The head of Meta AI’s Artificial Intelligence division does not anticipate the imminent arrival of superintelligent AI, suggesting that attaining intelligence comparable to “cats and dogs” might be achieved sooner than human-level intelligence. Jan LeCun, Chief Scientist at Meta and a deep learning pioneer, asserts that current AI systems are still far from achieving certain facets of consciousness that would make them genuinely intelligent, surpassing the creative text summarization capabilities.
This stance diverges from Jensen Huang, the CEO of Nvidia, who recently claimed that artificial intelligence would be “somewhat competitive” with humans in less than five years, outperforming people in various mentally demanding tasks. LeCun acknowledged Huang’s viewpoint, emphasizing the ongoing competition in the field of artificial intelligence and stating, “There is a war going on between companies in the field of artificial intelligence, and he is supplying us with weapons.”
LeCun made these statements at a recent event commemorating the tenth anniversary of the establishment of Meta’s parent company, Facebook, for the essential artificial intelligence research team.
Expanding on his thoughts regarding companies dedicated to developing general artificial intelligence, LeCun explained, “General artificial intelligence exists because of the number of processors you have to buy from providers like Nvidia.” He believes that as researchers at entities like OpenAI persist in their quest for advancing general AI, the demand for computer processors from Nvidia will increase.
LeCun envisions that humanity is likely to achieve artificial intelligence at the level of “cats” or “dogs” before reaching human-level AI in the years to come.
Leave a Reply