Google dives deep into generative AI during Google Cloud Next.

In Las Vegas this week, an impressive gathering of 30,000 individuals convened to receive the latest updates from Google Cloud. The focus of the event was unequivocally on generative AI, showcasing Google Cloud’s primary role as a leading cloud infrastructure and platform provider amidst the flurry of AI-related announcements.

While Google presented a multitude of AI advancements aimed at enhancing customer utilization of the Gemini large language model (LLM) and augmenting productivity across the platform, the core business received only cursory acknowledgment, mostly in relation to generative AI.

Throughout the primary keynote on Day 1 and the subsequent developer Keynote, Google complemented its announcements with numerous demonstrations highlighting the capabilities of these solutions. However, some of these demonstrations appeared overly simplistic, often confined to examples within the Google ecosystem, despite the reality that many organizations store their data in repositories external to Google.

Despite the potential of generative AI in various scenarios such as code creation, content analysis, and log data interrogation, there was a notable oversight in addressing the complexities that may impede successful implementation within large organizations. While Google aimed to convey simplicity in utilizing AI tools, the reality of integrating advanced technology into complex organizational structures poses significant challenges.

Big changes are rarely straightforward.

Similar to past technological advancements, such as mobile, cloud, containerization, and marketing automation, each promising substantial benefits, they also bring along their own complexities. However, adoption within large enterprises tends to proceed cautiously, contrary to widespread expectations. Artificial Intelligence (AI) appears to pose a more substantial challenge than portrayed by major vendors like Google.

Experience from previous technological shifts reveals a pattern of high expectations followed by widespread disappointment. Despite years of availability, many large corporations have only tiptoed into these advanced technologies or opted out entirely.

Numerous factors contribute to companies failing to capitalize on technological innovation. These include entrenched organizational inertia, outdated technology infrastructures hindering the adoption of newer solutions, and internal resistance from various corporate factions. Legal, HR, IT, or other groups may oppose well-meaning initiatives due to internal politics or other reasons.

Vineet Jain, CEO of Egnyte, a company specializing in storage, governance, and security, identifies two categories of companies: those successfully transitioning to the cloud, poised for easier integration of generative AI, and those slower to adapt, likely facing challenges in embracing such technologies.

In discussions with numerous companies that predominantly rely on on-premises technology, there emerges a realization of the considerable distance they must traverse before contemplating the potential benefits of AI integration. According to Jain, who conveyed insights to TechCrunch, a significant portion of these entities can be categorized as “late” adopters of cloud technology, either yet to embark on their digital transformation journey or just in its nascent stages.

The advent of AI prompts these organizations to confront the imperative of embracing digital transformation, yet grappling with the challenges posed by their current technological lag.

In essence, while AI presents opportunities for advancement, companies at the initial stages of their digital evolution must first address fundamental infrastructural concerns before effectively leveraging AI capabilities.

The focus is always on the data.

While big players like Google make the implementation of such solutions seem straightforward, the apparent simplicity upfront doesn’t necessarily reflect the complexity on the backend. As reiterated frequently this week, when it boils down to the data utilized for training Gemini and similar large language models, it still boils down to the old adage of “garbage in, garbage out,” especially when dealing with generative AI.

It all begins with data.

If your data isn’t organized, shaping it for training Language Models (LLMs) for your specific use case becomes incredibly challenging. Kashif Rahamatullah, a Deloitte principal overseeing the Google Cloud practice at his firm, found Google’s announcements this week impressive overall, but admitted that companies grappling with messy data will encounter hurdles in implementing generative AI solutions.

From Google’s standpoint, the company has developed generative AI tools to assist data engineers in constructing data pipelines linking to various data sources within and outside the Google ecosystem. Gerrit Kazmaier, Google’s Vice President and General Manager for database, data analytics, and Looker, told TechCrunch, “The aim is to accelerate data engineering teams by automating many of the labor-intensive tasks involved in data movement and preparation for these models.”

This should facilitate data connection and cleansing, particularly in companies further along in their digital transformation journey. However, for firms like those mentioned by Jain—those yet to embark significantly on digital transformation—it could pose greater challenges, even with Google’s tools.

Moreover, the challenges associated with AI extend beyond mere implementation, whether it’s deploying an app based on an existing model or constructing a custom model, according to Andy Thurai, an analyst at Constellation Research. “While deploying either solution, companies must consider governance, liability, security, privacy, ethical and responsible use, and compliance,” Thurai stressed. And none of these aspects are trivial.

Attendees of the recent GCN event, including executives, IT professionals, and developers, may have sought insights into Google Cloud’s upcoming offerings. However, those not actively pursuing AI or organizations unprepared for it may have been taken aback by Google’s exclusive focus on AI at the event. For organizations lacking digital maturity, fully leveraging such technologies, beyond pre-packaged solutions offered by Google and other vendors, could be a lengthy process.