While the rapid development of artificial intelligence has some concerned about the dangerous scenario of the technology being smarter than the humans who created it, some experts believe that AI has in some ways already reached that stage. I think there are.
“If you define it as performing intelligent but repetitive and limited problems, they are already smart. The best chess players and Go players are machines. And soon, all such tasks Examples include legal analysis, simple writing, photos on demand, and advanced preparedness and threat response simulations. Phil Siegel, the center's founder, told Fox News Digital.
Siegel's comments came after a new survey of nearly 2,000 AI experts found that opinions are divided on when the technology can outperform humans. It is something. To narrow down how smart AI is, respondents were given a list of human tasks ranging from writing high school history essays to fully automating all human tasks, and asked when the AI would perform the tasks. I was given the task of predicting whether I would be able to do so.
IMF warns AI will impact 60% of US jobs and increase inequality around the world
Robots from the Artificial Intelligence and Intelligent Systems Laboratory of Italy's National Computer Science Interuniversity Consortium were exhibited at the 7th Maker Faire. (Andreas Solaro/AFP via Getty Images)
Experts said the technology could already accomplish this feat within the next two years for some tasks, such as high school history essays. However, research shows that we are still a long way from replacing all human labor, with most experts predicting that such a feat will not be achieved by AI this century.
“They can write short stories now, but to write a bestseller they need more information about human nature. They can write a movie, but it probably won't be a hit. “You can write a scientific paper, but you can't carry out all the instructions,” Siegel said of current AI platforms.
“Maybe one day they'll be able to do those things, but training them to be good at something like that requires a lot of data. And then there's probably another level — human “It's about training people to be able to read their true nature on the spot and make complex decisions like the following.” “Management of companies and universities. The level of human training is very complex, and it is important to have a thorough understanding of these tasks.” Because they aren't, it can take a very long time and a lot of computation to excel at these tasks.”
“It's not a question of if or when AI will outperform us. We simply cannot compete with raw processing power.”
Samuel Mangold Rennett, staff editor at The Federalist, shared a similar sentiment, noting that some AI platforms can already perform tasks that are impossible for humans.
“AI is a relatively new field, and products like ChatGPT can already perform complex tasks and solve problems in seconds that would require months of complex thinking and a lifetime of practice for humans. , in some ways, AI is already outsmarting us,” Mangold-Renett told Fox News Digital. “Artificial general intelligence (AGI) is something else to consider. In theory, AGI could exceed all human intellectual capacity and perform all economically important tasks. We might as well ask whether we can gain sentience and what we can achieve.'' This has implications for humanity. ”
Some experts believe that the arrival of a world in which AI outperforms human creators is inevitable, and discussions have begun about how such technology will change society.
A Team Tartan Rescues CHIMP (CMU Highly Intelligent Mobile Platform) robot uses a hand-held power tool during a cutting operation during the Defense Advanced Research Projects Agency Robotics Challenge at the Fairplex in Pomona, California, June 6, 2015. (Chip Somodevilla/Getty Images)
Artificial intelligence and U.S. nuclear weapons decisions: How big is the role?
“It doesn't matter if or when AI will outsmart us. We simply can't compete with raw processing power,” said John Schweppe, policy director at the American Principles Project, on Fox News Digital. told. “This is the charm and value.” [added] Abbreviation for AI. The ability of a computer to process data and produce output in a way that is much faster and more efficient than when humans do the work. But this clearly has an incredible impact on society, both good and bad, so it's important that lawmakers guide technology companies and help them chart a responsible course. Dew. ”
Christopher Alexander, chief analytical officer at Pioneer Development Group, warned that some of these developments can be dangerous, especially if the technology falls into the hands of less responsible actors.
“Our obsession with the Skynet hypothetical situation is derailing the serious policy discussions we need right now to develop and deploy AL responsibly.”
“By policy design, autonomous weapons systems in the United States are not allowed to kill humans without human approval, but consider this very likely scenario: A defense contractor has developed an AI that can control autonomous vehicles. Develop. The project was canceled and the AI was said to be “incredibly effective, but has some flaws. The Chinese, who have already stolen trillions of dollars of US intellectual property over the past decade, The Chinese used flawed AI for autonomous drones and it ran.” [amok]”It's not the end of the world, but it's certainly a possibility,” Alexander told Fox News Digital.
But Jake Denton, a researcher at the Heritage Foundation's Center for Technology Policy, told Fox News Digital that some of the more extreme predictions about the dangers of AI are overstated.
A robot plays soccer in an artificial intelligence demonstration in the stands of the German Artificial Intelligence Research Center. (Sean Gallup/Getty Images)
CLICK HERE TO GET THE FOX NEWS APP
“At this point, fears that superintelligence Al will bring about some form of technodystopia feel misplaced. These sci-fi doomsday scenarios are no exception to the real and immediate challenges we face in Al policy today.” “It's a huge distraction from the problem,” Denton said. “Our obsession with the Skynet hypothetical situation is derailing the serious policy discussions we need right now to develop and deploy AL responsibly.”
Denton listed several ways that AI can be developed responsibly, including transparency standards, open sourcing of underlying models, and policy safeguards.
“Advances in AI need not be catastrophic or dystopian. In fact, these technologies can significantly enhance and improve human productivity and performance across industries. They don't need to replace human workers, but rather can amplify human capabilities,” Denton said. “Our path forward is not to replace workers with workers, but to augment them. There is an opportunity to improve.”
