Google Restructures Company Departments Around Future AI Development
Google has reorganized the company’s departments to further focus on AI as Google Gemini and related projects infuse more and more of Google’s project plans. Google CEO Sundar Pichai shared publicly his plans for the revamped division of work in an email to the company, aptly titled “Building Our AI Future.”
Google is AI
The restructuring revolves around key shifts across four main domains: Models and Research, Responsible AI, Platforms and Devices, and Mission First, and includes marquee brands like Android, Chrome, and Pixel. A central aspect of the overhaul is the consolidation of efforts in AI development, particularly the merger of the Google Brain team and DeepMind researchers into a unified entity named Google DeepMind. This integration is designed to centralize AI model development, facilitating smoother collaboration across teams and innovation.
Furthermore, Google Research is undergoing a strategic refocus, with Google DeepMind spearheading the advancement of advanced AI models, while Google Research concentrates on fundamental areas of computer science. This dual approach seeks to push the boundaries of AI while ensuring responsible development practices, with a keen focus on societal impact.
“Now, to accelerate this progress, we’re going to consolidate the teams that focus on building models across Research and Google DeepMind. All of this work will now sit in Google DeepMind and scale our capacity to deliver capable AI for our users, partners and customers. This will simplify development by concentrating compute-intensive model building in one place and establishing single access points for PAs looking to take these models and build generative AI applications,” Pichai wrote in his letter to the company. “This change also gives Google Research a clear and distinct mandate to continue investing in foundational and applied computer science research in three key areas that tie directly to Google’s mission: computing systems (including quantum), foundational ML and algorithms, and applied science and society.
In parallel, Google is expanding its Responsible AI development by integrating its Responsible AI teams within Google Research into Google DeepMind. This move aims to embed accountability within the core development process, enabling early identification and mitigation of potential issues. Additionally, Google is augmenting its central Trust and Safety team to fortify AI-related safeguards, including standardized launch requirements, enhanced testing protocols, and robust model evaluations.
Another key aspect of the restructuring involves the consolidation of the Devices and Services Platforms and Experiences (DSPA) teams into a new organization called Platforms & Devices. Led by Rick Osterloh, this unified entity aims to leverage AI across hardware and software domains to deliver enhanced products and experiences. Notable additions to this group include Google Research specialists in computational photography and on-device intelligence, further amplifying its capabilities. Accompanying the structural changes are leadership transitions, with Sameer Samat assuming leadership of the Android ecosystem and Hiroshi Lockheimer transitioning to an advisory role while also spearheading Google’s endeavors in Japan.
Google’s comprehensive restructuring is clearly designed to enhance its work while also painting a public picture of rapid AI advancements paired with an ethical foundation. However, it also lets the company keep consolidating amid feature shutdowns and layoffs. In January, Google let go of hundreds of employees from teams, including Google Assistant, as part of a long-term overhaul. It was not a surprise considering the melding of Google Assistant with what was then Bard and is now Gemini. The sunsetting of Google Conversational Actions in favor of Android apps also foreshadowed the reorganization.
Follow @voicebotai Follow @erichschwartz
Generative AI Stars at Google Cloud Next: Here’s The Most Notable News This Year
Google Assistant Snips 17 Features as Shift to Generative AI and Bard Continues
