Nearly 50 years ago, when computers were in their infancy, the late Arthur C. Clarke, the world’s greatest science fiction writer, created a fictional computer called HAL (Heuristically Programmed Algorithms). HAL was sentient and could act almost like a human. After that, no one could control it. The title of the book is 2001: A Space Odyssey. It popularized the idea of artificial intelligence. Clark foresaw a self-replicating computer that would destroy Jupiter.
Half a century later, we are at a point where what Stephen Hawking feared could come true. This is what makes AI different from all new technologies of the past. But being sentient or self-replicating is not enough. Both conditions must be met before the AI becomes dangerous. The fear is that if it does, it will bring about the doom of mankind. So how should humans deal with the problem? We know what AI is, but we don’t know its future trajectory. We also know that anything beneficial that humans have invented can be harmful to humans. And, as always happens when a lack of knowledge breeds anxiety, science is mixed with semi-superstition. There are a few things you simply have to do.
First, world-leading institutions like the United Nations must step forward to initiate the next steps. This is to convene a committee of the world’s best (and only) scientists, and no one else. The committee should develop its own terms of reference for assessing risk, namely sensory and self-replication. Second, there is a great deal of private intellectual property involved, and like the US FDA, we must find a way to verify the individual results of AI research. Institutions must have global authority. That is, it must be produced by a treaty, like other bodies such as the WTO, WHO, etc. Third, in parallel, individual governments must do the same within their jurisdictions. But their main focus should be regulation. But this requires understanding what you are regulating. As such, coaching and training to become a regulator is a necessary prerequisite. Under no circumstances should career bureaucrats be involved in this exercise. Again, this should be left to the scientists and engineers. Fourth, as the United States and China seek to control the outcome, we must recognize that while the United States is attacking for personal gain, China is attacking for the benefit of the CCP. must be These are opposites. Controlling their impulses becomes a major challenge.
Last but not least, there is India that does not apply when needed. There are good reasons to believe that governments have not had the luxury of giving AI much thought. This perception must be corrected without delay, and not after someone has received an order from the Supreme Court on Mandams’ warrant, as is often the case today.
This is the last free article.
