Over 10000 individuals gave an open call to pause the developments in AI. As a technology, AI is overhyped. Large language models (LLMs) that AI uses simply predict what a reasonable person would say next. The LLMs are trained on all online text, say the entire web, books and massive amounts of software.

LLMs use a technology invented in 2018 by Google called Transformers. Transformers let predictions get influenced by events in the past. The outputs of LLMs become very sophisticated because of this.

However, the word prediction is not something novel. There were statistical predictors 40 years ago in speech recognition, natural-language modelling and stock-market predictions. However, data used to condition those models then was sparse. These days machines cope with billions of documents, and mimic people with surprising accuracy.

It should be noted that apart from predictions ,LLMs do not plan, do not decide. They do not have intention or malice.

Movies paint AI as a monster out to destroy the world. How just predictions could enslave the mankind is beyond comprehension.

Of course, AI will disrupt the way we work. It will affect employment. It is to be used responsibly. Cybercriminals and bad actors can misuse it. We should be alert. We should regulate it when such misuses happen, What is paramount is safety and ethics. AI will make us more effective. Mankind would have been poorer if it had stopped electricity, printing press and steam-engine. All these technologies too frightened people when they arrived.