With the emergence of ChatGPT, Bard, and other Artificial Intelligence (AI) chatbots, understanding AI is important. There is a craze over the possibilities of AI.
By understanding what AI is and isn’t, people can judge for themselves what is just hype, and how much to fear the future.
What exactly is AI?
Artificial Intelligence, is defined in the Oxford Dictionary as “The theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.”
This means a computer program that makes decisions that humans would otherwise be required to make. It needs to be able to make correct decisions with math, logic, reasoning, and research skills. Programs of this nature have been around for decades.
A search engine that guesses what is being looked up is a good example of AI decision making. Games such as The Akinator, about a genie who uses a series of logic based questions to determine who the user is thinking about, have been around for years. Google Translate is a service that millions of people use every day.
ChatGPT, Google Bard, and AI Chatbots
ChatGPT, Bard, and other chatbots are programs that users can ask questions and receive researched, and articulate answers. While the answers are usually correct, there are many examples of AI producing wrong answers.
Google Bard will generally provide links to the sources used in an answer. ChatGPT currently does not have that ability. This makes it difficult to check the accuracy of the answers.
Most people ask these AI to write essays, stories, songs, poems, letters, and other forms of text documents. Provide a topic and the type of response you want, and it attempts to generate a text.
It is easy to see how this can be used to cheat in educational settings, so there are also websites, such as GPTzero that attempt to detect whether the text was generated by an AI model. Alternatively, this article, 5 Ways to Detect Text Written by ChatGPT and Other AI Tools, from PC Mag discusses ways to tell whether text has been written by AI or a human.
Google’s Bard, a chatbot competing with ChatGPT, is considering watermarking content to prevent misuse. This article, Google: We’ve Made Bard Even Smarter Using PaLM 2 AI Model, from PC Mag contains technical information about Bard and its development.
Examples of AI in action
In this example, we asked ChatGPT to write us a 100 word poem about frogs. This is what it came up with.
Furthermore we asked ChatGPT to write us a Blog Post about Chat GPT, this was part of its response:
AI chatbots can also write computer code based on user prompts. Coders and app developers could use it as a tool to help them string together their ideas using generated subroutines. The program cannot build an entire application, so users are still required to understand app development.
In this example, we asked ChatGPT to write us a simple HTML website that sells cars.
Although the code it provided is basic HTML, it can achieve much more when directed with a specific purpose.
Only about 14% of adults in the US have actually tried using ChatGPT according to this article by Pew Research Center, A majority of Americans have heard of ChatGPT, but few have tried it themselves. It also provides more statistics on chatbots presence in the USA in general.
Chatbots, Anthropomorphism, And Human Control
With the fear of AI sentience, users are concerned about how chatbots portray themselves to users.
A recent article from The Boston Globe touched on a Google Engineer claiming that their chatbot, Bard, seemed sentient. He was fired because the engineer knew there was no sentience in Bard, and felt he was misrepresenting their product to cause alarm. The article described ChatGPT’s more realistic approach of not pretending to express emotion, whereas Bard does the opposite. Many think Bard’s attempt to express mannerisms are irresponsible in that regard.
Ideally it is important to understand that a code generated notion of human emotion, is far from an AI program gaining sentience, and even farther from it lashing out on its own. Many are curious as to what the future holds. This well thought out piece by a man named Stuart Russel, How can humans maintain control over AI — forever?, touches on how we can maintain control on the AI programs we create, not just now, but forever.
It’s clear there is no stopping the freight train that is advancing technology, but hopefully we as humans are intelligent enough to prevent it from being our demise. We will continue to see how this plays out.
If you have specific questions about understanding AI, or if our terminology isn’t clear, please use our contact form to send us a message. Thank you!