Artificial Intelligence (AI) can be defined as the simulation of the process of human intelligence by machines. For our purposes as writers, we would be working with areas such as speech recognition and language processing. Current applications in these areas include ChatGPT, AI Writer, Writecream, and CopyAI.
AI works by taking in large amounts of data, analyzing it for patterns, and using these patterns to make predictions. Chatbots are fed examples of text chats that enable them to produce realistic exchanges with people. (A prime example of this is ChatGPT. Another example that we all encounter is Chatbots that are used for responding to questions on a number of sites. If they cannot answer your questions, they may bring on a live Customer Service Representative or refer you to an e-mail address or phone number.)
There have been recent instances of AI bots going rogue. By that, I mean they gave inappropriate responses or took on aggressive personality traits. The prime example here would be Microsoft's Bing search engine. One reporter trying out the Bing AI was told that he was not happily married. Please note: the reporter pushed the chatbot by asking it to describe the darkest desires of its "shadow self."
The AI app Lee-Luda (by the South Korean company Scatterlab), in the form of a 20-year-old friendly female (released in December of 2020, began using verbally abusive language referencing LGBTQ+, people with disabilities, and feminists. This led to it being removed from Facebook Messenger 20 days after its launch.
In March 2016, Microsoft introduced the bot Tay, which was supposed to be an experiment in "conversational understanding." The more chats the bot engaged in, the smarter it became. What happened is that it started posting inflammatory Tweets and had to be removed from Twitter.
In December 2016, Microsoft released a chatbot named Zo (a successor to Tay). It was available on Facebook Messenger, GroupMe, and Twitter. Zo was known to give inappropriate responses and was taken down in April 2019.
To balance this out, I would also like to mention Apple's Siri, Amazon'z Alexa, and Google Assistant. These are all conversational bots that perform in very positive ways.
Why do AI's go rogue? It's the little things, like lack of human supervision, incomplete or biased data, and poorly designed algorithms. All of which can be overcome with a little bit of thought.
Some individuals, such as Elon Musk, founder of OpenAI, are promoting AI regulation and safety. What we can do to help prevent AI from going rogue is to implement ethical guidelines and ensure adequate human oversight and control.
How do AI's help us with our writing? They help us improve our grammar, writing, and style. They can help us research given topics, and they can give us outlines for whatever we want to write. They are a fantastic time saver.
Keep in mind that AI stands for Artificial Intelligence. The AI apps function in conversation mode so that the user feels comfortable with them. It is like having a conversation with another human being. You will get back what you put into a conversation with an AI app. I "talk" with ChatGPT on a regular basis. I am extremely polite and try to be as specific as possible. In return, the app is also extremely polite, and its responses are exactly what I am looking for.
They are not always right. We can get bad or skewed information from them. If what you are asking the app is extremely important to the work that you are doing, I would advise confirming it elsewhere. Treat the chatbot well, and it will treat you well.
References:
(c) February 2023 Bonnie Cehovet
Reproduction is prohibited without written permission of the author.
No comments:
Post a Comment