Meta poaching AI talent for new lab
Meta has hired "top engineers" away from Google DeepMind and startup Sesame AI as CEO Mark Zuckerberg staffs up a new "superintelligence" lab, Bloomberg reports, citing anonymous sources. "Frustrated" with the pace of Meta's progress in artificial intelligence, per CNBC, Zuckerberg is offering "seven- to nine-figure compensation packages," The New York Times reported, also citing anonymous sources. Meta just announced a new AI "world model" that can be used in technology such as delivery robots and self-driving cars.
Zuckerberg is offering $100M+ compensation packages to poach AI talent 😳
Yep you read that right.
He's personally creating a new "superintelligence" team dedicated to building the world's most advanced AI platform, and splashing out nine figure packages to hire top talent.
Everyone is an AI expert these days, so he'll have plenty of talent to pick from!
Yep you read that right.
He's personally creating a new "superintelligence" team dedicated to building the world's most advanced AI platform, and splashing out nine figure packages to hire top talent.
Everyone is an AI expert these days, so he'll have plenty of talent to pick from!
💼 Meta poaches Jack Rae (Google DeepMind) and Johan Schalkwyk (Sesame AI) for its new “superintelligence” lab.
🏠 Mark Zuckerberg is personally interviewing recruits at his Palo Alto and Lake Tahoe homes, aiming for a 50-person AGI squad.
💰 A multibillion-dollar stake in Scale AI would bring CEO Alexandr Wang onboard next.
🎯 Mission: leap beyond today’s models toward full artificial general intelligence.
🏠 Mark Zuckerberg is personally interviewing recruits at his Palo Alto and Lake Tahoe homes, aiming for a 50-person AGI squad.
💰 A multibillion-dollar stake in Scale AI would bring CEO Alexandr Wang onboard next.
🎯 Mission: leap beyond today’s models toward full artificial general intelligence.
Meta just dropped its new AI “world model,” and while it might sound like another buzzword, this could be a major leap forward in robotics and autonomous systems.
At its core, a world model is about giving machines a sense of “intuition”—an internal simulation of how the physical world works. Instead of reacting solely to real-time data, AI systems with world models can predict, reason, and plan ahead, much like humans do. For robotics and self-driving cars, this means faster adaptation, fewer failures, and dramatically improved generalization in unfamiliar environments.
This move by Meta signals a shift from data-hungry, narrowly trained models to AI that understands context, cause-and-effect, and future state estimation. It's not just about perception anymore—it's about foresight.
For the robotics and autonomous driving industries, this could unlock new levels of flexibility:
🔹 Robots that can operate safely in dynamic, cluttered spaces
🔹 Self-driving systems that adapt to rare or unpredictable traffic situations
🔹 AI agents that collaborate with humans more naturally
It also hints at Meta's long-term strategy: bridging virtual and physical intelligence. A future metaverse won’t just be about avatars—it may also include real-world machines that “think” more like us.
At its core, a world model is about giving machines a sense of “intuition”—an internal simulation of how the physical world works. Instead of reacting solely to real-time data, AI systems with world models can predict, reason, and plan ahead, much like humans do. For robotics and self-driving cars, this means faster adaptation, fewer failures, and dramatically improved generalization in unfamiliar environments.
This move by Meta signals a shift from data-hungry, narrowly trained models to AI that understands context, cause-and-effect, and future state estimation. It's not just about perception anymore—it's about foresight.
For the robotics and autonomous driving industries, this could unlock new levels of flexibility:
🔹 Robots that can operate safely in dynamic, cluttered spaces
🔹 Self-driving systems that adapt to rare or unpredictable traffic situations
🔹 AI agents that collaborate with humans more naturally
It also hints at Meta's long-term strategy: bridging virtual and physical intelligence. A future metaverse won’t just be about avatars—it may also include real-world machines that “think” more like us.
Google is adding a new role to its leadership ranks: Chief AI Architect.
The company has tapped Koray Kavukcuoglu, Google DeepMind’s chief technology officer, for the new senior vice president position aimed at overseeing faster productization of the company's AI technology.
Kavukcuoglu will “accelerate how we bring our world-leading models into our products, with the goal of more seamless integration, faster iteration, and greater efficiency,” Sundar Pichai wrote in a memo to employees, according to Semafor.
The executive appointment comes as Google faces pressure to create money-making products from its flagship AI technology, much of which is housed in Google’s DeepMind unit. The company has been integrating its DeepMind unit within the broader Google organization since 2023. Since then, AI competition has moved quickly, and though advancements have been made in AI, tech companies haven’t been able to show significant returns.
Capital expenditures, however, have continued growing.
Google is investing $75 billion into AI infrastructure this year, alone. It's trying to mitigate costs elsewhere. Tuesday the company offered buyouts to employees across several organizations, including those working on Search and ads, CNBC reported.
The company has tapped Koray Kavukcuoglu, Google DeepMind’s chief technology officer, for the new senior vice president position aimed at overseeing faster productization of the company's AI technology.
Kavukcuoglu will “accelerate how we bring our world-leading models into our products, with the goal of more seamless integration, faster iteration, and greater efficiency,” Sundar Pichai wrote in a memo to employees, according to Semafor.
The executive appointment comes as Google faces pressure to create money-making products from its flagship AI technology, much of which is housed in Google’s DeepMind unit. The company has been integrating its DeepMind unit within the broader Google organization since 2023. Since then, AI competition has moved quickly, and though advancements have been made in AI, tech companies haven’t been able to show significant returns.
Capital expenditures, however, have continued growing.
Google is investing $75 billion into AI infrastructure this year, alone. It's trying to mitigate costs elsewhere. Tuesday the company offered buyouts to employees across several organizations, including those working on Search and ads, CNBC reported.