Renowned AI Pioneer Thinks Humanity is at Risk Because of AI
Yoshua Bengio is a prominent figure in artificial intelligence (AI). He is the founder and scientific director of the Montreal Institute for Learning Algorithms. He has expressed his concerns about AI’s potential risks. In an interview with CTV’s Question Period host Vassy Kapelos, Bengio stated that some AI technologies are rapidly approaching human intelligence levels, occurring much sooner than anticipated.
Dangers of Matching Human Intelligence
Bengio was asked whether he shared the concerns of others who believe that AI matching human intelligence puts humanity at risk. His answer was an unequivocal “yes.” According to Bengio, shifting views about something you have been working on for decades can be challenging, mainly when your career and motivations revolve around bringing good to the world. However, he believes that scientists are responsible for remaining neutral and considering the potential risks associated with AI.
Bengio emphasized the need to engage with citizens, other experts outside of AI, and governments to understand and prevent potential bad scenarios. He noted that as some systems like ChatGPT evolve, they could pass as humans in online conversations, which poses certain risks. These systems could quickly propagate disinformation in more powerful ways than we already have with social media.
Faster Advancements, Slower Policies
One of the challenges is that AI technology is advancing faster than the speed of government and officials’ ability to develop policies to protect people and society from these risks. Bengio said that he did not expect the level of AI competence we see now to arrive so quickly. He believed it would take 20 or 50 years, but the acceleration has been significant in the last few years.
Counterfeiting Humans Should be Illegal
Bengio believes that in the short term, one of the important ways to prevent AI from posing a risk is to make it illegal and punish those who impersonate humans. He believes that when a user interacts with an AI, it should be clear that it is not human. This transparency should also extend to identifying which company made the AI. Counterfeiting humans should be as bad as counterfeiting money.
While AI has brought many benefits and improvements to our lives, it is crucial to identify potential risks associated with this technology. Yoshua Bengio’s concerns and suggestions about AI matching human intelligence are critical in light of the rapid advancements within the field. It is essential to remain neutral, consider the potential risks, and engage with various experts to create policies protecting people and society from these dangers.