This week in Las Vegas, a crowd of 30,000 gathered to discover the latest innovations from Google Cloud. The focus was overwhelmingly on generative AI. While Google Cloud primarily serves as a cloud infrastructure and platform vendor, you might not have realized this given the flood of AI-related announcements.
Although Google's presentations were impressive, they followed a similar pattern to Salesforce's event last year in New York, where the company's main business received only minimal attention—except, of course, when discussed in the context of generative AI.
Google introduced several AI enhancements aimed at helping users leverage the Gemini large language model to boost productivity on their platform. These were highlighted with numerous demos during the main and developer keynotes, showcasing the capabilities of these new tools.
However, many of these demos appeared overly simplistic, possibly due to time constraints. They primarily featured applications within Google's own ecosystem, ignoring the fact that most companies store their data outside Google. In one e-commerce demo, the presenter had to call the vendor to complete a transaction online, a step that seemed unnecessary with AI since it could have been easily handled directly on the website.
Despite these oversights, generative AI does offer significant benefits, such as code generation, content analysis, and data querying for troubleshooting. Google also introduced task and role-based agents to assist developers, creatives, and other employees, showing tangible applications of generative AI.
Creating AI tools using Google’s models, however, presents challenges that weren't fully acknowledged. These challenges are substantial, especially when implementing advanced technology within large organizations. Over the past 15 years, technological advancements like mobile tech, cloud services, and marketing automation have promised much but also brought considerable complexity. Often, large companies proceed with caution, and AI adoption is proving to be no exception.
Some organizations struggle with technological adoption due to internal resistance, outdated tech infrastructures, or internal politics that hinder new initiatives. Vineet Jain, CEO of Egnyte, noted the difference between companies that have embraced cloud technology, who may find adopting generative AI easier, and those that are lagging behind, for whom AI adoption will be much more challenging.
The successful implementation of sophisticated technologies like generative AI often starts with having well-managed data. Companies lacking clean, organized data will find it particularly tough to benefit from generative AI. Kashif Rahamatullah from Deloitte emphasized that discussions about AI need to quickly shift towards improving data management for effective implementation.
Google aims to assist data engineers in creating pipelines that facilitate data connectivity and preparation for AI models. Yet, for companies not yet advanced in their digital transformation, these efforts might still fall short, presenting significant challenges in data management and integration.
Aside from the technical challenges, implementing AI involves navigating issues related to governance, liability, security, and ethical usage, as pointed out by analyst Andy Thurai from Constellation Research. These aspects are far from trivial and require careful consideration.
In conclusion, those who attended the Google Cloud Next event in Las Vegas this week, expecting general updates from Google Cloud, might have been surprised by the intense focus on AI. For many, particularly those from organizations with less digital maturity, fully leveraging these new technologies might still be a distant reality.