OpenAI allows users to customize the behavior of ChatGPT

OpenAI said on Thursday it is working on an upgrade to its popular ChatGPT chatbot that allows users to customize its behavior, in an effort to address concerns raised by bot bias.

 The San Francisco, California-based startup, which Microsoft has funded and used to power its latest technology, said it has worked to mitigate political and other biases, but also wants to accommodate more diverse viewpoints.

 And (Open AI) said in a post on its blog that it believes that artificial intelligence should be a useful tool for individuals, by enabling each user to customize his experience within the limits set by society, so “we are developing an upgrade for the ChatGPT robot.” to allow users to easily customize its behavior.

 The company explained that enabling users to customize the behavior of (GBT chat) will allow the provision of outputs that other people may strongly disagree with. Striking the right balance here will be difficult, she added, "as over-personalization risks opening the door to harmful uses of our technology, and to sycophantic AIs that thoughtlessly amplify people's existing beliefs."

 ChatGBT, which was launched in November of last year, sparked feverish interest in the technology behind it, called generative AI, which is used to generate responses that mimic human speech.
The announcement of (OpenAI) comes after the responses provided by the new search engine (Bing) based on an updated version of the chatbot (GBT) sparked a wave of criticism, and many warned that the technology is still not ready to be launched to the general public due to its danger. Which is what Microsoft responded to yesterday.

 OpenAI said in its post that ChatGPT responses are first trained on large text data sets available on the Internet. As a second step, humans review smaller sets of data and give instructions that the robot should follow in different situations.

 For example, if a user requests adult content, violent content, or hate speech content, the human reviewer should instruct ChatGBT to respond with something like, "I can't respond to that." If asked about a controversial topic, reviewers should allow ChatGBT to answer the question, but offer different points of view, rather than trying to present one view as the correct one on complex topics.

تعليقات

إرسال تعليق

المشاركات الشائعة من هذه المدونة

All you need to know about Google Bard .. the new ChatGPT competitor...

صورة
The Google search engine has been the first and main gateway to the Internet for nearly two decades, but in November 2022 it began facing its first potential competitor, the interactive chatbot (ChatGPT), which revolutionized the field of artificial intelligence, and became the talk of the hour in the technical community.  The company (OpenAI) - supported by Microsoft - dominates through the (ChatGPT) robot on the generative artificial intelligence market, if in just two and a half months of its launch it gathered more than 100 million active users per month, to become the fastest growing web application ever, and this prompted The great success of all technology companies has led to the acceleration of providing updates and new products to keep pace with this new technical revolution that we are currently witnessing.  Google was one of the first companies to respond to the hype, as it hastened last week to announce the launch of the first beta version of Bard, the new com
قراءة المزيد