ChatGPT now pretends to be a machine

ChatGPT, the AI-powered chatbot currently taking the world by storm, is even capable of simulating a Linux environment.

Powered by OpenAI, the research company launched by Elon Musk and backed by Microsoft, the discovery of its performance credentials was recently put into an open testing phase, allowing anyone to play with it.

To prepare for the position, ChatGPT used the information provided to them as part of their initial training data. These probably included excerpts from Linux manuals and actual shell session logs.

act of

Jonas Degrave, a DeepMind researcher who made the discovery, made it very clear to ChatGPT that he wanted it to pretend to be a Linux terminal and respond to his typed commands just like a Linux terminal would. The chatbot duly complied, responding to the code block format just like a terminal would.

ChatGPT works by determining which words are most likely to follow previous words in the conversation, optionally relying on the full history of the conversation when further prompting is done.

ChatGPT was fully immersed in the part of a Linux machine, even going so far as to run Python code. Degrave gave him a simple calculation written in the Python language, and he responded with the correct answer.

Another ChatGPT user got it to simulate an old bulletin board system (BBS), going from simulating a modem to creating an imaginary chat room with an imaginary chat named Lisa.

OpenAI and ChatGPT have recently lit up the world of AI with all kinds of surprising and original results, from simulating an ATM to explaining Einstein's theory of general relativity in the form of a rhyming poem.

However, when it comes to more vital tasks, like gathering factually accurate information, ChatGPT is less effective, with commenters pointing out that its new releases can be just as absurd as they are accurate, depending on user input.

Share This