Exploring ChatGPT: Understanding its Capabilities and Limitations

Exploring ChatGPT: Understanding its Capabilities and Limitations

Despite the hype surrounding ChatGPT, it's important to understand its capabilities and limitations.So lets understand ChatGPT and limitation

ChatGPT is a type of language model that uses deep learning techniques to generate human-like text. Here are a few key points about how it works:

  • The model is trained on a large dataset of text, such as books, articles, and websites. This allows it to learn patterns and structures of language, as well as how words are used in context.
  • Once trained, the model can generate new text by predicting the next word in a sequence, based on the words that came before it. This is done using a technique called "autoregression," where the model uses information from previous words to predict the next word.
  • The model can also be fine-tuned on specific tasks or domains, such as conversation or technical writing. This allows it to generate text that is more relevant and appropriate for a specific purpose.
  • ChatGPT is based on transformer architecture which allows it to handle context with much larger context window than previous architectures, this improve the quality of generated text.
  • Finally, the model can be used in a variety of applications, such as chatbots, automated writing, and language translation.


ChatGPT, like all language models based on deep learning, has several limitations and potential drawbacks, including:

  • Lack of common sense: ChatGPT can generate text that is grammatically correct and coherent, but it may not always make sense in the real world or have a complete understanding of the context.
  • Bias: ChatGPT is trained on a large dataset of text, which means it may pick up biases present in the data. This can lead to biased or discriminatory language being generated.
  • Limited understanding of the world: ChatGPT can generate text based on what it has learned from the training data, but it doesn't have any understanding of the real world. It can not make inferences or answer questions that require knowledge that is not in its training data.
  • Computational resources: Training and running a language model like ChatGPT requires significant computational resources and can be expensive.
  • Privacy concerns: Using large language model like ChatGPT for some applications can raise privacy concerns as it may require access to sensitive user data.
  • Safety issues: ChatGPT is capable of generating highly convincing text that can be used to impersonate or deceive people, this could lead to safety issues if not used properly.

Nitesh has also shared his quite useful thought on ChatGPT in his linkedin post.