Even if you're already advanced, learn data science faster with ChatGPT

directory

• Is data science still relevant?

• Why use artificial intelligence to learn data science?

• How to really use AI to learn data science

1. Step 1: Develop a roadmap
2. Step 2: Design ChatGPT as my mentor
3. Step 3: Develop a study curriculum
4. Step 4: Try advanced tools like AutoGPT
5. Step 5: Do the project
6. Project walkthroughs for beginners
7. Project walkthroughs for senior practitioners

•conclusion

Everything has changed in a short time. AI tools such as ChatGPT and GPT-4 are taking over and revolutionizing the landscape of education and learning tech skills. I feel like I need to write this article to address something important:

1. In the new era of artificial intelligence, is learning data science still important?
2. If so, what is the best way to learn these skills by leveraging new technologies available? What should I do if I have to start over now?
3. What does the future of data science look like?

Does data science still make sense?

As AI continues to evolve, will data scientists become obsolete or will their roles be more important than ever?

From a personal point of view, I still feel that I add more value to my clients than AI, and with these new tools available, I have been able to (at least) double my work output. Now, I don't feel like AI will replace my job, but actually, the future is more uncertain than ever.

Before you worry about jobs disappearing, let's look at the following scenario: At some point in the future, you run a company and let AI do the analytics for you.

Who do you want to run AI, prompt it, and supervise it? Do you want these programs to be supervised by someone with a background in data science or software engineering, or do you want them to be supervised by someone who is not trained?

I think the answer is very obvious. You need someone with the experience and knowledge to work with the data that runs these AI systems.

In the short term, this scenario is hopefully hypothetical. But it did give me some confidence that some aspects of these skills are resilient.

Even if circumstances change and data scientists reduce their hands-on coding work, I still think that in a world more closely integrated with AI, these skills developed by learning the field will be very useful. AI is rooted in data science, and to some extent, we're more integrated into this system than other professions.

Beyond that, AI still creates illusions, and we need as many knowledgeable people as possible to oversee it and act as a feedback loop.

While I'm not sure about the future of data scientist work, there's one thing I'm pretty sure of: data, analytics, and AI will become a more important part of our lives. Don't you think that people who study these fields will also have greater relative success?

Why use artificial intelligence to learn data science?

If I don't think data science is still worth learning, that's the end of this article. To be clear, I still think it's still 100% worth it. But, let's be honest, it's no longer enough to just learn data science. You also need to learn how to use new AI tools.

Interestingly, it's easier to learn data science and these AI tools than it is to learn data science alone. Let me explain.

As it happens, you are at the perfect time to learn both fields together.

If you learn data science by leveraging existing new AI tools, you'll reap the benefits of being twofold:

1. With AI learning the data domain, you get a more personalized and iterative educational experience
2. You can also improve your skills in AI tools at the same time.

If my calculations are correct, you will do more with less.

If the ability to use AI tools can help you get jobs and do better, it's better to know how to use them than to ignore them. In the last three months, I feel like I know more about data science than I have done in the last three years combined. I attribute most of this to the use of ChatGPT.

So, what do you do? How to really learn data science through artificial intelligence?

How to really use AI to learn data science

If I had to start over with all the tools available, that's exactly what I would do.

Step 1: Develop a roadmap

I will make a roadmap. You can do this by checking out other courses or having a conversation with ChatGPT. You can actually ask it to develop a data science learning roadmap for you based on your learning goals.

If you don't have a learning objective, you can also ask it to create a list for you where you can find what you like.

If you want to learn more about developing an education roadmap, check out my article on the topic for a deeper dive into the subject.

Step 2: Design ChatGPT as my mentor

I would design ChatGPT as my mentor. You can create characters using GPT-4, which is probably my favorite feature. You can use tips like this:

In this case, you are one of the best data science teachers in the world. Please answer my data science questions in a way that will help me better understand the field. Please use many real-world or practical examples and give me relevant practice questions.

Step 3: Develop a study curriculum

I'm almost certainly biased, but I think free courses or paid courses are still good options for creating a learning structure. During the learning process, you can ask a ChatGPT tutor to give you examples, expand topics, and provide practice questions.

Step 4: Try advanced tools like AutoGPT

If you're a little more advanced in AI, you can use tools like AutoGPT to generate class schedules for you. I'll probably try to do this and see what happens.

Step 5: Do the project

If you're already familiar with coding, you might be able to jump to doing projects. I personally learned a lot from the project I worked with ChatGPT.

If this is your first project, just ask it to do a few things, but as you progress, you want to be more intentional and interactive about how to use it.

Let's compare how beginners and advanced practitioners should approach project learning.

Project walkthroughs for beginners

An example of a starter project walkthrough might look like this:

1. You provide ChatGPT with information about rows and columns of data
2. You ask it to create boilerplate code to explore nulls, outliers, and normality for that data
3. You ask it what questions you should ask about this data
4. You ask it to clean the data and build a model for you to make predictions about the dependent variable

While it looks like it has done all the work for you, you still need to get the project running in your environment. When you proceed, you are also prompting and solving problems.

There's no guarantee that it will work like copying someone else's project, so I feel it's a good middle ground for participating in learning.

Project walkthroughs for senior practitioners

Now, let's consider how more advanced practitioners will use it:

1. You can follow the same steps for generating boilerplate code, but this should be extended. Therefore, you may want to try to explore the data and hypothesis testing more in person. Perhaps, pick a question or two that you want to answer with data and descriptive statistics and start analyzing.

2. For those who have done some projects, I recommend generating some code yourself. Let's say you make a simple bar chart with plotly. You can enter it and ask ChatGPT to reformat it, change the color or ratio, etc.

By doing so, you can quickly iterate on visualizations and see in real time how different adjustments to the code change the graph. This instant feedback is very conducive to learning.

3. I also think it's also important to check these changes and understand how they work. Also, if you don't understand something, just ask ChatGPT to expand its features.

4. More advanced practitioners should also pay more attention to data engineering and production code pipelines. These are all things you still need to get your hands dirty with. I've found that ChatGPT helps me with some of my tasks, but I need to do a lot of debugging myself.

5. From there, you might want to let the AI run some algorithms and make parameter adjustments. Honestly, I think this will be the fastest part of data science to automate. I think parameter tweaking will see diminishing returns for the average practitioner, but not for high levels.

6. You should focus your time on feature engineering and feature creation. This is also something that AI models can help but cannot fully grasp. Once you have some decent models, see what data can be added, what features can be created, or what transformations can be made to improve results.

In a world with these advanced AI tools, I think it's more important than ever to do projects. You have to build things and share your work. Fortunately, with these AI tools, it's easier than ever to do this. It's easier to make web apps. It will be easier to use a new package that you have never used before.

I highly encourage you to create real-world impact and tangible things in your data science work. This becomes a new way of differentiating when others also use these tools to learn and build.

conclusion

The world is changing, and so is data science. Are you ready to rise to the challenge and create real-world impact through your projects?

I mentioned it before, but I think the way we all work is changing. I think it's an uncertain time for all fields, including data science.

On the other hand, I think data science is the perfect combination of technology and problem-solving skills that scales well to almost any new world or field.

I discussed in detail how I consider data science to be one of the fields closest to pure entrepreneurship. I think in a world changed by AI, we need to leverage this entrepreneurial spirit as much as possible.