Creating an artificial intelligence that can solve all the world’s problems simultaneously is considered “absurd” by a prominent AI researcher, Michael Jordan, an American, who made this statement before a summit on the topic.
Jordan, a computer science professor at the University of California, Berkeley, was addressing claims made by industry leaders like Sam Altman and Dario Amodei that general artificial intelligence (AGI) is imminent.
The AGI is the desired goal of AI research that is believed to surpass humans in various areas.
Jordan, a specialist in machine learning and statistics, expressed disbelief at the idea of a super AGI that possesses all knowledge during an AI conference at Polytechnic Engineering School in France.
Jordan expressed disapproval of companies investing “billions” in pursuit of what he deemed currently an unattainable objective.
AI models today can accurately predict a wide range of outcomes due to their access to vast amounts of human-generated data, such as internet data, according to Jordan.
The AI is limited in its knowledge as it cannot predict the user’s immediate future actions.
“If you are unaware, you cannot provide me with valuable guidance,” he stated.
Language models used in AI chatbots, like ChatGPT, can be likened to a collection of specialized experts able to respond to detailed queries, yet they are not considered intelligent beings.
Jordan described recent advancements in AI as a novel field of engineering comparable to civil engineering, focusing on human decisions, data, and network flow.
“I don’t believe there has ever been a time in human history when a new technological field emerged with such excitement and frenzy,” he noted.
Scientists and developers are in a heated debate over the feasibility of achieving AGI and the potential risks it may pose if its creators are unable to maintain control.