Human Intelligence Will Still Drive AI Innovation

Since its launch in November last year, the OpenAI-developed ChatGPT language model has taken the world by storm. Its use of deep learning techniques to understand language and generate responses based on the input it receives has pushed artificial intelligence (AI) into the mainstream. But according to PBT Group, human intelligence will always be an essential component behind data analysis and the development of innovative technologies.

One area where the relationship between technology and people has been thrust into the spotlight is human resource management. Nicky Pantland, Data Analyst at PBT Group, says ChatGPT is a tool that can potentially aid businesses when it comes to the monitoring, management, and skills assessments of their employees.

“Because ChatGPT can analyse enormous amounts of data quickly, it can assist with identifying trends, patterns, and insights related to employees. But the need for human oversight and understanding remains important. Ultimately, the human workers must ensure that the information received from the AI is suitable and accurate for use,” says Pantland.

This requires companies to balance the benefits of AI with the need to preserve human autonomy and creativity. Because ChatGPT relies on data to learn and generate responses, a human must still ensure that the data consumed by the model is diverse and complete.

“Should any bias or omission become evident, the human operator is needed to identify and address this. Different cultures have different considerations regarding what is important or considered to be ‘normal.’ Only a member of that culture would be able to appropriately train the model on the social norms and the expectations of its people,” adds Pantland.

From a legal perspective, aspects of ChatGPT could impact business operations. One of the most significant of these is the regulations and laws governing the collection, storage, and use of data. There is also the potential for legal ramifications should the use of ChatGPT create an underlying gender or racial bias which could impact customers and cause reputational risk or damage to the business.

It is the technology behind ChatGPT that has catapulted it into the spotlight. Unlike other AI chatbots that are rule and script-based, ChatGPT uses a generative model. This means it can generate new text based on the input it receives.

“This feature allows ChatGPT to be useful in areas like customer support, where the output needs to be customised to the specific need of the user. Another key differentiator is that ChatGPT is a transformer model, which means it uses a mathematical technique called self-attention to process the input and generate the output. This allows it to handle longer sequences of text and generate more natural and coherent responses,” says Nathi Dube, Director of PBT Innovation at PBT Group.

Dube believes that while the human role will remain paramount, AI solutions like ChatGPT can deliver much-needed improvements to business efficiency.

For example, by providing accurate answers to customer queries, ChatGPT can help companies build and maintain stronger relationships with end users. The tool lets businesses customise model responses and integrate them into their existing systems to enable accurate real-time responses.

“ChatGPT can also be used for data cleansing where natural language queries are issued to identify errors in the dataset and provide suggestions to correct them. The tool can perform data exploration where users can ask questions and get responses that may help them better understand their data,” adds Dube.

As ChatGPT understands various programming languages, it can assist developers in writing and debugging their code, which can greatly improve productivity. The tool can also help create various content on various topics to equip companies with the means to quickly write newsletters and other internal communications.

“The technology can certainly help data-driven companies to stay ahead of the curve without the overheads and limitations that come with developing chatbots in-house. Forward-looking industry leaders are embracing ChatGPT and identifying new use cases to help improve operational efficiencies and thereby gain a competitive advantage,” says Dube.

However, businesses must ensure that they have the appropriate measures in place to mitigate legal, ethical, and people-centred concerns. To ensure that the use of ChatGPT is responsible, companies will rely on human experts to manage the training and monitoring of the software to ensure it is diverse and representative.