ChatGPT: Does the Software Becomes Aware of Itself?

By , in Sci/Tech on . Tagged width:

Late in 2022, ChatGPT started making waves in the internet community, and it doesn’t look like it’s going away anytime soon! The generative artificial intelligence system that was built by the technology company OpenAI has been proven a real success, and experts have increased their caution about the hazards associated with the system. But here’s the thing. Certain chatbots started acting weirdly, going off-script, fooling other bots, and chatting back. This unquestionably sparked additional worries about how near some AI tools are getting to human-like intelligence. The debate of whether or not we should be afraid of AI persists. It’s more like an SF movie’s plot, isn’t it?!

A group of computer scientists from all around the world have been working together to investigate the level at which large language models (LLMs) like ChatGPT can gain abilities that hint they might be able to become aware of themselves and the world around them. It’s more like we turn to science once again to alleviate our fears, right?!

In answer to a series of question prompts, generative artificial intelligence (AI) and the LLMs they are based on are given their name because of the method in which they examine the relationships between billions of words, phrases, and paragraphs to produce natural-sounding streams of text. They actually take in massive volumes of material in order to figure out which term is probably going to follow after it. The researchers concentrated their investigations on one aspect of situation awareness, which they refer to as “out-of-context” thinking. This aspect may have been a possible forerunner to situation awareness.

The researchers conducted a number of tests on LLMs of varying sizes and discovered that bigger models performed better in tasks evaluating out-of-context reasoning for both GPT-3 and LLaMA-1. This was the finding of the recent study.

These findings offer a foundation for further empirical study toward predicting and potentially controlling the emergence of situational awareness in LLMs, reads the team’s statement.

Tiesha loves to share her passion for everything that’s beautiful in this world. Apart from writing on her beauty blog and running her own beauty channel on Youtube, she also enjoys traveling and photography. Tiesha covers various stories on the website.