Is AI Safe?
By Alexandra Wade - November 3, 2022

The emergence of complete artificial intelligence, according to Stephen Hawking, “may mean the end of the human race.” AI, according to Elon Musk, is the “greatest existential danger” to humans. People could be questioning, “Wait, what?” as a result. But these significant concerns have a scientific basis. Leading individuals at Oxford and UC Berkeley, as well as many AI researchers, agree with Hawking and Musk that if sophisticated AI systems are used recklessly, they may irreversibly shut off human society from a bright future.

Getty Images / Westend61 / Westend61

Some of them feel that because it is so far off, there is no use in thinking about sophisticated AI right now. Others are concerned that overblown claims about the strength of their industry might lead to an early demise. Even among those who mostly concur that AI offers distinct risks, there are divergent opinions on the current course of action.

The argument that AI might endanger humanity’s existence is more nuanced and difficult to understand. So many of those who are striving to create safe AI systems must first explain why these systems are unsafe by design. AI systems pursue their goals whether or not those goals are what we actually meant — and whether or not humans are in the way. This reality is at the heart of the notion that AI might pose a threat.

It is simple to create an AI that avoids that particular hazard. However, there are other ways that releasing powerful computer systems might have unanticipated and perhaps disastrous repercussions, and preventing them altogether is a considerably more difficult challenge than avoiding just one in particular. A sufficiently strong AI system may behave unexpectedly, following its objectives down a path we hadn’t anticipated.

AI appears to be a technology that, when it is developed, will fundamentally alter society. It will be like launching a rocket, according to researchers from several significant AI groups; we must get it perfect before pressing the “go” button. So it seems imperative to start studying rocketry. Regardless of whether mankind should be scared, we should unquestionably be doing our research.