MISTAKES, SOURCES, AND CITATIONS

One of the challenges of ChatGPT is how to check for mistakes. ChatGPT is a Large Language Model (LLM) that uses multiple information sources. When answering a question, it creates the answer in the form of an article produced without attribution, citation, or benefit to those who published the original content [5].

This raises two issues.

Firstly, checking the accuracy and truthfulness of a document becomes problematic if it has no citations or sources. How do we know if what ChatGPT has produced is accurate and truthful if the source documents cannot be checked? The other aspect is that not all source documents have the same authority. For example, peer-reviewed documents are generally considered to have greater veracity and authority than non-peer-reviewed documents or an article by an individual on their personal blog site. When scraping the web for information, LLM does not differentiate between the authority of the sources they are scraping.

The second issue has to do with plagiarism and intellectual property laws. Intellectual property laws are designed to protect the intellectual property of an individual or organisation. Yet the law has not kept up to date with technological advances, particularly regarding artificial intelligence such as ChatGPT. The application of Intellectual Property Law to AI remains untested and potentially problematic.

Given these potential issues, businesses should be careful in relying on the information, particularly complex information produced by ChatGPT.

CHATGPT AND THE OTHER SIDE OF THE COIN

I wrote above that ChatGPT could free up staff from performing routine and mundane tasks allowing them to focus on creative, strategic work. While that is one side of the coin, the other is that staff can grow reliant on AI tools to do the work for them.

Creativity requires discipline, routine, and a commitment to persevere through failure. In the heat of the creative moment, inspiration and determination flower into an idea, concept or event that captures our imagination and other people's imagination.

Why spend time and energy being creative and original and persevering through failure when ChatGPT can produce an article or a social media post in minutes? 

Rather than ChatGPT giving staff the time to engage in strategic and creative work, staff may take the path of least resistance and become lazy by relying on it to do their work.

Does it matter if that happens?

There are two reasons why it does.

Firstly, when anxiety, depression and boredom are increasing in society, and the links between mental health and social media are documented, we need to do what we can to improve our mental health. One way is to be conscious of when we use technology and what for. Using technology and AI because it is easier is not necessarily constructive for our mental health.

Secondly, according to a study by Microsoft, people have an attention span of eight seconds, a decrease from twelve seconds in 2000 and a second less than a goldfish, which has an attention span of nine seconds. Furthermore, their study claims that the human attention span decreases by 88% yearly. The Statistic Brain Research Institute also confirmed this study. 

We need our attention, concentration, and ability to think deeply about complex and ethical issues. Thinking is like a physical muscle that must be used to be effective. Over-reliance on AI can reduce our ability to think effectively and in-depth.

While ChatGPT, like other forms of AI, has its uses in the workplace. We need to think about where it is used, how it is used and how it is monitored and checked. Businesses thrive best when they marry AI's strengths with their staff's effectiveness and strengths to create a constructive work environment to meet the needs of their clients.