Mark Zuckerberg Sets Sights on Creating General AI: A New Challenge for Meta
Mark Zuckerberg, the CEO of Meta, has revealed his company’s ambition to develop artificial general intelligence (AGI), placing them in direct competition with other tech giants like OpenAI. In an interview with The Verge, Zuckerberg admitted that he does not have a precise definition of general AI nor does he know when they will achieve it.
The primary goal for Meta is to recruit top AI researchers who are motivated to work on this groundbreaking technology. Zuckerberg emphasized the importance of tackling ambitious problems to attract the best talent.
Meta is not only focused on talent acquisition but is also ramping up its resources for AI development. Zuckerberg revealed that the company aims to possess a staggering 600,000 Nvidia GPUs by the end of 2024, ensuring a scale of operation unmatched by any other individual company.
However, despite the ambitious pursuit, Mark Zuckerberg himself is uncertain about the exact capabilities and boundaries of artificial general intelligence. He believes that the key aspect lies in its comprehensive nature: encompassing reasoning, intuition, and a wide range of cognitive abilities.
When asked about the timeline for the arrival of general AI, Zuckerberg acknowledged that its development will be a gradual process rather than an abrupt breakthrough. He cited the example of generative AI, which has made significant strides in recent years.
Meanwhile, Meta continues its work on Call 3, the AI model that will eventually replace Llama 2. Significantly, this new model will possess a unique feature: the ability to generate code.
Interestingly, the current Meta language model lacks coding capabilities, which the company deemed irrelevant to user needs. However, Zuckerberg now recognizes that coding is crucial in understanding the rigor, hierarchical nature of knowledge, and logic.
In terms of the AI’s accessibility, Zuckerberg discussed the possibility of making Meta’s general AI an open-source model versus a closed-source model limited to Meta’s use. Each option carries advantages and disadvantages. While a closed-source AI may hold greater value, an open-source one can address inequality in terms of access and opportunity. However, Zuckerberg stressed the importance of prioritizing safety and responsibility when making a decision.
Ultimately, Mark Zuckerberg’s inclination leans toward open-source as long as it aligns with safety and responsible practices. He emphasized that he doesn’t wish to be bound by previous statements, emphasizing the importance of flexibility.
As Meta ventures into creating general AI, the tech industry eagerly awaits its impact on the future of technology and its potential applications across various domains. While the specifics remain uncertain, it is clear that Mark Zuckerberg and Meta are determined to be at the forefront of this revolutionary field.
With this new challenge, Meta is poised to reshape the AI landscape, pushing the boundaries of technology and our understanding of artificial intelligence as a whole.
