Why Google Cloud Bests AWS, Microsoft In Generative AI: Kevin Ichhpurani
“We’ve all seen the situations that have happened with some of our competitors, where data gets into the public domain. We have a complete firewall of your proprietary corpus of business data—that’s a really big thing for enterprises,” Google Cloud’s Kevin Ichhpurani, corporate vice president, Global Ecosystem and Channels, tells CRN.
Google Cloud has some serious advantages in the red-hot generative AI enterprise market compared to rivals Microsoft and Amazon Web Services, says Google’s head of global ecosystems and business development.
“We’re really addressing the needs of enterprise clients, which is completely different than the requirements on the customer side,” said Kevin Ichhpurani, corporate vice president, Global Ecosystem and Channels at Google Cloud in an interview with CRN.
“We’ve all seen the situations that have happened with some of our competitors, where data gets into the public domain,” Ichhpurani said. “We have a complete firewall of your proprietary corpus of business data—that’s a really big thing for enterprises. … Organizations don’t want to lose their secret sauce.”
Google’s Generative AI Push
The Mountain View, Calif.-based company has been driving deeper than ever before into generative AI and artificial intelligence in general this year.
At Google I/O 2023 last week, the company unveiled several new generative AI solutions for customers and partners. This includes Duet AI for Google Cloud, a new AI-powered collaborator to help cloud users of all skill levels solve everyday work challenges. Duet AI serves as an expert pair programmer and assists cloud users with contextual code completion, offering suggestions tuned to a customer’s code base, generating entire functions in real time, and assisting with code reviews and inspections. Additionally, the new Duet AI for Google Workspace brings together all of Google’s powerful generative AI features to its Workspace suite of products—which includes Gmail, Docs, Slides, Sheets and Meet—and lets users collaborate with AI so they can get more done.
Google also launched a new version of its large language model PaLM. “PaLM 2 models are stronger in logic and reasoning thanks to broad training on scientific and mathematical topics,” said Google’s CEO Sundar Pichai during the event.
There’s a generative AI arms race right now. What is Google’s market differentiation versus AWS and Microsoft when it comes to gen AI?
On the Google Cloud side, one of the things that’s unique is we’re really addressing the needs of enterprise clients, which is completely different than the requirements on the consumer side.
So our Google Bard is on the consumer side. We have Vertex AI on the enterprise side. It’s a very different set of issues. Like you might use Bard similar the way you would use ChatGPT to plan a vacation. The enterprise requirements are completely different.
What we have addressed for enterprise clients, which is so alluring—and why many of the largest systems integrators and advisory partners are making very material investments in us—is really addressing those enterprise requirements.
I’ll give you some examples.
Enterprises care about, ‘How do I take my proprietary corpus of data and adapt the model?’ Now, when Google Cloud does that, we keep your data completely private—it is stored separately in your own tenant. And we don’t commingle that with our large language model.
We’ve all seen the situations that have happened with some of our competitors, where data gets into the public domain. We have a complete firewall of your proprietary corpus of business data—that’s a really big thing for enterprises.
Organizations don’t want to lose their secret sauce. That’s why the data privacy aspect—your data is compartmentalized and kept separately and is firewalled off—is so critical.
What are some other generative AI market differentiators for Google Cloud versus AWS and Microsoft?
Another differentiator is enterprises are very focused on accuracy of the data. They cannot afford hallucination in data [divergences from the source content]. So we’re very focused in on the efficacy of the model for the enterprise, and the accuracy of the results.
Number three, they care about price performance.
You obviously can’t have a model that is cost prohibitive, where the value that you’re receiving is less than your costs. So the price performance becomes very important. We have a lot of science going into ‘How do we create the right special purpose model for clients that has the right price performance?’
Another point is, just looking at the broad trust and safety issue, ensuring that there isn’t toxicity, harmful content, or harmful to specific user communities.
These are the kind of things that enterprises are very focused on that we’re addressing. What’s made us very unique, is that we’re really going after the cloud side of the enterprise use cases and those business issues that are top of mind for enterprise clients.
What are some important and emerging generative AI use cases Google Cloud is attacking in the market?
Pharmaceutical companies can significantly reduce the amount of time it takes to do clinical trials: all the documentation to run the clinical trials and get drugs to market faster, which has an immense ROI for them.
We have financial institutions that are talking to us about creating personal financial advisors. With media and entertainment, the ability to have deep personalization.
We have retailers creating completely new experiences.
We have a number of companies where they can automatically generate marketing content in a highly personalized manner.
You’ve gotten the generic emails that look like spam in the past, right? Imagine now you could tap into your CRM system and send 300,000 emails that are all custom tailored to every single individual. For example it could say, ‘Dear Mark, in your last visit into the store, you looked at the following products. I wanted you to be aware now that these new products are available.’ This is the art of the possible.
You kind of pick an industry and there’s tremendous opportunity. You could almost choose any industry, and there’s many, many use cases around, ‘How do I optimize support? How do I improve my customer service? In HR, how do I better engage with my employees?’
These are all different areas that we’re seeing significant potential on right now.
What are your bullish thoughts on Google Cloud’s generative AI strategy around driving growth for channel partners?
The ROI is immense. That’s why generative AI is not a hyper-fad.
The ROI that is achieved through efficiencies, the ability to reach new customers and grow your top line is so immense that this will be something that is sustainable.
The clients need help with ideation like, ‘Where do I start?’ Partners can quickly get from one use case to like 200 use cases.
We have a lot of interest from the partner community, from global system integrators and advisory firms, to build practices and to train up large numbers of individuals to have deep competency on Google Cloud, but to also help the clients.
Organizations are working on their strategy of how can they create completely new customer experiences with generative AI? How can they create new employee experiences? New experiences for their partners and develop entirely new business models for the company.
We’re still in the very early innings. There’s a lot more innovation to come from Google Cloud.