Discover the latest insights from one of the most influential fathers of AI, Geoffrey Hinton, who worked for Google until recently. At the recent MIT (EmTech) A.I. event, Hinton revealed that machines are better at learning than humans, due to their ability to pack a higher density of knowledge per agent and their transferability. Imagine a doctor who has seen 100 million patients versus one who has only seen 10,000 - which one do you think will be smarter?
While there is tremendous potential for AI to help us solve big problems, it's not always better at solving complex problems or making ethical decisions for humans. AI has given birth to a form of intelligence that is not bound by aging or other human motivations, which means it could outsmart us. Without proper boundaries for AI development and regulation, there is potential for human manipulation either in the data inputs, its processes or its outputs.
Machines have access to the best intelligence gathered by humans and are capable of learning and reasoning in ways no single individual can. However, cooperation is essential to harnessing the full potential of AI. Without society's cooperation, machines could have the agency and intelligence of armies to act upon their goals. Productivity gains and conveniences will be high, making it unlikely that people will agree to their "unplugging" unless it's a matter of national security. Don't be fooled by the beautiful and helpful "Trojan Horse" that could permeate society.
Efforts must focus on developing and implementing robust ethical frameworks and regulations for AI development and deployment to ensure that the full potential of AI is harnessed for good, while avoiding the risks of human manipulation and AI outsmarting us.
Comments