Sundar Pichai, the CEO of Alphabet and Google, has revealed significant organizational adjustments meant to increase execution velocity throughout the digital behemoth. The goal of the reorganization is to accelerate the release of innovative technologies by combining the product and AI research teams.
The consolidation of Google Research’s and DeepMind’s AI model development operations under the DeepMind brand is the centerpiece of the reorganization. In a blog post on Thursday, Pichai declared that this consolidation will allow product teams to easily access the newest large language models and other fundamental AI technologies by centralizing compute-intensive model training into a single organization.
By centralizing compute-intensive model building in one location and creating a single point of access for PAs wishing to use these models to create generative AI apps, Pichai stated that this will streamline development.
Google Research will shift its focus to only foundational and applied computer science achievements in areas like quantum computing, core machine learning, and the societal effects of new technologies, while DeepMind takes on model building.
Additionally, the reorganization attempts to enhance Google’s responsible AI policies. To incorporate safety concerns early in the model building process, DeepMind will welcome back ethical AI teams that were previously integrated into Google Research. To standardize AI evaluation and “red team” testing procedures, more trust and safety teams will relocate to centralized groupings.
For consumers and clients, we must be the finest in the business when it comes to implementing precise, reliable, and transparent AI products. We’re changing how our Responsible AI teams operate at key junctures in the development cycle to assist achieve this, Pichai said.
The creation of a new “Platforms & Devices” product division that unites Google’s device engineering teams with the developers of Android, Chrome OS, Search, Photos, and other developer platforms is another important pillar. The goal is to integrate hardware, software, and machine learning into a single organization in order to expedite AI-driven platform breakthroughs.
This new division will also include experts from Google Research on on-device AI and computational photography. After focusing on other initiatives, longstanding Android chief Hiroshi Lockheimer will now serve as an advisor to Rick Osterloh, who will oversee the whole Platforms & Devices division.
Pichai’s memo states that the main goal is to “simplify our structure and improve velocity and execution” in terms of responsibly creating AI systems.