Yan Lekun, chief AI scientist at tech giant Meta, recently offered a bleak view of the future of artificial intelligence and quantum computing—one of the more optimistic (and more pessimistic) predictions of the tech world.
During the 10-year anniversary meeting of the Meta Basic AI Research Group, Lekun spoke about the current state and future of AI and contrasted it with conventional wisdom.
“Train the equivalent of 20,000 years of reading material, and they still don't understand that if A is the same as B, then B is the same as A,” he said.
LeCun emphasized the significant gap between today's AI capabilities and the prospect of achieving human-level intelligence. Some may think that AI will save or destroy the world, but for LeCun, it's more likely that we'll have “cat-level” or “dog-level” AIs in the coming years.
Real intelligence, he says, requires a large amount of information that goes beyond text and other audiovisual resources.
Lekun is the type of researcher who always prefers to keep his expectations as low as possible – without ever looking at the bigger picture. Speaking at the World Science Festival two weeks ago, he said the amount of power required to reach the level of human intelligence cannot be replicated with the computers we have today.
Even so, he admits that AGI is within reach in the future, not as much as many think.
“There is no doubt that at some point in the future, perhaps decades from now, we will have AI systems that are as intelligent as humans in areas where humans are intelligent,” he asserted.
About fifty
LeCun also expressed skepticism about the immediate benefits of quantum computing, which is attracting significant investment from tech giants including Nvidia, Google and IBM. He argued that many of the problems believed to require quantum computing could be solved efficiently using classical computers—a view shared by Meta's former chief technology officer, Mike Schroepfer.
“Quantum computing is a fascinating scientific topic,” LeCun said.
Quantum computing is a form of computing that uses quantum-mechanical phenomena such as superposition and entanglement to perform operations on data. This approach is fundamentally different from classical computing, which relies on bits in 0 or 1 states.
If the technology is done right, quantum computers could solve problems in seconds that would take thousands of years with today's most powerful supercomputers. That means instant encryption codes, high-fidelity real-time simulations, and excellent AI training.
Lekun's cautious stance suggests a more balanced approach to AI and quantum computing in a field often filled with revolutionary narratives. While progress is being made, he warns that the road to mature AI is longer and more complex than we think.
Edited by Ryan Ozawa.
Stay on top of crypto news, get daily updates in your inbox.
