How will artificial intelligence (AI) change the market in the future? Will we have to relearn technological literacy to be competitive? How does ChatGPT work? And why can’t it be characterised as an understanding program? Where can one learn about these technological innovations and learn how to manage them? 

These questions were discussed by IT specialists from various fields during an event dedicated to artificial intelligence organised by Godel Technologies in Vilnius. Godel’s Lead Software Engineer, Eduard Mikhailau shared their thoughts at the event. Eduard is a Software Engineer with 18 years of experience in the field. Since his studies, he has been improving his programming skills, studying AI tools usage for software development and conducting educational events for employees. In his discussion, Eduard also answered some tough questions about navigating ChatGPT in the business landscape. 

How would you describe the ChatGPT trend?

Simply put, ChatGPT is the first technology that showed the world how ‘cool’ large language models (LLM) could be. Many specialists had doubts about neural networks’ functionality and mind of their usage only in the narrow scope of tasks. And suddenly we’ve met something that can generate code, analyse texts, write music, and process any information that could be presented in symbol view. 

How has ChatGPT evolved so quickly?  

AI will touch everyone in the form of “ChatGPT” in all professions, one way or another. AI could develop even faster, but this process is currently being held back by gaps in legal regulation or uncertainty about what that regulation should be. AI currently works and is quite effective, although relatively little has been invested in its development. Many people fear that they will end up on the street if AI suddenly takes over their jobs.” However, AI is not something to be afraid of, but like any new technology, to be understood and applied.  

The dark side of any powerful technology goes from the lack of knowledge and the absence of society-applied regulations for technology usage. And today we see the very first steps towards AI law regulations. To my mind, these new laws and limitations give humanity time to ruminate about the place of AI in the world. And despite all these, sometimes I think about the fact that OpenAI has spent dozens of millions of dollars to train ChatGPT, what if someone would spend billions…?  

Technologies are developing faster than legal regulation. We are constantly looking for ways to use AI without violating copyrights, legal acts, and protecting confidential information. Since legal regulation lags behind technological progress, companies have to create internal rules and procedures. 

How can we adapt ChatGPT to our everyday work? 

According to an MIT study, workers improved productivity by 37% using ChatGPT, an “Industrial Revolution-level large” efficiency gain. It’s enough to check the list of companies that have been started with AI investments and it soon becomes clear what discoveries and in what fields we are going to have – it’s pharma, medicine, space, and, of course, industrial processes automation. 

When it comes to using ChatGPT in your work, many question the types of queries you need to form that will give you the information. Not only will you receive a reply, but the information you sent will be unknown when it was transmitted, and then it is unknown by when and for what purpose it is used. For this reason, difficult dilemmas are faced by programmers using AI. 

To analyse and improve AI, it needs to present the current ones. AI does not work by itself, it needs some data, currently publicly available and placed on the Internet. Some authors have expressed opposition to artificial intelligence using their work. However, this is very difficult to ensure because AI works like a neural network. It cannot be instructed to “forget” any information because it is stored unstructured; this is the specificity of this network. Therefore, various filters are developed.   

What does the future of ChatGPT hold? 

The key to the future is for us to learn as much as we expect AI to. Let’s say someone instructs, “rewrite my text in the style of some particular author”, and the AI will say that I’m not allowed to do this, because, for example, copyright prohibits it. However, such bans can be circumvented in various ways. Therefore, it is a huge challenge to adjust everything properly.  

There have already been cases where people have used ChatGPT in a chaotic manner across different servers and important trade secrets stored on some servers were stolen. It is still not clear how to act in such situations as legal regulation lags behind the development of technology. The use of AI is a new literacy.  

I believe the future of ChatGPT will depend on the user. The desirable would be for our knowledge on the subject to grow together, learn more, gain a better understanding of what AI is, how it works and what it can be useful for. This is a key reason I will keep talking about this topic and run even more educational events. We are in a rapidly changing technology market, and it is very important that people from different IT companies and professional fields keep up with these, and chat about ChatGPT.