Adapted by:
admin
Administrator
Alba Mora
WriterAn established tech journalist, I entered the world of consumer tech by chance in 2018. In my writing and translating career, I've also covered a diverse range of topics, including entertainment, travel, science, and the economy.
Meta CEO Mark Zuckerberg is feeling very frustrated with his company’s progress in artificial intelligence. This dissatisfaction has led him to take personal action by hiring a new team of AI experts and researchers to redirect the company’s trajectory. According to Bloomberg, Meta’s Llama 4 hasn’t met expectations, prompting Zuckerberg to get heavily involved in the creation of this new team. His ultimate goal is to develop artificial general intelligence to maintain a competitive edge.
Top priority. Zuckerberg has entered the so-called “founder mode.” This means he’s not delegating the development of Meta’s new superintelligence team. Instead, he’s personally involved in every step of the process.
He’s reportedly created a WhatsApp group with other executives to discuss potential candidates and is the first to reach out to them. He’s even met some candidates at their homes in Lake Tahoe and Palo Alto and has reorganized Meta’s headquarters in Menlo Park so that the new team sits close to his office for increased privacy.
The team is expected to comprise approximately 50 people, including researchers, infrastructure engineers, and other professionals. However, it remains unclear how this new team will integrate with Meta’s existing AI teams or if there will be any layoffs.
Discontent with Meta’s progress in AI. In April, Meta announced Llama 4, its large language model aimed at competing with others like GPT-4. However, it faces a challenging task in unseating ChatGPT, which, at that time, was the world’s most downloaded app.
Despite the dedication of Meta’s AI team, who worked nights and weekends, the results have been disappointing. This dissatisfaction has caused a delay in the launch of Llama 4 Behemoth. The company is developing a larger model because the current version doesn’t meet the desired standards. Frustrated by this situation, Zuckerberg has decided to assemble a new team.
Goal: general artificial intelligence. In early 2024, Zuckerberg revealed his ambitious long-term goal: achieving general artificial intelligence. He also said Meta wanted to “open source it responsibly, and make it widely available so everyone can benefit.” General artificial intelligence is an even more challenging concept. It refers to AI that is capable of performing any intellectual task, making judgments, and reasoning with greater capacity than humans.
As many companies vie for leadership in the next significant leap in AI, Meta is determined not to fall behind. While Zuckerberg is committed to reaching this goal, it seems unrealistic given that previous launches haven’t met his expectations.
A billion-dollar investment. Alongside forming the new team, Zuckerberg is reportedly planning to make a major investment in Scale AI. This company provides services to train AI models and develops apps for businesses and governments. This investment could enable Meta to improve the quality of its AI by using better data labeling. Although unconfirmed, rumors suggest that the deal is at an advanced stage and could be worth as much as $15 billion, making it Meta’s largest external investment to date.
Zuckerberg seems unfazed by the financial implications of these investments. According to Bloomberg, he’s reassured potential recruits that Meta’s advertising business remains strong and that the company has ample liquidity to fund its ventures in the AI sector. Some of these investments may include the creation of a massive data center, which could cost up to $200 billion. While Meta initially denied this claim, if confirmed, it would become the most expensive data center ever built, surpassing Amazon’s planned $100 billion facility.
Image | Meta
Related | Meta Is Turning to Nuclear Power for Its AI Data Centers With New 20-Year Agreement
View 0 comments