Highlights:

  • Google launched several services, including its new computer-vision-as-a-service capability, Vertex AI Vision service.
  • Google is also using the Next 2022 event to demonstrate its ability to collaborate with others in the AI and ML field.

Organizations can deploy Artificial Intelligence (AI) workloads in various locations – on-premises or various kinds of cloud infrastructure locations.

When it comes to AI platforms, there is no shortage of cloud-based solutions, and it is evident that AI adoption is also driving cloud expansion. At the Google Cloud Next 2022 event, Google announced that it wants to be the deployment target of choice for corporations’ AI and Machine Learning (ML) workloads.

At the event, Google launched several services, including its new computer-vision-as-a-service capability, Vertex AI Vision service. With the new AI Agents service, Google also intends to make it easier for businesses to develop AI-enabled apps of a certain sort. Following up on Google’s AI-related announcements next is the OpenXLA Project, an open-source initiative that intends to unify machine learning frameworks.

Google giving computer vision an ‘easy’ button

Google introduced Vertex AI in May 2021 as a fully managed cloud AI solution. Over the past year and a half, Google has progressively increased the service’s capabilities to assist enterprises with specific use cases.

Computer vision is one of the most prevalent forms of AI applications, offering image recognition capabilities that can be utilized for various applications. With its new Vertex AI Vision services, the tech giant is providing its users with a managed service to facilitate the development and deployment of computer vision.

Gerrit Kazmaier, VP and GM of the database, data analytics, and Looker at Google, said, “This [Vertex AI Vision] is a fully managed service for computer vision applications that allows you to analyze video streams and pictures and detect objects automatically. Giving all of our customers the ability to build advanced machine learning applications in a very streamlined way.”

Google’s AI Agents also aim to make it simpler for enterprises to enjoy the benefits of AI. Google is developing unique ‘agents’ that are purpose-built services for a particular function powered by Google AI technology. At Google Cloud Next 2022, the business is announcing one such agent with the new Translation Hub service. Kazmaier noted that with the Translation Hub service, customers would be able to take documents and translate them into up to 135 languages.

Kazmaier said, “Translation Hub builds on this idea of using AI and machine learning to make knowledge more accessible and widespread. Think about a company [and] the ability you gain from this to understand and serve markets in different languages, so we are excited to make the translation more accessible to the world with the Translation Hub.”

Open-source machine learning moves forward with OpenXLA

Google is also using the Next 2022 event to demonstrate its ability to collaborate with others in the AI and ML field.

Sachin Gupta, VP, and GM, of infrastructure at Google Cloud, said, “Google is collaborating with several open-source AI frameworks with industry experts, including AMD, Arm, Intel, Meta, Nvidia, and others as part of our shared commitment to unite the AI ecosystem, helping customers avoid platform or model locking.”

The collaboration is taking place through the new OpenXLA initiative, which, according to Gupta, aims to lower the cost and complexity of ML implementation. OpenXLA will aid by solving the fragmentation issue that now plagues the ML infrastructure. OpenXLA, according to Gupta, is creating a modular compiler and infrastructure tools to expedite the development of artificial intelligence with the support of an open community of contributors.

“Our long-term commitment to open ecosystems is grounded in the belief that no one company should own AI and ML innovation,” Gupta said.