Photo by Rolf van Root on Unsplash

Recently, everyone has been talking about ChatGPT, the language model developed by OpenAI. With its advanced capabilities, ChatGPT is revolutionizing the way we communicate with machines. It has the ability to understand natural language and respond in a human-like manner, making it ideal for a wide range of applications, from customer service to virtual assistants. ChatGPT’s language processing abilities are based on deep learning algorithms that analyze large amounts of data to generate responses that are both accurate and contextually appropriate. As a result, ChatGPT is quickly becoming one of the most sought-after tools for businesses and individuals looking to improve their communication with machines (and humans).

The fact is, ChatGPT is only the beginning of a revolution that will transform the way humans interact with technology. Artificial intelligence has come a long way since the first AI program was written in 1951. Today, AI systems can perform a wide range of tasks, from playing chess to driving cars. However, AI is still in its infancy, and there is much more that can be done.

One area where AI is having a big impact is in natural language generation (NLG). NLG is the field of study that focuses on how computers can produce understandable human language. NLG is what powers chatbots like ChatGPT, which can converse with humans in a way that feels natural and intuitive.

A number of other companies are also active in this area, aiming to compete directly with ChatGPT. For instance, Google’s T5. T5 (Text-to-Text Transfer Transformer) is a NLG model developed by Google that is similar to ChatGPT in terms of its architecture and capabilities. T5 is pre-trained on a large dataset of text and can be fine-tuned for specific tasks, making it highly versatile and capable of generating high-quality text in response to various prompts.

In the coming years, thanks to this competition we can expect to see NLG technology continue to advance. Already, systems such as ChatGPT can understand and respond to human language at a level that was once thought impossible. These systems will be used to power virtual assistants, customer service bots, and digital humans.

Storytelling

In 2019 I published a book titled Storytelling and AI, arguing that machines were fast learning to speak like humans; and would soon master the secrets of storytelling that until now, for tens of thousands of years, have been the exclusive domain of humankind. The advent of ChatGPT is a strong confirmation of that argument.

Indeed, ChatGPT can be a very useful tool in the field of storytelling in several ways:

• Idea generation. ChatGPT can help writers generate new ideas for stories by providing prompts or suggestions based on a given topic, theme, or character.
• Character development. ChatGPT can assist writers in creating and developing complex characters by answering questions about their backgrounds, motivations, and personalities.
• Plot development. ChatGPT can help writers develop plot twists and turns by suggesting different scenarios or plot points to keep the story engaging.
• Dialogue creation. ChatGPT can assist writers in crafting compelling dialogue by suggesting different phrases, idioms, and expressions that fit the tone and style of the story.
• Editing and revision. ChatGPT can be used as a tool for editing and revising a story by suggesting alternative words, phrases, and sentence structures that may improve the flow and readability of the story.

In other words, today ChatGPT can be a valuable ‘storytelling assistant’ for writers to enhance their creativity, develop their writing skills, and create engaging and compelling stories.

We can expect to see even more exciting progress in the coming years.  As AI technology continues to advance at a rapid pace, it’s important to consider what the future may hold for NLG software such as ChatGPT. Here are some possible developments:

• Improved conversational abilities. ChatGPT and other NLG software may become even more advanced in their conversational abilities. They may be able to understand and respond to more complex queries, making them even more useful for customer service, personal assistance, and other applications.
• More realistic responses. In the future, NLG software may become even better at creating responses that are indistinguishable from those created by a human. They may be able to incorporate more nuance, sarcasm, and humor into their responses, making them even more engaging and enjoyable to interact with.
• Personalization. As ChatGPT software continues to advance, it may become more tailored to individual users. It may be able to learn from previous interactions and create responses that are more personalized and relevant to each user’s individual needs and preferences.
• Integration with other technologies. NLG software may become more seamlessly integrated with other technologies, such as virtual and augmented reality. This could create even more immersive and engaging experiences for users, allowing them to interact with NLG software in more natural and intuitive ways.
• Story invention. ChatGPT is already able to create stories on the basis of simple prompts. This ability will evolve making NLG software more and more able to invent meaningful narratives, even in the form of novels or movie scripts.
• More applications. As ChatGPT software becomes more advanced and versatile, it may find applications in a wider range of industries and contexts. For example, it could be used to generate personalized news articles, ads, financial reports, and other types of content.

In brief, as AI technology continues to advance, NLG software in general, and ChatGPT in particular, will likely become even more powerful, versatile, and useful for a wide range of applications.

Metaverse and brand storytelling

In early November 2022, just weeks before the launch of ChatGPT, I published a book titled Brand Storytelling in the Metaverse. There, I noted that new NLG software will soon be integrated with AI avatars in the metaverse, particularly with brand avatars interacting in a dialogue with customers in virtual worlds. This development is now going to be enabled by tools such as ChatGPT.

