Even though it's been a few years since ChatGPT's 2022 debut, odds are you're still getting started on your AI journey. We're all learning more about it every day, and knowing how the tech works can help you get the most out of your conversations.

While ChatGPT is easy to use on the surface, many complex computations that are customized to each user are happening behind the scenes. Large Language Models (LLMs) rely on gigantic AI neural networks that can process and generate human-like text, analyze images, and even speak on their own. Here's how it works.

What Is the Technology Behind ChatGPT?

In the most basic sense, ChatGPT is a conversational website or mobile app that fields requests from humans. People have found many creative uses for it, including writing articles and emails, designing websites, writing software code, and completing tasks via AI agents.

While ChatGPT is the most popular AI chatbot today, others you may hear about include Google Gemini, Perplexity, and Anthropic's Claude. They're all trained on vast quantities of data, which "teaches" them how to interact with humans in a convincing way, as if they are humans. But they're more like aliens (or toddlers) constantly trying to learn how to be a human adult. They also want to be well-liked; OpenAI recently had to roll back a ChatGPT update when it became too sycophantic.

This learning process happens by feeding the chatbot data, largely from the internet (Wikipedia is a big one), including copyrighted books, YouTube videos, and other original materials, prompting lawsuits in some cases. Chatbot aliens are hungry for as much information as possible so they can keep performing better.

The model learns by taking a chunk of text from the data (say, the opening sentence of a Wikipedia article) and trying to predict the next token in the sequence. It then compares its output with the actual text in the training corpus and adjusts its parameters to correct any mistakes. By doing this over and over across a very large body of text (or images or voice), it develops a model of language that can create coherent sequences of text when given a prompt.

This process relies on a software architecture called a deep neural network (DNN), specifically transformer networks. Transformer networks are adept at breaking down text into "tokens," which are basically parts of words ("words" is one token, "basically" is two tokens). Then, it predicts the sequence that's most likely to make sense with the user based on their interactions. The calculation is different for every person and conversation, requiring a huge amount of electricity and energy.

ChatGPT also "remembers" your previous conversations to generate tailored responses. The more you talk to it, the more it refines its interactions with you. If you say phrases like "that's not right," the model will take note and try a different approach next time. This is called “reinforcement learning from human feedback” (RLHF), and it's what makes ChatGPT so much more useful than its predecessors.

How Can I Try ChatGPT?

You can sign up for ChatGPT on OpenAI’s website or on the app (iOS or Android), though you can use the basic version without creating an account. The free version will suffice for occasional conversations, but it limits the number of exchanges you can have with the flagship GPT-4o model in one day and the number of photos you can upload.

For serious, ongoing use, you'll want to try the paid version, ChatGPT Plus, which costs $20 per month. It has fewer limits and extra capabilities like the Sora video creation model and custom GPTs. The latter are miniature models you can use for specific tasks, like language translation, whereas the main ChatGPT model is more of an all-purpose athlete.

OpenAI also offers other subscription tiers, like a $200-per-month Pro model, which has no limits and can do things like compile advanced research reports. There are also Team and Enterprise accounts for large organizations. Finally, developers can also access ChatGPT through OpenAI’s API, where you pay for it based on the number of tokens you use.

What Can I Do With ChatGPT?

• Writing: ChatGPT can be a helpful writing assistant. If you prompt it to write a full essay or article in one pass, it will give mixed results. But if you work with it step by step, ChatGPT can do impressive things. For example, you can start with an outline and flesh out each part with the help of OpenAI's chatbot.
• Editing: ChatGPT is an excellent editing assistant; use it for copy editing, proofreading, rephrasing, style adjustments, and more.
• Translation: ChatGPT can translate very well into dozens of languages. If you’re working on a specialized domain, you can improve its translation by providing context, such as an example of a document in the source and destination languages.
• Summarizing: ChatGPT can summarize articles, speeches, and papers. It becomes more accurate when you provide guidelines, such as which topics to highlight.
• Brainstorming ideas: ChatGPT can provide all kinds of assistance here, from suggesting discussion points for a presentation to planning a trip.
• Writing code: ChatGPT is a good coding assistant and turns functionality descriptions into working code in dozens of programming and scripting languages.
• Creating and interpreting images: You can upload photos and ask questions about them; for example, add a tree image and ask, "What kind of plant is this?" ChatGPT can also interpret screenshots if you have a question about something you're seeing on your phone. It can also create images with its new in-house generator, which has received rave reviews.
• Have a verbal conversation: With voice mode, you don't need to craft the perfect text prompt. Just speak into the microphone and start chatting.
• Create movie snippets: The Sora video generator cranks out custom clips, just a few seconds in length with no sound, which you can use to augment larger video projects or use as standalone snippets. It's available to ChatGPT Plus and Pro subscribers and in a limited capacity for free via the Microsoft Bing app.

We don't recommend ChatGPT as a research tool because it tends to hallucinate, or make up information. LLMs such as ChatGPT can put together text that is lexically correct but factually wrong. This also applies to using ChatGPT for coding: It might generate code that is non-functional or insecure. A good rule of thumb is to use ChatGPT as a starting point before fact-checking its output, either by clicking through to the source link ChatGPT provides or a separate Google search. (Perplexity is another citation-focused chatbot.)

What Are Some Alternatives to ChatGPT?

• Google Gemini: Google's AI model can do almost everything you can do with ChatGPT, with the added benefit of hooking into Google's ecosystem, so you can export its output to Gmail, Google Sheets, Docs, etc.
• Bing: Microsoft, which has invested billions in OpenAI, integrated ChatGPT into its Bing search engine. It's a conversational interface to search for knowledge and perform other tasks you would do with ChatGPT. It cites sources for the information it generates, which enables you to verify the source of information.
• Claude: Anthropic, a San Francisco–based AI lab, makes Claude, a ChatGPT rival with a strong reputation for writing and coding.
• Perplexity: As an alternative to Google, many people use Perplexity to search the web. It's gaining prominence rapidly and will be the go-to search platform on Motorola's new Razr phones. Samsung is reportedly in talks to do the same.
• Open-source models: The open-source community released LLMs that you can run on your own servers. These LLMs can help you control your data and avoid vendor lock-in. Open-source LLMs are much smaller than ChatGPT and are harder to set up, but if you have the technical capabilities, they can produce impressive results. Meta's Llama models are some of the most well-known. Other options include Open Assistant, Alpaca, Vicuna, and Dolly 2.

For more examples of how PCMag writers are exploring AI, check out our Try AI series.

Disclosure: Ziff Davis, PCMag's parent company, filed a lawsuit against OpenAI in April 2025, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.

Source: PCMag