At Google I/O, San Francisco, besides several updates, emerged a new drinking game — take a sip every time the speakers utter the word “responsible AI”. But hold on to your glasses, because the first-ever Google I/O Connect in Bengaluru was a carbon copy of the US event, albeit with an Indian twist. Playing the perfectly responsible parents of the AI world, the Bard creator made sure whatever they announced had a responsible AI approach. Here, they focused on updates tailored for India, but surprisingly, there was barely any announcement on Search.
Last year at the ‘Google for India’ event, the big tech announced a grant of $1 million to IIT Madras for setting up a centre for Responsible AI. However, at the same time, Google also cautioned its employees against its own chatbot Bard raising concerns about the privacy and security of these LLMs. Google was earlier promoting the use of Bard for its employees. However, its updated privacy note also asks users to not include confidential and sensitive information in Bard conversations.
In light of these recent developments, Google’s commitment to responsible AI becomes even more crucial, as it aims to ensure that AI systems produce reliable and trustworthy results while mitigating potential biases and inaccuracies in the AI models.
Will Grannis, VP and CTO of Google Cloud mentioned during his keynote session that all their “AI-generated content is labelled with metadata” to ensure that they employ AI responsibly. And not just Grannis, other speakers like Ambarish Khegne (vice president, product, Google Pay), Matthew McCollough (VP, product, Android) and Una Kravets (developer relations engineer) made sure to emphasise the ethical use of AI.
Indian IT & Google – A Love Story
Interestingly, in the past couple of months, several Indian IT giants like TCS and Wipro have partnered with Google Cloud to bolster their generative AI capabilities, giving up on competitors like Microsoft Azure, AWS, among others. Will Grannis, CTO of Google told AIM, “Our local engineering community in India is the second largest in the world, so it’s where we need and want to be, and we’re really thrilled about tapping into that developer community. This is why we’ve been offering starter credits to ensure financial barriers don’t prevent them from accessing our capabilities.”
Another significant aspect is the ever-growing startup ecosystem in India developing innovative solutions by leveraging the platform. Talking about the differing priorities of startups and large governmental bodies, startups favour speed and agility while larger entities prioritise accuracy and invest more time in ensuring it. Google acknowledges the diverse needs and priorities of its users and strives to develop AI technologies that cater to these different contexts, allowing for customisation and adaptability.
“As more people come, we’re working to simplify network solutions. This is part of what we believe is critical and will shape the future,” he added.
India is home to more than 60 generative AI startups, as per a report. With a focus on implementing generative AI, particularly with Vertex AI, for Indian enterprises, Google is offering AI capabilities through consumer-facing products and within the Cloud platform, empowering developers with new tools and capabilities. The introduction of PaLM API, MakerSuite, and features on Vertex AI are important steps in this direction. This development isn’t just about the models themselves but about providing a comprehensive platform that enables production-ready solutions at scale, marking a natural progression for Indian developers.
“There is a growing interest in gen AI, with numerous startups and enterprise customers exploring AI for innovation and improved customer experience,” Manish Gupta, head of Google Research India told AIM at the event.
Regarding Project Vaani, which Google announced at its ‘Google for India’ event, in association with IISC, the initial phase of data collection is now made available to developers as an open-source resource for speech recognition systems in multiple languages. “We are also launching the Open Buildings dataset, which includes satellite analysis of over 200 million buildings in India,” Gupta told during the discussion.
As per Gupta, among the 125 languages they had pledged to support, 75 lacked any data corpus. “In the 4,000 hours of speech data that has been put out, for a few languages, it’s the first such instance that digital data has been made available,” he added, opening an arena of new advancements and innovations in these languages with zero data corpus.
Reimagining the SEO Experience
As AIM reported earlier, Google is slowly inching towards realising cofounder Larry Page’s vision of an “AI-complete” search engine by incorporating generative AI. So now, search engines will resemble conversations with AI, making conventional SEO techniques redundant. “Harnessing the power of generative AI can enhance human creativity and advertisers will be able to automatically generate diverse content such as images, short videos, and compelling descriptions, enabling them to personalise their ads according to individual users’ preferences,” Gupta told AIM.
While generative AI has the potential to misrepresent facts, Google is dedicated to providing trustworthy information and will exercise caution in leveraging this technology to ensure the preservation of user trust.
The post Responsible AI Takes Center Stage at Google I/O Connect appeared first on Analytics India Magazine.
