Microsoft Researchers claim GPT-4 is showing \”sparks\” or AGI

A team of Microsoft AI researchers published a paper shortly after GPT-4 was released. They claimed that the OpenAI language models — which power Microsoft’s Bing AI, now somewhat lobotomized — show \”sparks\” or artificial general intelligence.

In the paper, the researchers place a lot of emphasis on \”sparks\” and describe GPT-4 as a \”first step towards a number of increasingly general intelligent systems\”, rather than a fully-hatched human-level AI. The researchers also stressed that the paper was based on a \”very early version\” of GPT-4 that they studied when it was still \”in active development by OpenAI\” and not necessarily a version that has been wrangled to product-applicable form.

These are serious claims. Disclaimers aside. A lot of people, including some in the AI industry, consider AGI to be a pipedream. Others, however, believe that AGI is the key to humanity’s future. The next-gen GPT-4, for example, is the most advanced version of the OpenAI Large Language Model (LLM), and is ranked near the top, if it is not the number one contender, on the list of theoretical AGI candidates.

Source:
https://futurism.com/gpt-4-sparks-of-agi