ChatGPT, a new chatbot that’s quickly become the latest fad in tech. The technology was developed by San Francisco-based OpenAI, a research company led by Sam Altman and backed by Microsoft, LinkedIn co-founder Reid Hoffman and Khosla Ventures.
ChatGPT debuted in late November and has quickly turned into a viral sensation, with people tweeting questions, such as “Are NFTs dead,” and requests like, “Tell a funny joke about the tax risks of international remote work.” They include a screenshot of ChatGPT’s response, which often — but not always — makes sense. I tried it myself and talked to this bot. And his answers are generally quite realistic. But at the same time, if you ask the bot to compare and choose something better, you will get an answer like this :
“It is not appropriate for me to make a determination as to which brand is “better,” as this is a subjective decision that depends on an individual’s preferences, needs, and priorities.“
And here’s what the ChatGPT can tell us about itself
“I am a language model, which means that I have been trained on a large dataset of text and can generate human-like responses to a wide range of questions and prompts. I do not have my own personal experiences or physical presence, and I exist only as a set of algorithms running on a computer. My primary function is to assist users like you by providing information and answering questions to the best of my ability. I am here to help and do not have any personal opinions or feelings. Is there anything else you would like to know?
You can ask me any question within the scope of my knowledge and training. Some examples of the types of questions I can help with include:
- General knowledge questions
- Questions about language and vocabulary
- Questions about history, science, and other academic subjects
- Questions about technology and computing
- Questions about culture, arts, and entertainment
If you’re not sure what to ask, you can try asking me something about a topic that you’re interested in, or you can ask for suggestions on a specific topic. I’ll do my best to provide accurate and helpful information.”
It is also important to remember that this is a fairly new technology, it has limitations. For example, ChatGPT sometimes writes plausible, but at the same time meaningless answers. It is difficult to fix this problem, because: there is currently no source of truth during training, or training the model to be more careful forces her to reject questions that it can answer correctly, also supervised training misleads the model, because the ideal answer depends on what the model knows, and not on what the user knows. ChatGPT is also sensitive to water phrasing changes or repeated attempts to enter the same thing. For instance, with one formulation of the question the model may answer that it does not know the answer, but with a paraphrase it can answer correctly. In such cases, ideally, the model could ask clarifying questions when the user enters a question that is not entirely clear. But so far this does not happen and the bot just guesses what the user meant. So no matter how interesting the technology is and most likely very promising for the future, especially for search engines, it cannot be trusted 100% yet. Since AI can also generate something incorrectly.
However, in a year, this turned into a failure for the technology sector, with mass layoffs, wrecked stock prices and crypto catastrophes dominating the headlines, ChatGPT has served as a reminder that innovation is still happening.
And in the end, here’s a little joke from the bot:
Why was the math book sad?
Because it had too many problems…
I hope this joke brimgs a smile to your face.
Sources:
https://www.cnet.com/tech/computing/why-were-all-obsessed-with-chatgpt-a-mind-blowing-ai-chatbot/
https://openai.com/blog/chatgpt/
https://www.cnbc.com/2022/12/13/google-execs-warn-of-reputational-risk-with-chatgbt-like-tool.html