23 Jan 2023 ChatGTP
ChatGTP
This article covers “Daily current events “and the topic is ‘ChatGTP’ which is in news, it covers “Science and Technology” In GS-3, the following content has relevance for UPSC.
For Prelims: About ChatGTP
For Mains:GS-3 , Concerns associated with ChatGTP
Why in news
OpenAI, a business that does AI research and implementation, recently allowed customers to test the capabilities of ChatGPT, its most recent and potent AI chatbot.
About ChatGPT
ChatGTP is a subset of OpenAI’s GPT (Generative Pre-trained Transformer), a sophisticated neural network-based language model.
To produce writing that is human-like, GPT models are trained on enormous volumes of text data.
It may produce responses to a variety of topics, including asking and answering questions, explaining things, and having dialogues.
The ChatGTP can also “address follow-up inquiries” in addition to being able to “admit its mistakes, contest misleading premises, and decline unsuitable requests.”
Additionally, reinforcement learning from human feedback was used to teach the chatbot (RLHF).
Limitations of ChatGTP
The chatbot was clearly biased against people of color and women, which is still a concern with almost all AI models.
Although some have noted that the chatbot’s responses lack context and depth, this is mostly true. The chatbot nonetheless provides well-written, grammatically correct responses.
The information that ChatGPT occasionally provides is incorrect, and the knowledge it has is limited to world events that happened before 2021.
The ethical issues surrounding it
Coding that is harmful and dangerous: Users have been experimenting with the chatbot’s capacity to commit sinister deeds. Despite their claims to be amateurs, a number of users have asserted that Chatbot produces malicious and dangerous programs.
Phishing Email Generation: Although ChatGPT is designed to refuse requests to write dangerous code or phishing emails, it really produces excellent phishing emails.
Biased Data/Information Creation: One issue to be concerned about is the possibility of bias in the created code, as the training data used to develop the code generator may include biases that are mirrored in the output code.
Limiting the Job Opportunity: It is feared that the use of code generators may result in the loss of programming jobs for people.
Plagiarism Chokepoint: Academic plagiarism is not a recent problem, but ChatGPT has changed how AI is used to create unique work. As a result, it is challenging to spot copied content. Concerns about ChatGPT’s effect on written tasks have also been raised by academics and teachers.
Way ahead
Increasing its capacity to recognize and react to context, such as the ability to carry on a discussion about a certain subject or discern the tone and intent of a communication.
Expanding its knowledge base and capacity to produce more precise and insightful answers.
Acquiring more sophisticated language-generating talents, such as the capacity to produce imaginative and compelling fiction or the ability to compose cohesive and persuasive essays.
Enhancing its capacity to comprehend and react to various languages and dialects.
Acquiring more sophisticated interactive skills, such as the ability to manage many tasks and participate in complex conversations.
Source:
Get Current Affairs from Yojna IAS
Yojna IAS daily current affairs eng med 23rd Jan
No Comments