On This Day in AI: January 18, 2024
On January 18, 2024, Meta chief executive Mark Zuckerberg announced that the company was making advanced general intelligence a central research and product goal. In a public video and interview, he said Meta would bring its two major AI research groups closer together and continue building models intended for broad use across the company’s services. He also said training of Llama 3 was underway and described plans to acquire roughly 350,000 Nvidia H100 graphics processors by the end of the year, with other chips bringing Meta’s total computing capacity to the equivalent of about 600,000 H100s. The scale signaled that Meta intended to compete directly at the frontier of model development.
The term artificial general intelligence has no single accepted technical test. Zuckerberg described it in practical terms as systems able to reason, plan, learn, remember, and support many cognitive tasks rather than one narrow application. Meta’s approach also differed from some rivals because the company had released Llama model weights under licenses that allowed extensive outside use. Zuckerberg said openness could help the technology become widely available, though he stopped short of promising that every future model would be released without restrictions. The announcement connected long-term research with immediate products, including assistants, creator tools, advertising systems, and smart glasses.
The January 18 statement was a strategic commitment, not proof that Meta had created AGI. Computing purchases could increase training capacity, but they did not guarantee capability, reliability, or safety. Open model releases also created a difficult balance: researchers and businesses could inspect and adapt the systems, while bad actors might do the same. Meta later released Llama 3 models and continued expanding its infrastructure, validating part of the near-term plan while leaving the broader AGI claim unresolved. The announcement mattered because it reordered one of the world’s largest technology companies around frontier AI and made access to chips, electricity, data, and scientific talent part of its public strategy. It also intensified debate over whether advanced models should be open, controlled through cloud services, or released according to capability-based safeguards. More broadly, the event showed how quickly AGI had moved from a speculative research label into corporate planning, capital spending, and products used by billions of people.
The restructuring also brought Meta’s long-established fundamental research organization closer to the teams building product models. That connection promised a faster route from scientific advances to services used at global scale, but it increased the importance of evaluation before deployment. A research result and a consumer feature face different standards for privacy, reliability, and abuse. Meta’s announcement therefore joined an ambitious technical objective to an equally large governance challenge inside the same company.
Zuckerberg described systems able to reason, plan, learn, remember, and support a wide range of tasks. No universally accepted AGI benchmark accompanied the announcement.
The goal connected fundamental research with assistants, creator tools, advertising products, and wearable devices across Meta's existing businesses.
Meta planned to hold about 350,000 H100 processors by year-end, with other accelerators bringing total capacity to the equivalent of roughly 600,000 H100s.
That scale demonstrated ambition but did not by itself guarantee more capable, reliable, or safe systems. Models also depend on data, research choices, evaluation, and deployment controls.
Meta had allowed broad outside use of Llama model weights and presented openness as a path to wider innovation and scrutiny.
The same access could enable misuse, leaving future releases to balance research benefits, commercial competition, and safeguards for increasingly capable systems.
Explore more of "On This Day ..."
Discover more events from the same date across news, politics, technology, sports, and other fields. Each link highlights significant moments that shaped history on different fronts.