Writing has been around since the dawn of time, and humans have developed to be very good at it. Today, writing–much like many other industries–faces challenges from artificial intelligence that seek to replace humans with computers and algorithms. Can AI write as well as a human? The answer isn't an easy one, but it will provide insight into how AI can affect the industry in the future.
Artificial Intelligence is simply the simulation of human intelligence processes by machines, especially computer systems. The goal of artificial intelligence research is to create machines that process information and perform tasks in ways that are similar to humans. Some very well-known examples of this are Siri, Alexa, Google Translate, and even self-driving cars. But what about writing?
A new study from Stanford University found that AI can write text at the same level as humans. Although it's not perfect yet, it's getting closer. But we have to remember that for now, AI can only write one specific type of text: technical or formal content. In other words, if you give your iPhone an order for a coffee on your way to work or tell your Amazon Echo where you live so it can set up an appointment reminder for you, then technically, these are written by computers. However, when you ask Alexa about your day or want to know how much money you have in your bank account, these questions would be processed through natural language processing (NLP) which is currently done by people instead of machines. The general idea behind NLP is figuring out what a person means when they say something and translating that into another form.
To understand how we taught machines to read and write, it's important to first understand the history of computing. The machine language used by computers is binary code. One basic way of representing this code is in the form of zeros and ones that can be represented in a string on an 8-Bit computer (since each bit can hold either 0 or 1). Early computers used punch cards and paper tape for input and output. The input was typically in binary code from a card reader and output from another card punch or line printer. Programmers would translate the human readable source code into sequences of zeros and ones, which were punched onto cards using holes or typed into the computer using Teletype terminals with keyboards. These programs are often referred to as handwritten since someone created manually them. As you might imagine, mistakes could happen when translating text into binary code. What made these errors more difficult to find was that programmers didn't have any good tools available to help them debug their work—no editors with syntax highlighting or advanced debugging features like those found in modern IDEs.
Machines are already able to do many things that were once exclusively human, and in short, they will be able to do even more. In some cases, this might not have any negative impact, but in others, it could have disastrous consequences. One example of this is when machines can do jobs better than people. This has been happening for decades, but recently many jobs that involve writing have been done by machines instead of humans. Many people believe that since computers don't require breaks, don't need to be paid well, or need healthcare benefits as humans do, it makes sense for them to replace them in these positions. However, there is a large amount of debate on whether AI can write as well as humans and if their style would become too uniform due to the lack of diversity among writers. As the number of people who work from home increases, it becomes increasingly difficult for companies to find new employees with diverse backgrounds and experiences. Some experts argue that companies should hire humans over machines to retain important skills that cannot be replaced with technology and also make sure their communications aren't too impersonal.
Machines can learn to be creative writers by copying the writing of other humans. In this way, they can develop a style that people might not even be able to tell is the work of an AI. -This is a problem because humans are starting to feel threatened by the idea that machines will soon be able to do everything better than them. However, since computers can only copy what we have already written and they cannot come up with ideas on their own, they can't have true originality in any form of artistry or creativity.
-While machines are getting better at reading and understanding our language, people still have a more nuanced understanding of things like tone and humour. So when you're trying to create something meant to make people laugh, you need a human touch. You also need human input when you want your story to be based on something real or historical events. For example, if I wanted my story about the Titanic sinking to include dialogue from eyewitnesses who were there during the incident, then I would have needed help from someone who understands how people talk so that I could get their dialogue right. If I had tried to write it without talking to anyone involved, then no one would believe that such detail came from anything but my imagination. When it comes to creating stories with human-like emotions, people are the best bet. Computers cannot feel emotions and understand how we react differently to certain situations, depending on where we grew up or which culture we live in. That said, using a computer wouldn't hurt if you just want to write quickly!
Therefore, AI writers are not the same as human writers. The two forms of writing do not have the same effect on readers and the techniques used to create them differ. AI-generated text may be able to mimic human writing, but it can never be exactly like it. Both types of writing will continue to exist because they each serve different purposes for different audiences. Some people may prefer one form of writing over the other; others will enjoy both. With further technological development, the gap between AI and human writers should close even more.
About the Creator
Manik Roy
Writer | Photographer | Handwriting Expert | AI Artist
Reader insights
Nice work
Very well written. Keep up the good work!
Top insight
Expert insights and opinions
Arguments were carefully researched and presented
Comments (1)
I think AI may actually write technical articles and documentation better than humans but it will never have any lightbublb moments. As for fictions and poetry I will always prefer human authors.