If your job is to analyze and report on data, it’s understandable that you might feel some anxiety about rapid developments in artificial intelligence (AI). In particular, the viral app ChatGPT has captured the imagination of the general public over the past few months, providing a powerful demonstration of what AI can actually do. For some, it can also appear as a warning of what the future holds.
There is no doubt that one of the strengths of AI is its ability to make sense of large amounts of data – looking for patterns and shaping them into human-friendly reports, documents and formats. It is the daily bread of data analysts and many other professionals in the knowledge economy whose work involves the use of data and analysis.
It’s true that artificial intelligence – a term that in business and industry usually refers to machine learning – has been used in these areas for years. What ChatGPT and similar tools based on Large Language Models (LLM) and Natural Language Processing (TLN) have to offer is that they can be used easily and efficiently by anyone. If only a CEO could say to a computer, “What should I do to improve my customer satisfaction?” or “How can I increase my sales?” Should he worry about hiring, training, and maintaining an expensive team of analysts to answer these questions?
Fortunately, the answer is probably yes. Indeed, as AI becomes more accessible and popular, this team may become even more important to the business than it already is. What is certain is that their work will change dramatically. So here’s a look at how this technology will impact the field of data and analytics as it becomes more prevalent in the near future.
ChatGPT is a publicly available chat interface (or chatbot) powered by an LLM called GPT-3, developed by the research institute OpenAI. LLM is part of the field of machine learning known as natural language processing, which basically means that it allows us to talk to machines and have them respond to us in “natural” (i.e. human) languages. In short, this means that we can ask him a question in English, or indeed, in any of the nearly 100 available languages. It can also read, understand, and generate computer code in a number of popular programming languages including Python, JavaScript, and C++. We’ve been used to interacting with natural language processing (NLP) for a while now, thanks in large part to AI assistants like Alexa and Siri, but the LLM running GPT-3 and ChatGPT is much larger, allowing it to understand more complex . inputs and provide more complex outputs.
LLM GPT-3 appears to be able to use the language in a very complex way because it has been trained on a massive informational dataset, which is said to include more than 175 billion parameters. This includes an open web data repository called Common Crawl and several online book archives. By processing all of this data, it is able to learn how words relate to each other and predict what is likely to be the most appropriate response to any prompt (question or other input) that is given to it. It is sometimes referred to as “generative AI” because it creates new results that have not been seen before.
What are the limits of ChatGPT?
Before getting excited about its potential, it’s worth noting that despite the hype, there are some very important limits to what the technology can do today. First, it often makes mistakes — sometimes very minor ones — that can easily make anyone using it professionally feel a little silly if they’re not careful.
For example, the author of this article, Bernard Marr, did research on ChatGPT to discover what parts of a data analyst’s work he was able to automate. One of the first responses was: “ChatGPT can generate graphs, charts and other visualizations”. This is obviously wrong, as it is only able to generate text.
When it comes to data analysis, ChatGPT is also limited by the fact that we cannot load data beyond any information that can be entered in text form. We cannot, for example, download an Excel sheet with sales figures and request information from it. Of course, we can’t know what future versions can do. In that case, let’s see how it can be used and speculate what might be possible with LLMs and LNPs in the near future.
How can ChatGPT, Large Language Models (LLM), and Natural Language Processing (NLP) be used in data and analytics?
Create code and applications that can analyze data or automate processes such as collecting, formatting, or cleaning data.
Define data structures – for example, the fields that should be included in database records or the row and column headings required for a spreadsheet.
Tell us how tables, graphs, graphs or graphs should be created and what information should be included.
Suggest information to include in reports so that different audiences – executives, department heads, managers, etc.
Create training materials to teach workers how to apply analytics to their own data.
– Identify data sources that may contain the information we need for a particular task – for example, “Where can I find data on financial fraud in India?”
Generate mock or synthetic data for various purposes, such as training other machine learning models or testing algorithms.
Provide advice on compliance, regulations and practical steps that can be taken to ensure data operations are legal, fair and ethical.
Define analytical processes and suggest best practices that are most likely to yield the desired results.
Is ChatGPT a threat to data and analytics functionality?
As we’ve seen, ChatGPT can easily automate some of the tasks that are traditionally performed in analytics functions – such as business analyst, data analyst, and financial analyst roles. Future iterations of the technology are likely to become more effective and efficient in this regard.
But that doesn’t mean that everyone who works in analytics will find themselves out of a job anytime soon. In fact, today’s most advanced LLM and LNP tools still lack capabilities such as critical thinking, strategic planning, and complex problem solving. Most experts agree that tools based on machine learning are unlikely to be able to perform these functions at the same level as humans in the near future.
It is likely that companies and other organizations will need human experts in this field for some time to come.
However, analytics jobs that require only repetitive work are likely to be largely automated in the near future, and some jobs will likely disappear because of this.
At the same time, new jobs will be created. These likely revolve around the ability to deploy tools like ChatGPT while practicing human decision making, problem solving, leadership, strategy, direction, and team building.
There are two very important rules to follow, if you are into data and analytics. First, whatever you do, don’t deny that artificial intelligence may be about to radically change the way you work.
Secondly, learn to use this technique as a tool. Understand its capabilities to increase your own skills by using tools like ChatGPT or any other upcoming tool to automate routine and repetitive tasks. In this article, a number of tasks that this technique can be applied to are immediately listed – go through them and make sure you understand how to implement each of them. Next, learn how to take advantage of the resulting time and efficiency savings to build your skillset and focus on areas where you can really make a difference.
By ignoring the reach of AI into your profession, you may find yourself falling behind as colleagues and competitors willing to move with the times reap the rewards. For now, we see only the tip of the iceberg. With the development of technology, more and more aspects of our daily work will be automated. To thrive in the age of artificial intelligence, it is essential to stay ahead of the game, learn to use new tools as they become available, and stay aware of where the key is. A human is still essential.
Translated article from the American magazine Forbes – Author: Bernard Marr
<< اقرأ أيضًا: كيف سيحدث ChatGPT ثورة في عالم العمل؟ >>>
