Google's impressive portfolio of cutting-edge technologies, including its latest Gemini model, has garnered significant attention in the tech world. However, a closer examination reveals that these innovations may not be directly applicable to mainstream enterprises, despite their impressive capabilities.

Google Cloud's revenue growth, albeit on a smaller base, is a testament to the company's strength in AI. Its newest Gemini model meets or exceeds OpenAI's GPT-4 in areas like complex reasoning, showcasing Google's prowess in tackling bafflingly hard computer science problems. The company has a history of releasing research papers, open-source projects, or cloud services that are now commonly used, such as Kubernetes, Angular, Bazel, and Zanzibar.

However, the key takeaway is that you shouldn't try to be Google. The tech giant's unique scale, homogeneous technology stack, and enormous monorepo set it apart from most enterprises. Google's approaches, such as relying on monorepos, site reliability engineers, and other techniques, may not be applicable to mainstream enterprises due to differences in scale and architecture.

A prime example of this is Google's Zanzibar, a globally consistent authorization system. While authorization is a near-universal need for enterprises, Zanzibar-style "authz" systems require centralization of all data, which may not be feasible for companies operating in distributed microservices architectures. This has led to startups attempting to deliver on Google's vision, but ultimately struggling to adapt it to mainstream enterprises.

In contrast, companies like Oso have found success by offering hybrid architectures that cater to the messy reality of enterprise tech. Oso's approach centralizes shared authorization data in its cloud service while leaving service-specific data in general-purpose application databases. This pragmatic approach has enabled enterprises to tackle hard problems like authorization without introducing the harder problem of mimicking Google.

The lesson to be learned is that adopting technology from Google's latest research paper or open-source project won't necessarily make you just as cool. Instead, the best strategy is to learn from Google's design thinking and apply it in a way that suits your organization's unique needs and architecture. By doing so, enterprises can leverage the best of Google's innovations without straining to fit their IT practices into Google's mold.

In conclusion, while Google's cutting-edge technologies are undoubtedly impressive, mainstream enterprises should be cautious about adopting them wholesale. By recognizing the differences in scale and architecture, companies can instead focus on applying Google's design thinking in a way that works for them, rather than trying to be something they're not.