that code as well?
26:22And I'm going to see what it does. I hope it makes it into more proper case. Cause I don't
26:26like using it in uppercase. Just not always what I like visually seeing, but let's see what it does.
26:33So it's basically going to use init cap and make it a more proper case instead of doing it all in
26:37upper or lower or something like that. And now what I'm going to do is I'm going to say,
26:41can you put that cleaned data in a table for me? And it's going to actually generate this
26:49and create a table for you to copy and paste. Um, and it's actually creating the table. That's
26:55not what I wanted. If I'm being honest, I wanted it to create just a list for me to copy and paste.
27:00So I'm going to say while it's generating it, I'm going to say, can you put this in a list? So I
27:07can copy and paste this into Excel. So let's see if it does better this time. Cause I didn't want it
27:17to create the table in SQL, but that's the context that it was understanding. And now it's doing it in
27:22Python again, not exactly what I wanted, but it's doing something right. It's definitely working hard
27:29to try to understand what I'm trying to tell it. I just want it to literally give me like
27:33something I can copy and paste. And it's not doing that exactly. Um, I could just take this if I
27:37wanted to and break it out, but it's not doing exactly what I wanted. Although that happens
27:43with chat GPT, the very last thing I'm going to do, and I'm just going to go back here to have the
27:48second data set and it's a lot of data. So I'm just going to take like the first two columns.
27:52And what I'm going to do is just ask it to basically explain what this data set is. I'm
27:58just going to say, can you please explain what is in this data set? Now I'm being specifically
28:07vague for a reason, because you can be much more specific. You can say, explain to me what these
28:12columns are and how they correlate to each other. You can explain, ask it to explain almost anything,
28:16but it's basically going to give us some context, some overview of what kind of data is in
28:21here. Much like a data dictionary would, it's kind of doing the same thing. Now this is going
28:26to take a long time. So there's a lot of columns. There's like, like 30 or 40, but it's doing its
28:30thing. It's going to keep going until it explains all of them, unless it errors out because it's
28:35just the system's overwhelmed. But, um, while this is running, these are some ways that you can
28:40actually analyze data. I'm just giving you a lot of examples because I want you to go and test this
28:44out. It's, I don't know. There's no other word to use other than just, it's really impressive.
28:51Um, I'll have an entire other video that I'm going to be making on this, which I'm basically
28:57going to be talking about how it'll help with data analysis. Um, the data analyst job market,
29:02how it'll impact it in the future and some predictions that I have about its capabilities
29:07and things like that. There's just a lot going on in this world. Um, Google's about to come out.
29:12I'm just kind of giving some filler for my next video on this, but Google's coming out with its own
29:17version. Uh, I believe it's called Sparrow. Um, it's using their Lambda, um, model that they've
29:22built and I'm super excited for that one. So I probably will do some comparisons between those
29:27as well. This type of technology is just, it's going to change things. I don't know how exactly
29:34I have some predictions. I have some, um, you know, information that I'll be providing. This is
29:39going to change things. This is just incredible. And it can analyze data like this on these small
29:43data sets, but when it's incorporated in things like Azure, like with Microsoft, which is, you
29:47know, doing some big things, if it's incorporated with Azure and it can analyze these huge, massive
29:52data sets a lot faster, a lot of things could change. So in this video, I kind of showed you
29:58a lot of my favorite things that I've been using ChatGPT for. I've showed you some of the
30:02functionalities as well. And then we even analyze some data, just kind of poking around and asking
30:07it questions and seeing how it interpreted those questions, as well as asking it to write it
30:11in Python and SQL. So, you know, I just gave you a few different variations and flavors.
30:16It wasn't like a full analysis. This is just like an example. In another video, I might take
30:21an entire real data set and see if I can get ChatGPT to analyze it well. Um, but not in this
30:27video. So I hope that this was helpful. And I hope that this was exciting for you. This
30:30stuff is super, super exciting to me. I find it just incredibly interesting. And I think
30:35it's just really impressive. So go try it out yourself. Go see if you like it. I will
30:41be making more videos on stuff like this because I personally think it's just incredible. So
30:46with that being said, thank you for watching this video. I hope you learned something. I
30:50hope this piqued your curiosity. Be sure to like and subscribe, and I'll see you in the next video.
Recommended
2:11
|
Up next
12:00
14:59
1:21
12:50
1:23:56
1:03:10
1:08:17
1:06:27
48:44
56:23
1:22:27
1:22:34
59:43
1:59:39
1:19:21
1:28:40