The unintended consequences of progress, including the risks and possible harms caused, are always a factor that scientists and engineers consider in building any new technology. In today’s age, the spotlight is on AI – but the range of possibilities that progress might bring is near unfathomable.
“When I talk to the leading researchers in the field of AI right now, for example, they literally refer to this — right now — as their Oppenheimer moment. They’re looking to history to say, ‘What are the responsibilities for scientists developing new technologies that may have unintended consequences?'”
I am become death
Chris Nolan, director of a string of cinematic successes across the last three decades, was speaking on the subject matter of his 2023 blockbuster ‘Oppenheimer’ during a press tour.
Quote of the day
This article is part of TechRadar Pro’s QOTD project to provide an insight into the minds of the brightest and most recognized figures in the technology industry today and in years gone by. Read the full series here.
During a New York panel following a preview screening of the film, Nolan was speaking on the rise of AI, which has fallen into perspective after showing immense promise.
Latest Videos FromTechRadar
Although the mainstream image of AI has been coloured by a string of science-fiction movies throughout history, there are still valid questions to be asked about the unintended consequences of building and releasing new tools without an awful lot of testing.
Similar to the nature of the work that scientist J Robert Oppenheimer did on the Manhattan Project, there is a “dual-use dilemma” in the way that AI could either be used for good or to cause catastrophic harm.
Unintended consequences
Similar to fears expressed during that period of history, scientists and researchers frequently warn about ways in which AI can either be misaligned or spiral out of control, or ways in which people can exploit it for malicious intent.
A study published in May 2024, for example, highlights the possible risks of a future AI that could deceive humans. It uses documented examples as a springboard to speculate that a future system may decide to mislead humanity for purposes that differ from the humans that built and use it.
There are also flaws with these parallels, namely, unlike nuclear weapons, AI development spans thousands of organizations around the world. There’s also a dominance in the private sector, a much lower barrier to entry, and more difficulties in verifying and inspecting whether AI is being used in positive or harmful ways.
https://cdn.mos.cms.futurecdn.net/cWGheTBMBnt6hbxVEHjjhB-1090-80.jpg
Source link
