Artificial Intelligence has been a cause for several tech-outcomes that proffer different fields including education, medical, art and life. At the same time, AI has become a chatting person too. This ChatGPT by OpenAI has been trained to provide conversational answers to users’ queries on any topics in a wide range from coding, thermodynamics to cinematic scripts.
OpenAI, the company co-founded by Richest of the world Elon Musk, prospects AI’s capabilities in number of ways, – conversational chatbots like this and chatting with dead-persons, converting text to video via its Dall-E algorithm and more.
The firm’s latest released prototype ‘ChatGPT’ is nothing but an AI bot that can converse on any topic with users, showing off new capabilities. However, the AI bot raises concern of misinformation or cogent waffle that it is prone to deliver.
ChatGPT is adapted from OpenAI’s GPT-3.5 model and is strained to provide more conversational answers as humans chat with one another. GPT-3.5 is an update of GPT-3, whose job is to simply predict what text follows any given string of words. The bot is designed to be fluid enough of being able to converse as humans to the users. ChatGPT is made available to everyone and you can test out the model by heading here.
A number of users have already conversed with the ChatGPT on topics like coding, thermodynamics, romcom scripts based on existing shows, various scientific and school topics, essays and more.
Pros aside, the chat-bot is vulnerable and teased for its susceptibility of confidently delivering false or invented information as a matter of fact. The simple logic behind this criticism is the reason of how these bots are programmed – reinforcement learning.
ChatGPT becomes the biggest talk of the decade, as the AI spreads widely gaining 1 Million users in just 5 days and 10 Million daily users in just 40 days. The AI chatbot is great in providing answers that students and professionals had started using it for their academic and professional reasons. Almost 30% of the workforce had used the chatbot for their own benefits or tasks, a report said.
Recently, a new research conducted by a professor at University of Pennsylvania’s Wharton School found that the artificial intelligence-driven chatbot GPT-3 was able to pass the final exam for the school’s Master of Business Administration (MBA) program, scoring B on the exam. It had passed in a law exam too, though securing an average C+ performance.
I had experimented the trending ChatGPT to ask controversial questions and see how it was enacting for answers. It was factual on basic information and for few questions, it was so diplomatic to the surprise. See how it answered for range of questions.
ChatGPT’s maker OpenAI now gets a huge $10 billion investment from Microsoft in plans to incorporate the AI chatbot into its Bing Search, seeking to enter the race against Google again. Google, however, urges the inmates as the firm sees “Code Red” situation due to the launch of this revolutionary AI Chatbot. Google wasn’t expecting that. In action to that, the search-giant plans to launch over 20 AI products or services on May of 2023. It’s Google vs ChatGPT now.
Cons of AI Chatbots
People misconceive that AI bots are intelligent information organizers using potentials of internet and web, rather they are trained user-friendly answer-givers, which might not know what they are talking about.
In this instance, ChatGPT’s AI has been fed with answers or queries of human’s conversations to get trained enough to converse like them. OpenAI had explained in a blog post that the bot itself was created with the help of human trainers who ranked and rated the way early versions of the chatbot responded to queries. This information was then fed back into the system (reinforcement learning), which tuned its answers to match trainer’s preferences.
As these input data may mislead, the AI bot too could fall into falsity. Likely as some AI researchers call these chatbots “stochastic parrots”. Their knowledge is derived only from statistical regularities in their training data, rather than human-like understanding of the world as a complex and abstract system.
The company also says that while ChatGPT has certain guardrails in place, “the system may occasionally generate incorrect or misleading information and produce offensive or biased content.”
The bot did confidently assert obvious false information, a number of examples show. For instance, when a computational biology professor Carl Bergstrom asked the bot to write a Wikipedia entry about his own life, ChatGPT does with aplomb — while including several entirely false biographical details.
Anyhow, People liked it, it seems!
Many have tested out the chatbot and excitingly took the experience to Twitter reviewing on the platform.
Questioning on Thermodynamics to the ChatGPT:
Apart from the curses, the bot might be very useful for the school & college assignments.
And the bot can combine its fields of knowledge in all sorts of interesting ways. So, for example, you can ask it to debug a string of code … like a pirate, for which its response starts: “Arr, ye scurvy landlubber! Ye be makin’ a grave mistake with that loop condition ye be usin’!”
Would you try this AI Chatbot?