Meta, under Mark Zuckerberg's leadership, is making a significant $14.3 billion investment in ScaleAI, acquiring a 49% stake. This is arguably Meta's most ambitious move since purchasing WhatsApp, aiming to build a 'superintelligence' AI laboratory. The acquisition includes bringing ScaleAI's founder, Alexandr Wang, and his team to Meta.
Meta's Chief AI Scientist, Yann LeCun, expresses caution, suggesting that achieving Artificial General Intelligence (AGI) – AI surpassing human intelligence – could take a decade or more. This contrasts sharply with the more optimistic timelines predicted by figures like Sam Altman (OpenAI) and others who believe AGI is achievable much sooner. LeCun argues that current AI focuses too much on marketing and subscription-based models and neglects the foundational development needed for true AGI.
ScaleAI, founded by Wang in 2016, provides high-quality training data to major tech companies. However, past investigations have revealed that ScaleAI employed a large workforce in the Philippines at wages significantly below the minimum wage to annotate and refine massive datasets for its AI models.
Zuckerberg's decision to acquire only 49% of ScaleAI is strategic, avoiding potential regulatory hurdles. The substantial investment, representing a significant portion of Meta's revenue and net profit, shows Zuckerberg's commitment to achieving leading-edge AI capabilities. This move comes after the company's substantial investment in the metaverse which failed and resulted in significant stock losses.
The article highlights the ongoing debate within the AI industry, pitting cautious experts like LeCun against more aggressive proponents like Altman. Zuckerberg's bet on ScaleAI is a major step toward reshaping Meta's future in the competitive AI landscape. The success of this venture remains uncertain, given the differing views on timelines and challenges related to achieving AGI.