Interest in AI technology and, more specifically, OpenAI’s ChatGPT product has skyrocketed in recent years.
People are looking for information about both topics.
Source: Google Trends
Millions are writing about ChatGPT across the web…
…and talking about it in various communities.
Interest shot up almost immediately. This is the first couple of weeks after ChatGPT launched to the public.
And as you can tell from the graphs, all of this happened quickly.
Whether your X and LinkedIn feeds have been persistently inundated with threads and posts about AI in general and tools like Claude, DeepSeek, and ChatGPT (like mine), or you’re just stumbling on the topic, you may want answers to two questions before investing your time and energy into learning ChatGPT:
Is ChatGPT specifically likely to be an enduring product?
What does it actually do and what can you personally use it for?
In this article, I’ll help you answer these questions by telling you:
What ChatGPT is.
How it works.
Who built it and is behind the technology.
Why it’s important for SEOs specifically.
Some of the current and likely future uses for it.
What is ChatGPT?
ChatGPT is an AI-powered chatbot created by OpenAI that can be accessed at https://chatgpt.com/.
As of this writing, there are a few different product offerings:
A free version of the tool, providing access to ChatGPT 4o mini (a specific model).
A Plus plan for $20 per month, which includes extended limits, access to more advanced ChatGPT models (o1 and o1 mini), scheduled tasks, custom GPTs, and limited access to Sora for video creation.
A Pro plan for $200 per month, offering unlimited access to all Plus features, advanced voice capabilities, higher limits for video and screen sharing, an advanced version of the o1 model, and access to Operator, a feature that can perform tasks in a dedicated browser.
Out of the box, the free version’s interface is simple, with an empty dialog to enter a prompt.
The tool can perform various tasks and return text, files, images, and videos in response. Some examples of tasks ChatGPT can execute include:
Answering questions.
Writing things like ads, emails, paragraphs, whole blog posts, or even college papers.
Writing, commenting, or marking up code.
Changing the formatting on a block of text for you.
ChatGPT launched in late November 2022, on the heels of AI Content Generator Jasper.ai receiving $125 million in funding at a $1.5 billion valuation earlier the same month. The tool reached a million users in less than a week.
But each session has a specific cost associated with it.
In the interest of helping fund those costs (and further growth), Microsoft invested $10 billion in OpenAI at a $29 billion valuation.
This move, combined with ChatGPT’s growth and word of mouth, might have fueled Google’s subsequent reported concerns about ChatGPT as a possible threat.
Most recently, OpenAI raised another $6.6 billion at a valuation of $157 billion in October 2024.
It’s been rumored that OpenAI is in talks to secure another $40 billion in funding at a $340 billion valuation (on the heels of new competitor DeepSeek, which is rumored to have spent only $5.5 million).
Now that you understand what ChatGPT is, it’s equally important to learn how it works, who built it, and the goals and motivations behind its development.
How does it work and how was it trained?
That said, when using tools like ChatGPT, you will want to know where the information it generates comes from, how it determines what to return as an answer, and how that might change over time.
That way, you can understand what level of trust to put in ChatGPT answers and output, how to craft your prompts better, and what tasks you may want to use it for (or not use it for).
Before you start using ChatGPT for anything, I strongly recommend you check out OpenAI’s blog post about it and become aware of some of its failures and limitations.
There, they have a nice graphic explaining how it works and a more in-depth explanation.
AssemblyAI also has a detailed third-party breakdown of how ChatGPT works, some of its strengths and weaknesses, and several additional sources if you’re looking to dive deeper.
One of the most important things to remember about how ChatGPT works is its limitations. In OpenAI’s own words:
“ChatGPT sometimes writes plausible-sounding but incorrect or nonsensical answers. Fixing this issue is challenging, as: (1) during RL training, there’s currently no source of truth; (2) training the model to be more cautious causes it to decline questions that it can answer correctly; and (3) supervised training misleads the model because the ideal answer depends on what the model knows, rather than what the human demonstrator knows.”
Another that’s important to highlight:
“While we’ve made efforts to make the model refuse inappropriate requests, it will sometimes respond to harmful instructions or exhibit biased behavior. We’re using the Moderation API to warn or block certain types of unsafe content, but we expect it to have some false negatives and positives for now. We’re eager to collect user feedback to aid our ongoing work to improve this system.”
ChatGPT was fine-tuned on a GPT model that completed training in June 2024, meaning it won’t have knowledge of events that occurred after that unless prompted to access the web.