While the tech world celebrates Nvidia's staggering earnings, a quieter but potentially more significant shift is occurring in artificial intelligence. Behind the headlines of soaring stock prices and record revenues, a debate is intensifying about the fundamental direction of AI development.
The Nvidia Success Story
This week, Jensen Huang, the charismatic leader of Nvidia, dominated business news as his company revealed astonishing third-quarter results. The chipmaker reported revenues of US$57 billion, representing a 62 percent increase compared to the same period last year.
This performance sent technology stocks soaring, reversing earlier concerns that AI valuations had entered bubble territory. Huang confidently addressed these fears, stating that from his company's perspective, they see "something very different" given the extraordinary demand for Nvidia's AI chips.
The LeCun Counterpoint
While Huang captured the spotlight, another significant development emerged from Yann LeCun, the celebrated AI pioneer and Meta's chief scientist. The Financial Times confirmed that LeCun will soon leave his position at Meta to launch his own startup.
This move follows Meta founder Mark Zuckerberg's decision to appoint 28-year-old Alexandr Wang to lead the company's new "superintelligence" team, creating an unusual reporting structure where the 65-year-old LeCun would answer to a member of Gen Z.
A Different Vision for AI's Future
More significant than his career move are LeCun's technical plans. In recent years, the AI race has been dominated by large language models (LLMs) based on transformer technology outlined in a groundbreaking 2017 paper. This approach has powered revolutionary products like ChatGPT.
However, LeCun believes LLMs are approaching their limits. Instead, he advocates for "world models" – an approach to information processing inspired by human learning patterns. "LLMs are great, they're useful, we should invest in them," LeCun noted this month. "[But] they are not a path to human-level intelligence... so for the next revolution, we need to take a step back."
This philosophical divergence comes at a critical moment for the AI industry. While current corporate enthusiasm for AI remains high, surveys indicate wide variation in whether companies can actually achieve meaningful productivity gains from existing AI technologies.
The simultaneous announcements from Huang and LeCun represent two competing visions for artificial intelligence's future – one celebrating current commercial success, while the other looks beyond present limitations toward more human-like intelligence.
