AI has become a hot topic of discussion at Davos Forum, and all walks of life are more concerned about the accuracy of results.

According to news on January 19, at the World Economic Forum 2024 Annual Meeting held in Davos, Switzerland, the world’s largest technology companies, including Salesforce, Microsoft and Google, demonstrated their capabilities in artificial intelligence. In terms of technical strength, generative artificial intelligence has also become the focus of discussion at this year’s conference.

Intel CEO Pat Gelsinger said in an interview that although the capabilities of generative artificial intelligence technology have emerged in 2023, the industry’s focus in 2024 should be on improving The results are accurate so that executives in high-risk industries such as hospitals and manufacturing can freely apply artificial intelligence.

“The current application of artificial intelligence has reached its peak,” Gelsinger said. "I believe that the next stage of development of artificial intelligence is to establish regulatory accuracy in the underlying model."

Gelsinger emphasized that whether it is doctors relying on general artificial intelligence to diagnose diseases, or warehouses using This technology is used to check assembly line faults or self-driving cars. Humans need to be more familiar with the accuracy of artificial intelligence technology.

"Some of the problems in artificial intelligence have been solved so far, but there are still many problems," Gelsinger said. "Basic prediction, detection, visual language, those problems have all been solved now. But there are a lot of other problems. Like how do you prove that a big language model is actually correct? There are a lot of bugs out there. So you still need to be clear, Essentially, artificial intelligence technology is improving the productivity of knowledge workers. But in the final analysis, I need knowledge workers to determine that the results are correct." Clara Shih, CEO of Salesforce AI, said that improving accuracy The best way is to increase adoption through various experiments and manual collaborative testing. Shi Zongwei said in an interview that artificial intelligence needs to be adjusted to different standard deviation confidence levels so that users can recognize the credibility of artificial intelligence technology in various high-risk scenarios.

Shi Zongwei said that the adoption of general artificial intelligence will be divided into three stages. The first stage is to actively use artificial intelligence technology as an auxiliary means in work. The second phase is to purposefully observe the AI ​​in autonomous mode to ensure accuracy. The final stage, she said, is letting go and trusting that the technology will work, achieving the level of trust that people need to choose it.

Shi Zongwei explained: "You can ask the AI ​​to be conservative in high-risk situations until human assistance trains it to operate automatically."

Open AI CEO Sam This three-stage approach, which relies heavily on human acceptance of the reliability of the technology, should make general artificial intelligence less powerful than some think, Sam Altman said during a panel discussion on Tuesday. So terrible.

“This is more of a tool than I expected,” Altman said. "It's going to get better, but it's not replacing jobs yet. It's an incredible tool in terms of increasing productivity. It's a tool that expands human work, by using artificial intelligence to do some of the work, and also allows people to do better Get the job done."

Nasdaq CEO Adena Friedman said in an interview that the past year has been a "year of discovery" for artificial intelligence technology. . Friedman said that the financial industry, including Nasdaq, will use artificial intelligence to update old codes, upgrade aging systems, and improve automated workflows, which can save employees a lot of time every day.

“It came out more than a year ago,” Friedman said. "We did some experiments. We started to understand its potential. It's going to be a very dynamic year for us and for everyone."

The above is the detailed content of AI has become a hot topic of discussion at Davos Forum, and all walks of life are more concerned about the accuracy of results.. For more information, please follow other related articles on the PHP Chinese website!