Chat-GPT

Chat-GPT launched in November 2022 with an upgraded version that became available to the public in March 2023. More people downloaded Chat-GPT in the first month of public availability than other well-known applications such as Facebook. Chat-GPT is a form of Natural Language Processing (NLP) that understands human language. Generative Pre-trained Transformers (GPT) involve the interaction of humans to “train” machines on Large Language Models (LLMs).

Despite the fanfare, Chat-GPT has limitations. It is “trained” with data through September 2021. Thus, anything that occurred subsequent to that date would not be in Chat-GPT’s LLM. Most notably, Chat-GPT suffers from “hallucinations.” Common to LLMs, hallucination occurs when an AI program offers plausible-sounding, yet inaccurate or non-sensical, responses to queries. Finally, there is concern Chat-GPT responses reflect the biases of its developers.

Reception of Chat-GPT has been mixed, with some immediate calls for regulation. At one end of the spectrum, Henry Kissinger co-authored an article in the Wall Street Journal that states “generative artificial intelligence presents a philosophical and practical challenge on a scale not experienced since the start of the Enlightenment.” The article also compared it to the invention of Gutenberg’s printing press, and it is clear artificial intelligence should enhance productivity.

On the other hand, The Guardian, a British newspaper, questioned whether any content found on the Internet after Chat-GPT’s release “can be truly trusted” and called for government regulation. Other critics expressed concern over the potential impact on democracy because of Chat-GPT’s ability to generate automated content. There are also reasonable fears productivity gains could come at the expense of displaced job functions.

Chat-GPT is available for public use. However, it is not the only publicly available GPT trained on LLM, and there will be more. Large, well-capitalized competitors, such as Google, are developing their own products. Over time, the public will gravitate to models they consider reliable and easy to use. Similar to the introduction of browsers in the early days of the Internet, investors will finance a plethora of business models that promise to leverage AI. The market has already inflated the values of AI-related stocks. Many business models will fail. Some will prove ill-conceived, perhaps even fraudulent. A few will thrive and grow. The process will be, as with many new technologies, Darwinian.

Chat-GPT and similar programs could disrupt traditional keyword search engines. Keyword searches organized the web (i.e., Google and Bing facilitate research, shopping, and knowledge acquisition). Search engines incorporate artificial intelligence, and they also operate in the background of other applications. They are an integral aspect of our relationship to digital information. Given the types of responses Chat-GPT generates, keyword searches could become obsolete. Generally, Chat-GPT could alter how we interact with information, and its impact could be broad (e.g., academia, education, business, commerce, finance, healthcare, and professional services). 

In the original “Jurassic Park” movie, a character pondered the ethics of technological advances. He argued that scientists, more focused on whether they could, never stopped to consider whether they should. The moral and ethical debate about AI will grow more intense, and there is no current consensus on how it should be deployed. There is much more to this story, but in the meantime, I assure you that Chat-GPT did NOT write this article.

Ray Ryan is the president of Patten and Patten, an investment management firm, and a registered investment adviser in Chattanooga. Ryan is a CFA charter holder, a member of the advisory board for UTC’s College of Business, and an adjunct professor of finance at UTC. He is a graduate of Princeton University, where he had the privilege of taking a course taught by former Federal Reserve Chairman Ben Bernanke.