How Meta is reshaping its operations to become an AI powerhouse
California, May 25
,: Last week, Meta laid off 8,000 employees from its global workforce and transferred another 7,000 to new artificial intelligence roles, kick-starting a process of reshaping its operations as it gets battle-ready to stay ahead of competition in the war for AI dominance.
As it ramps up capital expenditure and plans to spend up to $135 billion on artificial intelligence infrastructure, the layoffs and the restructuring point to the scale of the AI pivot at the social media behemoth.
Its founder and the Chief Executive, Mark Zuckerberg, has drafted his man Friday, Andrew Bosworth, to undertake the mammoth exercise with a clear mandate to make Meta an AI-first company, WSJ reported. Andrew Bosworth - or Boz, as most people know him - is Chief Technology Officer at Meta, and Head of Reality Labs, the company's wearables and metaverse organisation.
The understanding at the top deck of Meta is clear that the next leg of growth will come from AI, and focus on more high-end chips, data centres and hiring top talent will determine the future of the organisation.
Meta is one of the many Silicon Valley tech players that have put AI at the centre stage of their existence. On May 18, Blackstone announced a joint venture with Google to set up a new AI cloud company, with the alternative asset manager pouring in an equity infusion of $5 billion into the company.
The new company will offer efficient data centre capacity, operations, and networking, as well as Google Cloud's Tensor Processing Units (TPUs), as a compute-as-a-service offering, Blackstone said.
The AI game is fast shifting beyond the chatbots to agentic functions. In fact, agentic AI has completely changed the game for tech players as they vie to gain a bigger share of the pie. That's the next frontier for AI players.
Agentic AI is capable of functioning autonomously, performing different functions and doesn't rely on prompts to generate responses.
In a new policy, Meta informed employees through an internal memo about a tracking system that will be installed on the company's computers and internal apps to track keystrokes and mouse clicks to help train AI models for performing some of the functions autonomously.
Bosworth had informed the employees through a separate memo that the company would accelerate the internal data collection as part of "AI for Work" efforts, now rebranded as Agent Transformation Accelerator (ATA), Reuters had reported.
He also told the employees that Meta is moving towards a future where agents primarily do the work, the WSJ report said.
"Our role is to direct, review, and help them improve," the report quoted him as saying.
Bosworth also spearheaded Zuckerberg's "Metaverse" project on virtual reality. However, despite a lot of attention and funding, it never really shaped up.
However, Bosworth continues to enjoy Zuckerberg's confidence and is now focusing entirely on scaling AI at the Facebook and WhatsApp parent to put the company firmly on the AI map.
— ANI
Reader Comments
135 billion dollars on AI infrastructure? That's more than the GDP of many countries! 🤯 As someone who works in tech in Bangalore, I can see the AI wave happening here too. But I'm skeptical about this 'agentic AI' thing - if their metaverse flopped (and let's be honest, it kinda did), what makes them think this will be different? Zuckerberg is just throwing money at problems.
The tracking of keystrokes and mouse clicks is a privacy nightmare! 🚨 I understand they need data to train AI, but this feels like a step too far. In India, we already deal with so much data privacy issues - Aadhaar, WhatsApp snooping, etc. Meta needs to be more transparent about how they'll use this employee data. Just because it's for 'training AI' doesn't make it okay.
I work in a startup here in Gurgaon and we're already using AI agents for customer support. The potential is huge - imagine AI handling all the data entry, report generation, and repetitive tasks. But the idea that 'agents will primarily do the work' while humans just 'direct, review, and improve' them? That sounds like a recipe for massive unemployment. Where will the 7,000 transferred employees end up when the AI takes over their new roles too?
Interesting to see Blackstone partnering with Google for AI cloud services. That $5 billion investment is huge! But I'm concerned about the concentration of power. Meta, Google, Microsoft - these giants are all racing to dominate AI, and it's creating a monopoly-like situation. In India, we need more homegrown AI initiatives like what NASSCOM and IITs are doing. We can't just be consumers of Western AI technology forever.
We welcome thoughtful discussions from our readers. Please keep comments respectful and on-topic.