One year of ChatGPT: usage still to be balanced

The impact of ChatGPT in just 365 days has been impressive! From the media agenda to academic research, OpenAI’s software has created unprecedented disruption in the industry and influenced society’s perception of artificial intelligence. At the same time, with only one year of existence, it is still difficult to evaluate the long-term impact of the massification of generative AI.

At the end of one year, there are still many uncertainties, and several topics must be weighed on this balance of “pros and cons” about the impact of ChatGPT on society.

Firstly, ethical questions. AI algorithms, programmed in their genesis by humans, inevitably have moral and ethical inclinations that may be reflected in the presented content. On the other hand, questions about the integrity of the information provided require a critical look rather than an uncritical use of these tools.

There is also the responsibility for decisions based on the information provided by ChatGPT - in harmful cases, who is responsible? OpenAI? The user?

On the other hand, safeguarding the previous issues, it is undeniable that efficiency gains are considerable, for example, in data analysis. AI tools allow for analyzing large amounts of data and “interpreting” them in seconds. Potentially, we are talking about saving hours of work that can be allocated to more complex tasks that require the surgical intervention of a qualified professional.

However, democratizing access to AI also entails enormous security challenges for organizations. Taking a simple example, if a particular employee shares a set of data, which may be sensitive and confidential, from their organization with these tools, what safeguards exist to protect that data? Where are they stored? Who can access them? Under what circumstances? These are some of the significant questions IT managers currently face in their companies. What is the balance between productivity, efficiency, competitiveness… and security? It is a concern company leaders must address immediately!

The evolution of AI tools and ChatGPT itself has been exponential. Natural language processing (NLP) models are becoming more advanced, and interactions are becoming more “human,” making them more accessible for the “common” user to adopt. The ability of generative AI to assess the context of a particular “conversation” and provide coherent and contextually relevant responses is one of the great attributes of this type of solution and a crucial element in creating meaningful and engaging interactions, improving the user experience.

Given the identified challenges and benefits, it is difficult to predict precisely the long-term impact of ChatGPT and similar tools on society. But we have one certainty today… it will be significant!

It will change how we work, study, access information, manage (and generate) knowledge. This impact is only comparable to the massification of access to the internet at the beginning of this century. Like all revolutions, some pros and cons weigh on the balance… but the expectation is that the best is yet to come!

If ethical challenges are addressed correctly, some security risks are mitigated, and technology is incorporated into our companies’ processes and systems, I am confident we will achieve the desired balance. One thing is sure: one year after the emergence of ChatGPT, no one remains indifferent to it!

This article was originally writen for SAPO and published in November, 2023 -Um ano de ChatGPT: as pessoas est?o a ser substituídas por máquinas? - Tecnologia - SAPO 24