Meta, the parent company of Facebook, Instagram and WhatsApp led by CEO Mark Zuckerberg, made major announcements this week regarding its artificial intelligence (AI) efforts and plans to develop an advanced general intelligence system.
Zuckerberg Merges AI Teams, Accelerates Work Towards AGI
On Wednesday, Zuckerberg announced a reorganization and consolidation of Meta’s AI division, with the company’s Fairness, Transparency and Accountability in AI team being moved under the purview of Meta’s Chief Product Officer.
This restructuring aligns with Zuckerberg’s aim to accelerate Meta’s progress towards developing artificial general intelligence (AGI) – AI systems that can reason, plan, solve problems and think abstractly across a wide range of domains, similar to human intelligence.
In an interview with The Verge, Zuckerberg framed Meta’s AGI push as one of the “most interesting and promising” areas he wants the company to pursue more ambitiously. He expressed a desire to “focus the company on building for the metaverse first, but also having this mission within Reality Labs focused on the next generation computing platform and artificial general intelligence.”
Massive Investment in AI Hardware
To power its AI ambitions, Meta made huge announcements about expanding its AI computing capabilities through strategic partnerships and chip investments.
Zuckerberg revealed that Meta will have access to over 600,000 Nvidia H100 GPUs by the end of 2024, with plans to acquire a total of 350,000 H100s specifically.
The Nvidia H100 is the company’s latest and most advanced AI accelerator chip for training large machine learning models.
This amounts to a multi-billion dollar investment by Meta, with Nvidia H100 GPUs costing up to $209,000 each.
Additionally, Meta will also utilize AMD’s next-generation MI300 GPU for AI training and inference workloads.
Overall, Meta is on track to have the largest AI supercomputer infrastructure in the consumer internet industry, powered by over 600,000 GPUs total by year’s end.
Concerns Over Safety of AGI Development
While Zuckerberg and Meta are charging full speed ahead on AGI, concerns have been raised by AI experts over the safety and ethics of developing human-like artificial intelligence.
Yoshua Bengio, a renowned AI pioneer who received the Turing Award in 2019, called out Zuckerberg’s statements on AGI as “irresponsible”, given the risks associated with advanced AI systems.
Meta’s push for AGI and its consolidation of AI ethics teams under product executives has alarmed industry observers, who worry it prioritizes progress over responsibility. Stanfords’s AI Index founder Yoav Shoham warned that Meta may be underestimating safety challenges in its race to develop human-level intelligence.
OpenAI CEO Sam Altman also expressed concern over AGI developed by a single company, given its potentially enormous societal impacts.
These apprehensions over Meta’s approach indicate that while rapid progress in AI capabilities brings excitement, responsible development minimizing downside risks should be the priority. Meta’s decisions to make its work open-source may help increase transparency and trust.
The Path Forward – AGI Assistants and The Metaverse
While Meta’s AGI system remains years away from fruition, Zuckerberg has a vision for how human-like intelligence could transform the company’s products and vision for the metaverse.
Zuckerberg believes life-like VR avatars and digital companions powered by AGI could become mainstream by the end of the decade. He envisions AR glasses as the “holy grail” for delivering helpful AI assistants to billions of people.
As Meta dedicates tremendous resources to establish itself as a leader in AI, it remains to be seen whether responsible development practices keep pace with technological achievements. Meta’s path will have profound impacts on the future of artificial intelligence and its integration into peoples’ lives.
To err is human, but AI does it too. Whilst factual data is used in the production of these articles, the content is written entirely by AI. Double check any facts you intend to rely on with another source.
