Generative AI

Google consolidates AI teams into DeepMind to scale capacity – Computerworld


Aimed at accelerating progress in AI development and responsible AI deployment, Alphabet-owned Google is consolidating its teams that are responsible for building AI models across Google Research and Google DeepMind, its CEO Sundar Pichai said Thursday in a note to its employees.  All AI “work will now sit in Google DeepMind,” Pichai specified in the note.

This restructuring will “scale our capacity to deliver capable AI for our users, partners and customers,” Pichai said in the note. “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.”

The Google DeepMind team will be led by Demis Hassabis, said the note.

Google formed Google DeepMind exactly a year ago by combining two research teams in the AI field – the Brain Team of Google Research, and DeepMind. This focused team, backed by the computational prowess of Google “will significantly accelerate our progress in AI,” Pichai had said in an April 2023 note. Gemini models were created by Google DeepMind.

Google Research is the research arm of Google, dedicated to AI and computer science to develop next-generation technologies that benefit Google products having key focus areas including AI/ML, Responsible Human-centric Technology, Science & Societal Impact, Computing Paradigms, and Algorithms and Optimization.

What’s new, now?

Google is now consolidating all its AI units one to “simplify our structure and improve velocity and execution – such as bringing together the Brain team in Google Research with teams in DeepMind, which helped accelerate our Gemini models; unifying our ML infrastructure and ML developer teams to enable faster decisions, smarter compute allocation, and a better customer experience; and bringing our Search teams under one leader,” the note added.

The move, Pichai said, also gives Google Research a clear mandate to continue investing in three key areas that align with Google’s mission — computing systems, foundational ML and algorithms, and applied science and society.

“Consolidating all of Google’s AI teams, including Google Research and DeepMind, into one unit under Google DeepMind likely reflects a strategic move aimed at streamlining and optimizing AI development and deployment across the company,” said Pradeepta Mishra, an AI expert and co-founder of cybersecurity firm Data Safeguard.

Besides, Google is reaffirming its commitment to responsible AI deployment by ramping up its Responsible AI Team’s roles and accountability. Teams focusing on Responsible AI within the Google Research team will now move to Google DeepMind to be closer to ‘where the models are built and scaled’, said the note.

Similarly, “other responsibility teams” are moving into our central “Trust and Safety” team where the company is investing more in “AI testing and evaluations” to enhance product accuracy and responsiveness. “These shifts create clearer responsibility and accountability at every level as we build and deploy, and strengthen the feedback loop between models, products, and users,” Pichai added in the note.

Recognizing the potential of AI, Google is also formalizing collaboration between its AI divisions, software, and computing platforms. “So we are formalizing the collaboration between DSPA and P&E and bringing the teams together in a new PA called Platforms & Devices.”

Having a unified team across Platforms & Devices will help Google deliver higher-quality products and experiences for its users and partners, Pichai said. It will help us turbocharge the Android and Chrome ecosystems, and bring the best innovations to partners faster — as we did with Circle to Search with Samsung. And internally, it will also speed up decision-making.”

How it helps Google

Merging teams eliminates redundancy and fosters closer collaboration between researchers and developers. This could accelerate the development cycle for new AI products and features. All these moves, Pichai said in the note, “will help us work with greater focus and clarity towards our mission.”

“With one central unit, decision-making around resource allocation and project priorities becomes more efficient,” Mishra added. The AI landscape is fiercely competitive. Data Safeguard’s Mishra said this consolidation could help Google “stay ahead of the curve by accelerating innovation.”

“By integrating research and development under one roof, Google might create more unified and impactful AI products across its platforms (Search, Assistant, etc.). Streamlined operations could lead to better resource utilization and potentially cost savings,” said Mishra.



Source

Related Articles

Back to top button