Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Meta is reported to have been invested nearly $ 15 billion In the scale of information and starting the AI and starting a 49% stake, the company brings the beginning of the CEO to help a new “Super Title” laboratory within the company.
The deal reminds the previous large and risky bets of methane, for example Obtaining $ 19 billion whatsapp and her Buying $ 1 Billion Instagram. When these compounds are closed, many people have passed a lot for meta platforms, and today’s disclosure has no difference. This weekend does not have a lack of investors and founders who scratches their heads on the latest talist.
Finally WhatsApp and Instagram CEO Mark Zuckerberg’s empire was an integral part of the empire. The question is, the AI transaction met’s grace, zuckerberg’s presscientific strategy or saying that the company will be among the wrong diligence or the company’s caught in the company’s challenges.
In this situation, Meta does not bet on a social media application, but in the data used to prepare the best AI models. Over the past few years, AI laboratories such as AI laboratory and relying on a scale AI to produce and label information used to prepare models. In recent months, the AI scale AI and its information annotation began to hire high quality people to create high quality data for Frontier AI laboratories, and high quality people for Frontier AI laboratories.
It can benefit meta to build a close relationship with a data provider like a scale. Methane leaders complained that a person familiar with the issue is the lack of innovation in the company’s leading EU teams.
Substitute Generative AI unit of methane earlier this year The family of AI models began in Llama 4 China did not match the capabilities of the AI Lab DeepSeek models, and it was widely seen as frustration. It does not help matters, Meta is trying to fight a charming problem. According to the information compiled by Signalfire, Meta lost 4.3% to the best talent’s AI laboratories In 2024.
Meta does not only bet on the scale AI to re-change the effort, but also in Wang to lead the above new Super Management Group. The 28-year-old General Director proved that he was a strong starting founder – a ambitious, good seller and very well connected in the silicone valley. Wang for the last few months Meeting world leaders To discuss EU’s impact on society.
Wang has not led a similar AI lab before, and it does not have the same AI research source as many other AI laboratory leaders like many other AI laboratory Says Sutskever or Mistral Artur Mensch. Perhaps the meta is also said to be hired by high-level talent Like Deepmind’s Jack Rae Round the new AI research group.
The public’s post-purchase fate of the Society’s scale is a bit. The role of real world information in the preparation of AI, some AI laboratories have increased their internal data collection efforts at home, while others trust in synthetic (ie the EU). In April, the information reported this Scale AI misses income targets.
According to any cooriter Robert Nishihara, several border EU laboratories explore Roman ways to use and optimize information, multiple calculating and optimizing information.
“The information is a moving target,” he told Techcrunch in the interview. “It’s not just a last effort to catch – you need to innovate.”
Meta and Wang relations are traditionally feasted from other AI laboratories working with a scale AI. If so, this deal can be a grace for EU’s rivals, such as Turing, Surge AI and even non-traditional data providers such as LM Arena recently launched.
Turing CEO Jonathan Siddharth reported to Techcrunch via e-mail that customers were interested in the lighting of the meta deal with the meta-scale meta scale.
“I think there will be some customers who prefer to work with a partner who is more neutral,” he said.
Only the time of the time investment will be said to be conveyed for its AI efforts, but the company has an important key to makeup. In this time, the competition does not slow down. Openai is preparing for the release of the next flagship model, GPT-5, as well as the first open model of the years – a model that will compete with the current and future LLAM issues.
