AI language tools like ChatGTP and others made a big entrance in 2023.
So what does it actually mean?
Here is some basic explanations…
GPT (Generative Pre-trained Transformer)
GPT-3 produced by OpenAI is a language model has more than 175 billion parameters and can produce 2048-token long-form content.
GPT-4 has more than 100 trillion billion parameters.
Tokens
Tokens are the basic unit that OpenAI GPT models (including ChatGPT) use to compute the length of a text.
Tokens can be thought of as pieces of words. Before the API processes the prompts, the input is broken down into tokens. These tokens are not cut up exactly where the words start or end – tokens can include trailing spaces and even sub-words.
The token limit includes BOTH the prompt and reply text count.
Here are some helpful rules of thumb for understanding tokens in terms of lengths:
1 token ~= 4 chars in English
1 token ~= ¾ words
100 tokens ~= 75 words
2048 tokens ~= 1,536 words
There are two versions of GPT-4, with context windows of 8,192 and 32,768 tokens, a significant improvement over GPT-3.5 and GPT-3, which were limited to 4,096 and 2,048 tokens respectively.
ChatGTP
ChatGPT is a specific application that runs on GPT.
(Image source: Zapier)
ChatGTP Plus
ChatGPT Plus is a GPT-4 backed version of ChatGPT available for a 20 USD per month subscription fee (the original version is backed by GPT-3.5).
OpenAI also makes GPT-4 available to a select group of applicants through their GPT-4 API waitlist; after being accepted, an additional fee of 0.03 USD per 1000 tokens in the initial text provided to the model (“prompt”), and 0.06 USD per 1000 tokens that the model generates (“completion”), is required to use the version of the model with a 8192-token context window; for the 32768-token version, those prices are doubled.
Auto-GPT
Auto-GPT is an autonomous “AI agent” that given a goal in natural language, can perform web-based actions unattended, assign subtasks to itself, search the web, and improve its own code.
Instead of asking GTP to do ONE specific task, with Auto-GTP you can give it a GOAL instead. It will then figure out the tasks to complete that goal. And then create multiple workers (agents) who would then complete their tasks and help each other to do so. e.g. Auto(self) prompting.
Auto-GTP can:
– Use the power of GTP-4
– Access to the internet
– Abilility to search
– A short-term and long-term memory (long term using PINECONE API)
– Able to save the output to your computer
– A voice so it could speak back to you (ElevenLabs API)
Auto-GPT is an experimental, open-source Python application that uses GPT-4 to act autonomously. This means that Auto-GPT can perform a task with little human intervention, and can self-prompt.
Auto-GPT was posted on GitHub on March 30, 2023 by developer Significant Gravitas. However, the application is driven by GPT-4 which is OpenAI’s latest and most advanced AI model.
Auto-GPT has internet access, long-term and short-term memory management, GPT-4 for text generation and file storage and summarization with GPT-3.5, according to the Github post.
The program plans step-by-step, justifies decisions and develops plans, which it documents. In addition to integrating GPT-4 for text generation, the system has Internet access for data retrieval, can store data, and can generate speech via the Elevenlabs API. Supposedly, the system is even capable of self-improvement and bug-fixing by generating Python scripts via GPT-4.
Unlike ChatGPT which lives on a browser that you can easily visit, for Auto-GPT you need to download specific software and be familiar with Python to access it.
Other reading:
Chat gpt 4 parameters, How to use GPT 4, chat gpt 4 features, Login and more
https://sscnr.net.in/chat-gpt-4-parameters/