AI has taken the world by storm. How is it being implemented and what effect does it have on written media?
When the topic of AI is brought up in conversation, it tends to spark a bit of apprehension. You got the typical mind control, human extinction, all that I,robot stuff. But one of the more common concerns relates to the realm of written content. Keep reading on as we dive into the details.
What is AI?
The term AI, or “artificial intelligence” has been around since the 1950s. It’s defined as a machine that possesses the ability to perform tasks commonly attributed to humans. I'm happy to say that this includes pop culture icon, R2-D2. Nowadays when people mention AI, they’re usually referring to specific tools known as chatbots.
These are programs that understand and respond to typed-out prompts, allowing its user to feel like they’re having a conversation with a real person. Chatbots also have the capability to create written content on demand. This includes stories, scripts, essays, and of course, blogs.
ChatGPT is the name that usually makes an appearance when discussing AI. (The GPT standing for Generative Pre-Trained Transformer). It’s by far the most popular chatbot, having over 180 million current users. Almost 50% of all businesses have implemented this program in some capacity, with many others intending to follow suit.
How does ChatGPT work?
Launched by OpenAI in November of 2022, ChatGPT was in development for around 4 years. It was originally created using the GPT-3.5, which is a natural language processing model.
ChatGPT was given access to a crazy amount of online data, (570 gigabytes to be exact). This allows it to predict which words come after a given prompt. So, despite what you might think, the AI doesn’t truly retain its information. It’s essentially just very good at guessing.
…This also happens to be the method I used in high school.
That said, ChatGPT can retain knowledge regarding users and their discussions. It can currently remember up to 64,000 words within a single conversation.
What really sets ChatGPT apart from other AI was the way it was taught. Trainers would use Reinforcement Learning from Human Feedback, or RHLF for short. Through this punishment-reward method, the AI had each of its responses ranked, with the human-like responses being favored.
Does Google know if a blog is written by AI?
Google’s ability to recognize AI within blogs depends heavily on how it’s being implemented. Using ChatGPT for inspiration or creating rough drafts has an extremely low chance of being noticed. However, something that was simply copy and pasted from an AI would be detectable by Google.
Most AI use NLG or Natural Language Generation when formatting their content. This pattern is what causes it to be recognized by Google’s algorithm. That said, this doesn’t mean AI-written work will automatically be penalized or taken down.
Google cares a lot less about AI-written content, so long as it’s beneficial and informative to humans. What they do intend to crack down on is automatically generated content.
What is Auto Generated Content?
Automatically generated content is created by AI, but with almost no human oversight. It is usually made with works from other sites and littered with keywords. Auto generated content does not have human viewers in mind, with its sole intention being to exploit Google’s system. This results in the work being almost unintelligible to the average reader, like most of my high school projects.
Conclusion
Advancements in AI have skyrocketed in the past few years. With this sudden shift, it can be difficult to understand where and how to implement it ethically. But if we use proper guidelines and supervision, I believe this technology will only enhance human work, not replace it.
And I’m definitely not just saying that because I’m a robot…
