Google has announced that it is going to carry out a series of structural changes and unify the work divisions dedicated to the development of ‘hardware’ and ‘software’ of the brand into a new team, which will be called Platforms & Devices (Platforms and devices), to advance the development of artificial intelligence (AI).
Currently, Google divides its work teams into different platforms. One of them is focused on ‘software’, Platforms & Services (‘Platforms and ecosystems’), known as P&E. It is followed by ‘Devices and Services PA’ (‘Devices and Services or DSPA’), dedicated to the firm’s ‘hardware’. That is, Pixel, Fitbit and Nest.
The company has decided to unify both areas into a single team, which will be named Platforms & Devices (‘Platforms and devices’), a space in which Google research teams focused on computational photography and intelligence will also have a place.
This was explained by the CEO of Google and Alphabet, Sundar Pichai, who said that with this latest membership, these teams will have the objective of “providing deep experience in AI on all platforms and devices.”
With these new measures, the technology firm seeks to simplify business decisions and streamline its work, so that it is “better and faster,” as stated in a statement published on its blog.
These changes continue the work done over the past year to simplify your business structure and “improve speed and execution” since the launch of Google DeepMind, which brings together the Google Brain team, DeepMind and other researchers focused on creating “increasingly capable and general” AI systems.
With the aim of continuing to offer “the most advanced, safe and responsible AI”, it will bring together in the same team the divisions that currently focus on creating models, Google Research and Google DeepMind.
All of this work will now be located in Google DeepMind, where it will move the responsible AI research teams “to be closer to where the models are built and scaled.” Likewise, it seeks to expand its capacity to offer AI to users, partners and clients, in addition to simplifying its development “by concentrating the creation of computing-intensive models in one place”, as this letter has commented.
As for Google Research, it will continue to invest in fundamental computing research applied to computing systems (including quantum ones), algorithms and foundational machine learning; and applied science and society.
On the other hand, he has commented that he will transfer other responsibility teams to the ‘Trust and Security’ area, where he is investing. “more on AI testing and evaluations.” In addition, it has advanced that it is increasing investments in the so-called red team – which tests the coherence of AI results – “to detect vulnerabilities and carry out broader evaluations” that guarantee accurate responses.
With this restructuring, Google seeks to “work with greater focus and clarity” and strengthen the feedback loop between models, products and users. “Our goal is to organize information and make it universally accessible and useful,” she concluded.