Conversational AI: a new frontier with endless possibilities
Written by Ilankir Matungulu, Venture Associate at ChingonaVentures
In the past few months, there’s been a surge in conversations online about conversational artificial intelligence (AI). These have led many big players in the tech industry to invest in, or launch technologies to strengthen their presence in the space. The product driving the conversation right now is ChatGPT. After the platform launched on November 30, 2022, it has exploded in popularity, reaching 1 million users in 5 days and hitting over 100 million users by January 2023, making it the fastest-growing platform on record. This explosive growth has led many to wonder how ChatGPT, and similar products, can help change the landscape of human interaction and efficiency.
ChatGPT is a leading-edge natural language processing technology developed by OpenAI. Natural language processing (NLP), as a discipline, is at the intersection of linguistics, computer science, and artificial intelligence. It is focused on building digital systems that are capable of understanding the contents of written documents and human speech alike while including contextual and cultural nuances. ChatGPT is based on a deep learning algorithm that uses neural networks to generate human-like responses. Before diving into how ChatGPT works, it’s important to explain deep learning and neural networks. It’s easiest to think of deep learning as a type of technology system that helps computers learn to do things by practicing. I could use the analogy of a puzzle: a computer is given pieces of a puzzle (in this case, data) and it puts them together to figure out how to do something, such as recognize an image of a cat. It’s able to do this with a special model known as a neural network, which is essentially a big brain that learns from the puzzle pieces. In the same way the human brain helps us learn and understand things, neural networks help computers learn and understand things.
ChatGPT is built on top of a system that uses massive amounts of data to train and learn to produce coherent responses to a given prompt. Users can log on to ChatGPT and type in a question or a request. That input is subsequently analyzed, and the platform generates a response based on the learned patterns it has developed from the massive amount of data it has been trained on. The more conversations people have with ChatGPT, the smarter it gets. This system has been trained on a diverse set of text, including books, articles, social media posts, and chat logs. This process has enabled ChatGPT to understand many of the nuances of the English language and generate responses that are not only accurate but also coherent and natural. As a matter of fact, some people think the responses are so natural that people can be easily fooled into thinking a human wrote an article. ChatGPT has been used to write poetry, blog posts, presentation outlines, and more.
ChatGPT could have many implications for venture capital, from both sides of the funding interaction. On the investor side, ChatGPT could be used by VC firms to improve and accelerate their due diligence processes. The platform could be used to analyze large amounts of data to identify patterns or insights that could potentially be overlooked by humans. This could reduce the risk of investing in companies with un-viable business models. ChatGPT could also potentially be used to identify emerging trends and technologies, helping firms to find and evaluate emerging companies more quickly. Some VC firms are already looking at ways to implement ChatGPT into their current fact-checking processes. On the founders’ side, ChatGPT could be used to better identify investors that are a good fit for their business ideas and for their desired growth timeline. By analyzing large datasets, ChatGPT could readily identify investors that have invested in similar companies or have a strong interest in a given industry. This could help accelerate the matching process for founders looking for funding.
ChatGPT has very significant implications for the future of work. The most obvious effect is that ChatGPT could quickly learn to manage routine and repetitive tasks, creating efficiencies that would give employees more free time to focus on more complex or creative tasks. For example, it could handle more basic customer service inquiries and shorten wait times. ChatGPT is already being used to help debug code. The platform can not only find errors but also provide an explanation to the developer so that they can understand why things went wrong. These solutions only scratch the surface of what is possible with ChatGPT. Indeed, ChatGPT could potentially be used to improve job-related education and training, by providing detailed feedback and coaching to employees as they look for opportunities to upskill. It could also be used to build more personalized career development programs and give custom recommendations on professional development opportunities for employees based on their desired career goals. In the healthcare space, ChatGPT could realistically be used to help to design scheduling systems that minimize employee burnout, an issue the industry has been struggling with since the COVID-19 pandemic.
At this point, it’s important to emphasize that ChatGPT is a tool and is not yet at a stage where it can effectively substitute human experience and expertise. The product has several limitations, the most important one being that its model is trained on large, but not all-encompassing, datasets. This means that it is not all-knowing and has limited-to-no knowledge of information outside of its training set. The product is not at a stage where it can create purely original content, and has limited contextual understanding, especially on topics outside of its training data.
Here at Chingona Ventures, we already have a few portfolio companies leveraging ChatGPT’s technology to improve their processes, but we know that this is only the beginning. As investors, we understand that the technology still has a lot of room to grow, but we see the Future of Work vertical expanding at an accelerated pace with this latest innovation. There are already many interesting startups in this space such as Hugging Face, a chatbot application designed to facilitate artificial intelligence-powered communication, Textio, an augmented writing platform designed to help companies in writing effective job listings, and Viable, a user software designed to use historical qualitative data to answer new questions instantly. We’re excited to see the new investment opportunities that will arise as ChatGPT, and similar products, slowly move into everyday adoption.
