Corporations are historically overly conservative and cautious, so it is with AI. Amid all the chatter and excitement about zillions of use cases, most experimenters at the enterprise level are still leveraging AI for internal document analysis and communication. Internal only, completely de-risked,

At Google’s Cloud Conference, visionary AI applications were discussed alongside their more practical business implementations. Companies are currently leveraging AI for routine tasks such as summarizing communications or extracting information from internal documents.

During his keynote at the annual event in Las Vegas, attended by 30,000 people, Google Cloud CEO Thomas Kurian painted a future where AI agents automate diverse business tasks, from data analysis to creating advertising content and enhancing consumer spending.

However, transitioning customers to these advanced applications is proving challenging. Many are still in the experimental phase with AI, using it for basic tasks as they navigate its limitations and high costs, according to empirical data and discussions with executives at the event.

Key Insights from the Conference:

• Many Google Cloud customers are still exploring generative AI’s possibilities.
• Discover reports a 70% reduction in response time to customer inquiries using AI tools.
• Skepticism remains about AI’s readiness for more complex applications.

Vineet Jain, CEO of Egnyte, is experimenting with Google’s AI to potentially enhance his company’s file storage solutions. However, the accuracy issues and “hallucinations” of current AI models make it difficult to justify charging customers for these new features.

Good Use Case

Jain personally uses an AI summarization tool to prepare for meetings. He appreciates the concise summaries it generates, acknowledging the technology’s potential despite the prevailing hype.

Google executives recognize the cautious approach their customers are taking. Oliver Parker, a vice president at Google Cloud, noted that the industry remains wary, particularly in scenarios involving direct customer interaction.

Some companies are making progress, integrating AI to streamline internal processes. For instance, Discover Financial Services has deployed Google AI tools among its customer support staff, significantly reducing response times. This integration is expected to expand, even internally though a cautious and slow rollout, with ongoing testing to ensure the accuracy and appropriateness of automated responses.

Conference Highlights and Industry Perspective: The conference showcased numerous applications of AI across business sectors with attendees discussing potential transformative impacts. Despite the excitement, industry leaders are moderating expectations about AI’s current capabilities and cost-effectiveness.

For example, a demonstration featuring Gemini, Google’s flagship AI model, humorously showcased its ability to recommend online shopping choices, highlighting both the potential and the frivolous applications of the technology. Gemini should stick to its knitting in an X variant market like this, with no predictability that will help pave the way for a unique differentiation.

In the healthcare sector, companies like Bayer are exploring AI for diagnostic support, though product deployment remains years away. Bayer currently uses AI to assist with internal document analysis and communication.

Arvind Jain, CEO of Glean, expressed caution regarding the premature reliance on AI for automating business processes without significant human oversight.

Concerns over AI Costs: The high cost of implementing AI is a significant barrier for many companies. Laurent Gil, chief product officer at Cast AI, pointed out the financial challenge in justifying AI investments, which often require expensive hardware and extensive data processing. Factoring in cooling and adjacent supportive infrastructure, the costs may go quite higher by next year.

Since AI will be a cloud native service and require hyperscale, and since we only have 3 hyperscalers on the planet, we additionally have set up a tempting and vulnerable target for nation state bad guys to consider. A single systemic, catastrophic cloud failure would make a sci-fi movie by itself.

Top of Form

Bottom of Form

Google Cloud sees a future in which businesses use artificial intelligence agents to automate all manner of tasks, from analyzing data, to creating ad copy, to prodding consumers to spend more money.

Now he must get customers to follow his lead—and that’s proving to be a challenge. Many are still just experimenting with AI. Those deploying it more widely are using it for relatively mundane tasks like summarizing correspondence or finding information buried in internal documents, executives told The Information during the 30,000-person event in Las Vegas. Increasingly aware of AI’s limitations—and high cost—many customers are moving cautiously.

The Takeaway

• Many Google Cloud customers say they are still experimenting with generative AI
• Discover says AI tools cut time to respond to customer queries by 70%
• Look, this is not ready for prime time,’ one customer said

Vineet Jain, CEO of Egnyte, a Silicon Valley startup that sells file storage and sharing software to midsize businesses, is looking to use AI models from companies including Google to enhance products for his own customers. But many clients aren’t ready to adopt—or pay for—new AI-fueled features.