ChatGPT: (in)Human Shortcomings of AI
Conceived from the astute minds of Silicon Valley, ChatGPT has left the entire internet spellbound by being just too good. At first glance, the chatbot’s capabilities not only uncannily human, but immeasurable as well. From passing MBA Exams, to writing school essays, coding websites, even composing propaganda pieces.
For all intents and purposes, ChatGPT has all the makings of a real nasty industry disrupter. Its ever-growing popularity, coinciding with very little in the ways of competition has worked in the chatbot’s favor. Its mere presence has sent Google’s parent company Alphabet Inc. in crisis mode. According to an insider report by the New York Times, CEO Sundar Pichai has declared a “code red” and accelerated its own rival AI development. Not so bad for being the new kid on the block.
But ChatGPT goes even further beyond just industrial concerns. Automation is the current dirty word being thrown in the everyman’s workplace. A new generation of artificial intelligence–just like ChatGPT–could potentially upend the job landscape.
So, can ChatGPT render the human race obsolete? Let’s take a closer look.
What is ChatGPT?
Introduced to the world in November 2022, ChatGPT is an AI-powered language model developed by OpenAI. The chatbot uses deep learning algorithms to generate human-like text responses based on the input it receives.
However, ChatGPT is not the first chatbot to be developed. The history of chatbots dates back to the 1960s when the first computer program designed to simulate conversation with a human user was created. These early chatbots were primitive by today’s standards, but they represented a significant milestone in the development of AI and natural language processing.
In the decades that followed, chatbots continued to evolve and improve. Now we have Google Assistant, Alexa, and Siri. So what makes ChatGPT hits different than all of its predecessors before it?
ChatGPT has been trained on a massive amount of data, which includes a diverse range of topics and conversations, allowing it to have a broad range of knowledge.
It’s not an omniscient, however. Nor always correct. The prompt page warns users that ChatGPT, “may occasionally generate incorrect information,” and, “may occasionally produce harmful instructions or biased content.”
Unlike traditional chatbots, ChatGPT is not connected to the internet and does not have access to external information. Instead, it generates responses based on the data it has been trained on. This data set contains a wide but again, limited range of texts from various sources, such as books, articles, and websites. ChatGPT also has limited knowledge of events that occurred after 2021.
Conclusion: Is the threat of mass AI take-over upon us?
In short, nope.
The fear of replacing the human element for cold, unfeeling technology goes beyond the fictional sphere of Sci-Fi. But realistically, we still have a long way to go before we reach the likes of Harlan Ellison’s AM. Machine learning is, at best, not yet out of its metaphorical cradle.
Originally published at https://www.redwoodcreative.com on March 1, 2023.
