ChatGPT is an artificial intelligence chatbot developed by OpenAI and launched in November 2022 and known for its detailed responses and articulate answers across many domains of knowledge. Its uneven factual accuracy, however, has been identified as a significant drawback
It is built on top of OpenAI’s GPT-3.5 and GPT-4 families of large language models (LLMs) and has been fine-tuned (an approach to transfer learning) using both supervised and reinforcement learning techniques.
What ChatGPT Can do :-
- Can write and debug computer programs
- Compose music, Teleplays,Ffairy tales and Student essays
- Answer test questions (sometimes, depending on the test, at a level above the average human test-taker)
- Write poetry and song lyrics
- Emulate a Linux system
- Simulate an entire chat room
- Play games like tic-tac-toe
- Simulate an ATM.
ChatGPT’s training data includes man pages and information about internet phenomena and programming languages, such as bulletin board systems and the Python programming language.
it also attempts to reduce harmful and deceitful responses as well as remember previous prompts given to it in the same conversation. To prevent offensive outputs from being presented to and produced from ChatGPT, queries are filtered through OpenAI’s company-wide moderation API, and potentially racist or sexist prompts are dismissed.
What are ChatGPT Limits :-
- OpenAI acknowledged that ChatGPT “sometimes writes plausible-sounding but incorrect or nonsensical answers”. This behavior is common to large language models and is called artificial intelligence hallucination. The reward model of ChatGPT, designed around human oversight, can be over-optimized and thus hinder performance, otherwise known as Goodhart’s law.
- In training ChatGPT, human reviewers preferred longer answers, irrespective of actual comprehension or factual content. Training data also suffers from algorithmic bias
- Limited knowledge of world and events after 2021
