Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124

It is said that Meta is investing almost 15 billion dollars At Scale Ai Scale AI and takes 49 % at startup, with Alexandr Wang CEO to help lead the new “SuperIntellign” laboratory within the company.
The deal reminds us of the big stakes and the previous risks in Meta, like it Earn 19 billion dollars from WhatsApp and Buy Instagram worth $ 1 billion. While this merger was closed, many people suggested that Mita had greatly pushed the platforms – and today’s speech is no different. There is no shortage of investors and founders who left their heads this week on the latest dead connections.
In the end, WhatsApp and Instagram became an integral part of the Empire CEO of Mark Zuckerberg. The question is whether the Scale AI deal will actively work for Meta, which proves the Zuckerberg Prescient strategy again, or whether the company accommodates straw in a misleading attempt to catch competitors like Openai, Google and Anthropic.
In this case, Meta does not bet on the emerging social media application, but on the data used to train the top AI models. Over the past few years, the leadership of artificial intelligence laboratories such as Openai has adopted on the artificial intelligence scale for production and naming data for models training. In recent months, Scale AI and my competitors have started explaining data in employing high skill, such as doctoral scientists and senior software engineers, to create high -quality data for Ai Frontier laboratories.
Meta may help you have a close relationship with a data provider like Scale. Meta leaders complained that there was no innovation about data in the leading AI teams in the company, according to a person familiar with the matter.
Earlier this year, Amnesty International Unit in Meta Llama 4 launched an artificial intelligence family This failed to match the capabilities of the Chinese Ai Deepsek models and was widely seen as disappointment. Do not help things, Meta is trying to combat the problem of attrition. According to the data collected by SignalFire, Meta lost 4.3 % of its best talents for artificial intelligence laboratories In 2024.
Meta is not only betting on the Amnesty International scale to clarify AI’s efforts, but also on Wang to lead the new SuperINGEGENCE team mentioned above. The 28-year-old CEO has proven that he is a strong founder to start operating-known about Silicon Valley as ambition, good sales, and well connected. During the past few months, Wang was A meeting with world leaders To discuss the impact of artificial intelligence on society.
However, the Amnest International Laborator has not previously led before, and he has no same background of artificial intelligence research that many other artificial intelligence laboratory leaders, such as Elia Soskfar from Safe Superintelligence or Arthur Mensch from Mistral. Perhaps for this reason it is said that dead recruits a high -level talent Such as DeepMind’s Jack Rae To assemble the new artificial intelligence research collection.
The fate of the post -acquisition from AI from the artificial intelligence company is a little mysterious. Data role in the real world in training the artificial intelligence model-some AI laboratories have brought data collection efforts at home, while others have increased their dependence on artificial data (IE, which has been created). In April, the information mentioned this Scale AI has missed some of its revenues.
According to the co -founder of Anyscale Robert Nishihara, many Ai Frontier laboratories explore new ways to take advantage of the data and improve it, many of which are calculated.
“The data is a moving goal,” Nishahara told Techcrunch in an interview. “It is not just a limited effort to catch a knee – you have to innovate.”
The Meta and Wang’s relationship can be able to scare other Amnesty International laboratories that have traditionally worked with Scale AI. If so, this deal may be a blessing for Scale AI, such as Turning, Turge AI and even unconventional data providers such as the recently launched LM Arena.
The CEO of Turning Jonathan Siddharge told Techcrunch via email that he had received increasing attention from customers in light of rumors about the Meta deal with Scale Ai.
“I think there will be some customers who prefer working with a more neutral partner,” he said.
The only time will be determined by Meta’s investment for its efforts of artificial intelligence, but the company has a large floor for compensation. Meanwhile, competition is not slowing down. Openai is preparing to release its next main model, GPT-5, as well as the first publicly available in years-a model that competes with the current and future Lama versions in Meta.