Meta AI boss Yann LeCun sounded off on the industry-wide state of synthetic intelligence (AI) and quantum computing throughout a latest occasion to rejoice the 10-year anniversary of the founding of Meta’s Fundamental Artificial Intelligence Research (FAIR) workforce.
During LeCun’s commentary, he commented on Nvidia’s present stranglehold on the AI {hardware} {industry}, the chance that human-level AI will emerge in the close to future, and why Meta isn’t at the moment pursuing quantum computing alongside its rivals.
By “not any time quickly”, I imply “clearly not in the subsequent 5 years”, opposite to various people in the AI {industry}.
Yes, I’m skeptical of quantum computing, significantly on the subject of its utility to AI.https://t.co/5t63w1GNfL— Yann LeCun (@ylecun) December 3, 2023
The synthetic intelligence conflict
LeCun’s views on the imminence of so-called human-level AI are well-documented. He’s lengthy maintained that we’ll want to realize “dog-” and “cat-” degree AI earlier than the discipline advances sufficient to assist human-level AI. And he’s to this point been reticent to offer predictions on when these early milestones will occur.
By comparability, Elon Musk lately made the daring prediction that a “digital god” would arrive within the next three to five years.
In the center, maybe, lies Nvidia CEO Jensen Huang. He lately stated that AI would have the ability to full assessments in a fashion “fairly competitive” with people in the subsequent 5 years. While this stops wanting claiming full human-level capabilities, some consultants view test-taking as a measure of an AI’s capabilities.
LeCun, apparently, is not amongst them. During FAIR’s 10-year celebration, he gave commentary associated to the Nvidia CEO’s assertions:
“I know Jensen. There’s an AI war and he’s supplying the weapons.”
The Meta AI boss’ assertion seemingly refers to the latest information that Nvidia is now the world’s most beneficial chip producer, a feat largely attributed to the firm’s GPUs changing into the established order {hardware} for coaching massive language fashions corresponding to ChatGPT.
LeCun additional defined that the present know-how cited as a possible spark for AGI, generative AI, was merely not adequate. “Text is a very poor source of information,” he stated, including, “Train a system on 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.”
Related: Nvidia posts record $18B Q3 revenue, cites generative AI as primary driver
Quantum computing
Another space LeCun touched on throughout the occasion was quantum computing. Unlike rivals Google and Microsoft, Meta has strayed comparatively removed from the quantum computing race.
“Quantum computing is a fascinating scientific topic,” stated LeCun, however different feedback made it clear that it was his view that the know-how wasn’t ripe. “The number of problems you can solve with quantum computing you can solve way more efficiently with classical computers.”
Meanwhile, Microsoft lately entered right into a $100-million partnership with Canadian quantum computing agency Photonic to deliver a fault-tolerant, absolutely practical quantum networking system to market within the next five years.
