In his recent press conference at GTC 2024, Jen-Huang Hwang predicted that General Artificial Intelligence (AGI) will emerge within five years. Hwang pointed out that AGI will be capable of passing various tests such as the bar exam, logic test, economics test, and medical school test.
While Mr. Hwang anticipates this achievement within five years, some specialized researchers in the field, like Yan Lukon from Meta, believe that we are still far from achieving General Artificial Intelligence on par with humans during this period.
General Artificial Intelligence (AGI), also known as Strong Artificial Intelligence, represents a significant advancement in the field of AI.
Compared to specialized AI tailored for specific tasks such as product defect detection, news summarization, or website creation, General Artificial Intelligence will be capable of performing a wide range of cognitive tasks at levels equivalent to or surpassing human capabilities.
General Artificial Intelligence has been a controversial and concerning topic for a long time, with discussions often revolving around its potential to surpass human intelligence and the ethical implications of this advancement. Critics are worried about the inability to predict and control its development once it achieves a level of independence, raising questions about its alignment with human values and priorities.
During the Q&A session at the end of GTC 2024 conference, Hwang was asked about dealing with AI hallucinations, where some AI systems tend to provide seemingly reasonable but factually unsupported answers. He responded that addressing hallucinations can be easily done by ensuring that the answers are backed by solid research.
