ChatGPT: Beyond the excitement
There has been a lot of excitement following the release of ChatGPT to the public. In simple terms, ChatGPT is a virtual assistant that operates over an extremely vast amount of digital content. The service, undoubtedly, demonstrates the ability of Artificial Intelligence (AI) to push existing boundaries and achieves new levels of innovation. ChatGPT seemingly opens the door to “endless” possibilities. Whether it is to seek help with a resume for a job application, writing a speech, assessing a CV, debugging a computer programme, or doing some homework for primary school students just to name a few examples. This has led many organisations, including public institutions, to discuss the integration of such services in and the impact on their day-to-day operations.
From the ChatGPT provider’s perspective, the economic advantage it yields is probably unrivalled. At the time of this writing, it is probably not yet fully understood by authorities, by the public and maybe the service provider itself. The only guarantee is that there is a lot to gain through such a service. This is corroborated by Microsoft’s recent investment in OpenAI, the company behind ChatGPT. More importantly, we can expect that other tech giants will follow suit and release similar services over the coming months.
It can be argued that one of the key advantages such a service yields to its users is that they can appear more knowledgeable while having done only the absolute minimum of research at best and little to none at worst. Despite this benefit, the long-term impact on its users’ ability to formulate independent thoughts is probably not receiving enough attention.
Importantly, to a great extent, a service like ChatGPT has the ability to “control” what its users express, and can therefore influence their thoughts and beliefs, affecting the overall “Diversity of Thoughts”. One can argue that using a search engine as of today equally shapes one’s thoughts through the sometimes seemingly arbitrarily relevant content returned by the search engine. However, the key difference is that the search engine does not provide its users with a completely formulated thought but offers leads that they need to investigate by themselves.
To provide some context, in 2013, Deloitte released a research report in which it describes the concept of Diversity of Thoughts as:
The idea that our thinking is shaped by our culture, background, experiences, and personalities
It is often argued that a key element to a successful business or society is diversity – this is an idea that is well established in fields such as economics. This begs a multitude of questions. For example, can we continue to diversity our thoughts when most of the thought process, if not all of it, has been outsourced to a service like ChatGPT? Does the service incorporate enough of our individual background, experiences, and personalities to fully account for what we stand for? How much control will we give away by embracing such services? How much control and power does the service provider have by controlling such a service and what controls are in place to prevent them from potentially using it to their full advantage?
Regulatory bodies would eventually need to intervene not only to identify boundaries and limitations of these tools, but, more importantly, to clearly define, to the extent possible, the rights and responsibilities of parties involved. This will require answers to various practical questions, beyond the above existential questions: For example, who bares the responsibility for unintended consequences when one is acting on the advice of such services?
