The Cloud Was Never About Technology. It Was About Control.

Most people think cloud computing started when Amazon launched AWS in 2006. They are partially right and completely wrong at the same time.

The real origin goes back further. In 1961, MIT computer scientist John McCarthy stood at a convocation and said something that sounded radical: computing would someday be organized as a public utility, like electricity. You plug in, you consume, you pay. Nobody owns the power plant. You just use power.

For four decades, that idea sat quietly. Enterprises bought their own servers, built their own data centers, hired armies of people to maintain iron they owned. Then the internet exploded in the late 1990s, and suddenly companies like Amazon were running massive internal infrastructure just to keep their websites alive. Amazon’s engineers noticed something inconvenient: most of their hardware sat idle most of the time. Peaks were seasonal. The rest was waste.

In 2002, Amazon began quietly building a standardized internal platform. By 2006, they did something no one expected from a bookstore. They opened that platform to the public. Amazon Web Services launched with two offerings: S3 for storage and EC2 for compute. The idea was staggeringly simple. Why should every company build a data center when they could just rent Amazon’s?

The cloud had a commercial birth date.

The Gold Rush That Followed

Microsoft watched and panicked slightly. Google watched and moved fast. Both launched their cloud platforms around 2008-2010. Microsoft Azure went public in 2010. Google Cloud Platform became a serious commercial offering around the same period.

What followed was not gradual growth. It was a supernova.

By 2015, enterprises that once laughed at putting their data “on someone else’s computer” were signing multi-year contracts with cloud vendors. The cost economics were undeniable. No capital expenditure. No hardware refresh cycles. No three-year depreciation headaches. Infinite scale on demand.

Today, global cloud infrastructure is dominated by five serious players: Amazon Web Services, Microsoft Azure, Google Cloud, Alibaba Cloud, and Oracle Cloud. A second tier includes IBM Cloud, Salesforce infrastructure, Tencent Cloud, and a handful of regional players across Europe and Southeast Asia.

The market share picture as of recent estimates is roughly this: AWS holds approximately 31-33% of global cloud infrastructure spend. Azure sits at 22-24%. Google Cloud is at 11-12%. Alibaba Cloud leads Asia-Pacific at around 4-5% globally. Everyone else divides the remainder.

That is not a level playing field. That is an oligopoly with AWS and Azure as a duopoly at the top.

The Azure Lock-In Question: Myth or Real Risk?

Here is a question that deserves an honest answer rather than a comfortable one.

Large enterprises already running Microsoft Active Directory, Office 365, Teams, SQL Server, SharePoint, and Windows Server environments do face a gravitational pull toward Azure. It is not conspiracy. It is product architecture. Microsoft has deliberately built integration pathways that make Azure feel like the natural extension of infrastructure you already paid for.

When your identity lives in Entra ID, when your compliance lives in Microsoft Purview, when your SIEM is Sentinel, moving your workloads to Azure is genuinely easier than moving them to AWS or GCP. Your security teams already know the tools. Your licensing agreements already cover some services.

But calling this inevitable is an overstatement. Multi-cloud is increasingly real, not just a buzzword. Major enterprises routinely run Azure for Office 365 and identity, AWS for specific application workloads where pricing or service maturity is stronger, and GCP for data analytics and machine learning pipelines. Vendor lock-in is a risk that serious architecture teams actively design against.

The honest answer: Microsoft has structural advantages with enterprises that already run their stack. But enterprises that are serious about negotiating power and resilience are deliberately diversifying. The gravitational pull is real. Surrender to it is a choice, not a destiny.

Where the Jobs Actually Are

Cloud has stopped being a niche and become the floor. If you are entering technology today without cloud literacy, you are entering a building and ignoring the foundation.

For freshers, the entry points are clear. Cloud support and operations roles require Azure, AWS, or GCP fundamentals. Get certified first: AZ-900 for Azure, AWS Cloud Practitioner, or Google Associate Cloud Engineer. These are not prestigious certifications. They are a threshold you must cross to be taken seriously in screening rounds. Once in, roles like cloud support engineer, infrastructure analyst, junior DevOps, and cloud operations analyst are accessible.

For experienced professionals coming from on-premise backgrounds, the translation is direct. If you understand networking, storage, or backup on-premise, those skills map cleanly to cloud equivalents. The learning curve is steep in the first six months and then it flattens. RBAC, identity, governance, and compliance experience travels well across cloud environments.

The areas where recruitment will grow hardest over the next three to five years are not basic cloud support. They are cloud security, specifically identity and access management, zero-trust architecture, and cloud SIEM operations. FinOps, meaning cloud cost management and optimization, is becoming a standalone function in large enterprises. Multi-cloud management platforms require engineers who can navigate more than one vendor simultaneously. AI and ML infrastructure on cloud platforms, managing GPU clusters, data pipelines, model deployment infrastructure, is the frontier.

Platform engineering, which means building internal developer platforms on top of cloud infrastructure, is also growing. Companies want their developers to stop thinking about infrastructure entirely. Someone has to build that abstraction layer.

The McCarthy Prophecy Completed

John McCarthy’s 1961 vision has essentially come true. Computing today is closer to a utility than at any point in history. You spin up a server in under three minutes. You shut it down when you are done. You pay for what you consumed.

What McCarthy did not predict was the concentration of that utility in three or four private hands. The cloud democratized infrastructure for users. It consolidated power for vendors.

That tension, between the openness cloud promises and the lock-in it quietly creates, is the defining conversation of enterprise technology for the decade ahead.

The engineers and architects who understand both sides of that equation are the ones who will be impossible to replace.

Nishani writes on technology, geopolitics, and Indian enterprise at nishani.in

Comments

comments