Pi the chatbot that aims to focus more on "human" conversations
A few days ago the news broke that Inflection AI, a California-based company founded by LinkedIn co-founder Reid Hoffman and DeepMind co-founder Mustafa Suleyman, introduced Pi, a chatbot that aims to prioritize human conversations with a high level of emotional intelligence, the company said. Pi has a more limited use case than ChatGPT, Google's Bard, or Microsoft's Bing chatbot, but it's better than those apps for conducting conversations in a human way.
With Pi, Inflection AI joins a race with big and small tech companies to develop chatbot products based on generative artificial intelligence.
About Pi
Pi (personal intelligence) is a new chatbot that made its debut in the highly competitive AI chatbot market. Pi is meant to act as an active listener, helping users ask questions or solve problems in a back-and-forth dialogue that they later remember. The goal is to get to know the users.
Inflection AI has placed a high priority on developing the emotional intelligence of the chatbot and to ensure that you behave appropriately in all situations. The company has trained Pi to be aware of its behavior and avoid any violation of their policies, including not engaging in romantic conversations. The goal is to ensure that the chatbot never tries to imitate a human and frequently remind users that it is an AI.
“Pi is eager to help you in any way he can. Talk to Pi about a topic that interests you, from quantum physics to movie trivia. Ask for advice on big and small decisions in his life. Or get helpful tips to help you learn new hobbies and skills," the company said in a tweet.
Mustafa Suleyman, CEO of Inflection AI, described the chatbot as a sympathetic sounding board, instead of a source of information. He said the company's chatbot has much more limited use cases, making it safer and easier to control.
Pi was designed to prioritize human-like conversations with a high level of emotional intelligence, including kindness and support.
“Pi is very balanced and impartial on political issues or sensitive topics, but he can also be funny, silly and creative at times. It is suitable for personal daily tasks,” said Suleyman. He said the company has limits in place to ensure the AI doesn't violate its behavior policies, including engaging in romantic conversations.
“The goal is to make sure that the AI always knows that it is an AI and never tries to impersonate a human. So it frequently reminds the human user that they are an AI,” he said. According to him, users can interact with the Pi on different platforms, including its website, app, and social media platforms like Instagram. The service is free, and the startup may launch premium subscriptions in the future. The company says it uses user data, including conversational content, to train its AI systems. Pi is currently not connected to the Internet.
During a live demonstration of Pi, distinguished itself from other chatbots such as ChatGPT or Bard due to the fact that he often ended his answers with a question for the user, thus encouraging dialogue.
"That's what Pi does really well, it helps facilitate their own line of research," Suleyman said. However, he does not provide citations or references, though Suleyman says that will change. Furthermore, Pi also sometimes makes up facts. Simply put, he wows like he does with all the major language models (LLMs), the technology that underpins the next generation of AI-powered chatbots.
Chatbots using generative AI technology have become a hot topic since OpenAI's ChatGPT was released last November. By using large language models, leveraging large amounts of text to summarize information and generate content, chatbots allow people to have deep conversations for business and personal needs.
Finally If you are interested in knowing more about it, you can check the details In the following link.