How to Talk to AI using ChatGPT
ChatGPT’s ability to generate human-like responses has generated much buzz. Here’s how to talk to AI with ChatGPT?
ChatGPT: Optimizing Language Models for Dialogue
Artificial Intelligence (AI) Research Company OpenAI on Wednesday announced ChatGPT, a trained model which interacts in a casual way.
This is a prototype model dialogue-based AI Chabot capable of understanding natural language and responding in natural language. The dialogue format makes it possible for ChatGPT to answer follow-up questions, admit its mistakes, challenge incorrect premises, and reject inappropriate requests.
It has taken the internet by storm and since then millions are following this app. As per study, more than millions has started following it in just a week.
While some are applauding its intelligence, some of them are calling it a replacement of Google, since it is trained to follow instructions in a prompt and response in detail.
What is ChatGPT?
ChatGPT is based on GPT-3.5, a language model that uses deep learning to produce human-like text.
It is a better version in itself as it generates detailed text and can even come up with poems.
Another outstanding feature of this model is its memory. It has a unique memory power to remember earlier comments in a conversation and recount them to the user.
As per access, the OpenAI has only opened up the bot for evaluation and beta testing but API access is expected to follow next year. However, under its beta testing phase, ChatGPT’s abilities are already quite remarkable.
Sample Testing and responding
Software start-up founder Amjad Masad got ChatGPT to spot errors in his code and found a detailed output on what’s wrong with it and how it can be fixed.
Canadian Musician Grimes was all about the sentimental side of things. When she asked ChatGPT if it felt “trapped,” ChatGPT responded by saying it lacks the ability to feel so.
YouTuber Liv Boeree feels that kids spending hours on homework will be a thing of the past – ChatGPT will do the job for them. After asking ChatGPT to do the task, the bot wrote a full 4-paragraph essay and also solved a complex math equation.
Limitations
While many of them are in gape with the model, there are many who were quick enough to catch the cons of the bot.
The bot is incapable of handling algebraic problems. And people often get mislead as it is confident enough for its solutions.
OpenAI understands these flaws and has noted them down on its announcement blog: “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.”
To summarize, keeping limitations aside, ChatGPT still makes for a fun little bot to interact with. You can try it out from its official website after signing up for the same.
During the research preview, usage of ChatGPT is free. Try it now at chat.openai
