One of the “godfathers” of modern artificial intelligence has predicted a further revolution in the technology by the end of the decade, and says current systems are too limited to create domestic robots and fully automated cars.
Yann LeCun, the chief AI scientist at Mark Zuckerberg’s Meta, said new breakthroughs are needed in order for the systems to understand and interact with the physical world.
LeCun spoke as one of seven engineers who were awarded the ￡500,000 (US$625,000) Queen Elizabeth prize for engineering on Tuesday for their contributions to machine learning, a cornerstone of AI.
Photo: AFP
Recent breakthroughs in the sector, led by the launch of OpenAI’s ChatGPT chatbot, have heightened expectations — and fears — of systems gaining human levels of intelligence.
However, LeCun said there was some way to go before AIs matched humans or animals, with the current cutting-edge technology excelling at “manipulating language” but not at understanding the physical world.
“There are still a lot of scientific and technological challenges ahead, and it’s very likely that there’s going to be yet another AI revolution over the next three to five years because of the limitation of current systems,” he said. “If we want eventually to build things like domestic robots and completely autonomous cars, we need systems to understand the real world.”
LeCun is working on systems that attempt to “understand” physical reality by forming a model that can predict how the world behaves. On current progress in AI, he said: “We’re not talking about matching the level of humans yet. If we get a system that is as smart as a cat or a rat, that would be a victory.”
However, LeCun’s fellow QEPrize winner and AI “godfather,” Yoshua Bengio, warned more progress was needed on the technology’s safety and called on next week’s global AI summit in Paris to grasp the issue.
“I’d like to see the leaders of this world to better understand the magnitude of what we are doing, both in terms of the power we’re creating, which could be for good or be dangerous, and the risks that come with that power,” he said.
In 2018, Bengio and LeCun shared the Turing award — seen as the equivalent of a Nobel prize for computing — with Geoffrey Hinton, who was also announced as a QEPrize winner on Tuesday. The award comes after AI pioneers won two Nobel prizes last year. They included Hinton — who shared the Nobel for physics with another recipient of this year’s QEPrize, the US physicist John Hopfield — and the Google DeepMind scientists among the winners of the chemistry prize.
Machine learning is a core process in developing AIs. Instead of being instructed directly, computers “learn” by analyzing data and then making informed decisions or predictions — such as what will be the next in a sequence of words.
The other winners of the 2025 QEPrize are: Li Feifei (李飛飛), the Chinese-American computer scientist who created ImageNet, a crucial dataset for training AIs to recognize objects; Jensen Huang (黃仁勳), the chief executive of Nvidia, the lead maker of chips used in operating and training AI systems; and Bill Dally, the chief scientist at Nvidia.
Patrick Vallance, the chair of the QEPrize foundation and the UK science minister, said the impact of machine learning was being felt across “industries, economies, and the planet,” The annual prize recognizes engineers whose innovations have a “significant impact on billions of lives around the world,” he said.
Taiwan’s English education system is being pulled apart by three opposing forces. Bilingual Nation 2030 pulls students toward English and global communication. Artificial Intelligence (AI) readiness pulls them toward digital judgment, verification and AI-mediated work. But Taiwan’s old exam culture pulls them back toward memorization, grammar drills, timed reading and correct answers. If the education system keeps using old exams to define success, it risks producing graduates who are neither genuinely bilingual nor genuinely AI-ready, but trained for tasks machines can already perform. The first force is Bilingual Nation 2030. Launched in 2018, the policy aimed to “help Taiwan’s workforce connect
It seems every few days one bumps into one of those “real man” comments in which Taiwan is urged to “face reality” or similar, and “make a deal,” with the speaker implying that soon it will be too late. “Deal” advocates always present themselves as having a superior grip on reality, and the manly ability to make the “hard choice.” Their testosterone-laden language often echoes that of Taiwan sellout advocates. Note that such commentary always specifies a process (“make a deal, work with, make progress”), never the end state of what occupation by a violent authoritarian colonialist state will entail. In
There are shadowy cabals plotting to sell out Taiwan to be annexed by China, by invasion if necessary. Fortunately, they are buffoons. In 2019, former Bamboo Union gangster and founder of the China Unification Promotion Party (CUPP), Chang An-le (張安樂, colorfully known as “White Wolf”), led a protest at the Legislative Yuan against comments made by then-premier Su Tseng-chang (蘇貞昌) that in the event of an attack by China, he would never surrender, but would protect the nation by fighting to the end, even if he only had a broom. Chang had party members bring a wooden casket that they
June 1 to June 7 "If all Taiwanese were as afraid of dying as you, then what would happen?” Physician Shih Chiang-nan (施江南) reportedly said this to his wife Chen Chiao-tung (陳焦桐) after she urged him to stop intervening on behalf of Taiwanese soldiers stranded overseas after serving in the Japanese Army during World War II. Shih had clashed with high-ranking officials over the issue, engaged in several heated arguments with Taiwan governor-general Chen Yi (陳儀) and allegedly shouted at general Ko Yuan-fen (柯遠芬), chief of staff of the Taiwan Garrison Command, over
RSS