Recently, Eric Jang contributed the following article to Smart Customer Service.
The advent of ChatGPT by OpenAI has made generative AI accessible to the public. In this short time, it has gained immense popularity as a cultural sensation, and although most individuals might encounter generative AI for the first time through this platform, it's just the beginning.
In fact, in just the short time that AI has gone mainstream, we've already seen companies like Microsoft announce Bing Image Creator mere weeks after its ChatGPT integration launch, and more recently Adobe's Firefly, its own creative flavor of generative AI.
With more and various iterations of generative AI on the horizon, it's poised to revolutionize numerous aspects of our everyday business and personal lives. And this potential for innovation is further amplified when multiple generative AI models are integrated, leading to even more advanced capabilities and enriched experiences.
If you think the last few months have been wild, just wait to see what's to come as each advancement builds on AI-powered advancement, opening new doors for creative innovation and revenue opportunities we've not even begun to consider.
Mastering AI with Baby Steps
As different sectors come to grips with the technology, they're focused on training their large language models (LLMs) with the industry-specific terminology and processes needed to ensure the specificity and accuracy of information the output delivers.
This is a vital step since before any language model can be useful to professionals in a specific industry it must first be trained on the relevant datasets. These ensure they are armed with the knowledge and information needed to provide accurate and focused outputs.
After all, you only get out what you put in, which is something we are already seeing with the waves of generative AI tools coming to market in terms of the output that can be delivered with more descriptive inputs and prompts.
In the financial services space, for example, we saw the March release of Bloomberg GPT becoming the first LLM trained to support natural language processing (NLP) tasks based on industry data and knowledge; assisting customers with sentiment analysis, named entity recognition, news classification, and question answering. This was quickly followed by IndexGPT by JP Morgan Chase. Similar initiatives are already under way in other sectors, including retail, healthcare, manufacturing, telecoms.
What remains missing, however, is a humanistic interface, one that can be provided by pairing the LLMs with a virtual human to take the interface beyond a keyboard and screen experience to one where you can interface with a lifelike human avatar.
This approach to integrating LLMs with lifelike human avatars has the potential to bring about a number of benefits for both businesses and society seeking to provide more personalized interactions.
Customer Service Potential
Customer service applications of this technology for virtual assistants are already in demand. According to Grand View Research, the intelligent virtual assistant market is expected to reach $14.10 billion by 2030. Combining the video generation models and text generation capabilities of LLMs opens a new world of possibilities when it comes to providing customers with 24/7 access to expert customer service through AI powered virtual agents.
By combining virtual humans with LLMs, businesses will be able to revolutionize how they interact with customers. Virtual humans will be able to simulate human-like conversation, facial expressions, and body language, creating more personalized and engaging experiences for the consumer. The more sophisticated they become, the more frictionless the experience and the more personalized the service delivered will be, enhancing customer retention, customer stickiness, and, of course, sales conversions and upsells.
By integrating with LLMs equipped with deep understanding and vast knowledge, the conversational capabilities of virtual humans will be greatly enhanced, enabling them to answer queries, make recommendations, offer tailored solutions with remarkable accuracy and efficiency, and ultimately lead to exceptional customer service, improved satisfaction levels, and strengthened brand loyalty.
Content Creation and Personal Assistants
Content creation and personal assistants are other areas where considerable traction is already being made and demand is high. From fake images of the pope wearing a puffer jacket to the fake Pentagon explosion wiping billions off the stock market, the impact of AI-generated content has been felt. However, while these are attention-grabbing and causing a stir in the news, there are many more positive applications of the technology.
Combining virtual humans with LLMs offers exciting possibilities, such as enabling virtual humans to serve as virtual content creators or presenters, providing new, engaging and interactive ways for audiences to experience content. Furthermore, they can also serve as personal digital assistants, helping with tasks and providing information. For example, imagine having your very own copy editor or content critique to provide feedback when and where you want it as you iterate on your creative projects.
Language barriers have been a significant hurdle throughout history to global interactions and language learning, and while we are not quite at the stage of the Star Trek Universal Communicator, it seems we are not too far off.
With companies like Meta announcing plans earlier this year to create translation software that works for everyone in the world and other initiatives in play, such as smart glasses that bring translated text straight to your eyeballs, essentially providing the wearer with translated speech to subtitles, it will be within our lifetime. Especially with the amalgamation of virtual humans and LLMs bridging these gaps.
With virtual humans able to converse in multiple languages, facilitating cross-cultural communication without the need for human interpreters, many of these barriers will be quickly and simply overcome. Furthermore, language learners will be able to practice conversational skills with virtual humans, receiving real-time feedback and guidance from LLMs, as well as ongoing and regular exposure to the new language they are learning.
It's evident that the convergence of virtual humans with large language models is paving the way for a world brimming with opportunities and benefits. This combination of human avatars and LLMs is re-writing the rule book on traditional human-computer interaction and opening new avenues for enhanced customer experiences, transformative education and training, immersive simulations, personalized content creation, and cross-cultural communication.
But this description still doesn't quite do justice to the opportunities and benefits this fusion of technology will deliver.
Yes, providing businesses with a better way to connect and engage with customers is great, and the progress we have seen from a content development point of view is astounding. However, when we look at the larger-scale issues this technology could address in education and cross-cultural understanding, the benefits are staggering.
By harnessing the lifelike presence of virtual humans and the vast knowledge of LLMs, we can unlock unprecedented potential across industries, domains, and the world at large.
As we continue to explore and refine this remarkable combination, businesses need to embrace the boundless opportunities it offers and work toward a future where humans and intelligent machines collaborate to create a more innovative, interconnected, and inclusive world.
Eric Jang is founder and CEO of Deepbrain AI.