By integrating ChatGPT with brand avatars, customers could have natural language conversations with artificial humans in real-time, allowing for personalized and responsive interactions. This could help to create a more engaging and interactive customer experience, as customers would feel like they are having a real conversation with the brand rather than simply receiving canned responses.

One potential benefit of using ChatGPT with AI avatars in the metaverse is that these artificial entities could provide customers with a much wider set of answers to their queries. For instance, a digital assistant for a car maker could be knowledgeable about all details of a 200-page instruction manual. A brand avatar empowered by ChatGPT could also help to reduce the workload of human customer service representatives. By automating some of the simpler interactions, such as answering frequently asked questions, human representatives could focus on more complex issues and offer more personalized assistance to customers.

However, there are also potential challenges associated with integrating ChatGPT with AI avatars in the metaverse. For example, as with any AI technology, there is a risk of discriminatory language if the training data is not diverse and representative. Additionally, it may be challenging to ensure that the avatar’s body language and facial expressions are conveying the intended tone and emotion of the conversation.

Clearly, there are still significant hurdles to integrating ChatGPT with AI avatars in the metaverse; but there are also many potential benefits. By creating more engaging and immersive customer experiences, brands will be able to foster deeper connections with their customers and increase customer loyalty.

Jobs

Many people worry that AI systems such as ChatGPT could ‘steal’ a lot of  jobs in the coming years. Undoubtedly, there is a risk that ChatGPT and other NLG technologies will automate some jobs that currently rely on human language skills, but it is important to consider the nuances of this issue.

On the one hand, NLG technologies like ChatGPT can be used to automate tasks such as customer service, content creation, and data analysis, which traditionally require human language skills. This could lead to job displacement in certain industries, particularly those that rely heavily on repetitive or routine tasks. For example, ChatGPT could be used to automate some aspects of content creation, such as generating news pieces or social media posts, which could potentially displace human writers and editors.

On the other hand, NLG technologies can also create new opportunities for human workers, particularly in areas where NLG can enhance or augment human skills. For example, ChatGPT could be used to help journalists and writers generate more content in less time, freeing them up to focus on more complex or investigative reporting. Additionally, ChatGPT could be used to help language teachers create more personalized learning materials for their students, or to help doctors and medical professionals create more accurate and efficient medical reports.

Furthermore, it is important to note that NLG technologies are not yet advanced enough to completely replace human language skills. While they can generate human-like text, they still lack the creativity, nuance, and subtleness that human language use requires. Human language is also highly contextual and often requires emotional intelligence and cultural knowledge that machines may not be able to replicate.

So, there is actually a risk that NLG technologies like ChatGPT could displace some jobs that rely on human language skills. However, there are also many potential benefits and opportunities for human workers. It is important for policymakers and industry leaders to consider the potential impacts of NLG technologies on the workforce and to ensure that the benefits are widely shared and that the transition to an automated future is managed in a responsible and equitable manner.

Other risks

Job displacement is not the only risk associated with a wider use of technologies such as ChatGPT. Here are some other potential dangers and some suggestions on ways to control and limit them:

1. Bias. NLG software can perpetuate biases present in its training data, resulting in offensive language or conclusions. To control and limit this risk, it is important to ensure that the data used to train the NLG software is balanced and representative of different groups. Additionally, regular audits should be conducted to ensure that the NLG software is not perpetuating any biases.
2. Misinformation. ChatGPT may inadvertently generate false or misleading information. To control this risk, it is important to incorporate fact-checking mechanisms into the NLG software. This can involve cross-referencing with reliable sources and verifying information with human experts.
3. Privacy. NLG software may collect personal information about users, potentially raising privacy concerns. To limit this risk, it is important to adhere to established privacy guidelines and regulations, such as GDPR. Additionally, users should be provided with clear information about what information is being collected and how it will be used.
4. Dependence. Technologies such as Chat GPT may become so widely used that humans become overly reliant on them. This can lead to a loss of critical thinking skills and the ability to evaluate information independently. To control this risk, it is important to encourage the development of critical thinking skills and provide education on how to evaluate information effectively.
5. Security. NLG software may be vulnerable to cyber attacks, potentially leading to the compromise of sensitive information. To limit this risk, it is important to incorporate robust security measures into the NLG software, such as encryption and regular vulnerability assessments.

Overall, while tools such as ChatGPT offer many potential advantages, it is important to be aware of the potential risks and take steps to avoid them. By addressing these risks, we can ensure that NLG technologies are used in a safe and responsible manner that benefits society as a whole.