I started my technology journalism career in the early 1980s, when IBM’s market dominance was so complete that each year it sopped up half of the industry’s revenues and 80 percent of its profits. I also covered Microsoft during its ascent into desktop dominance. These two amazing companies achieved almost unbelievable power in very competitive markets because they owned the platform.

With the arrival of the open-standards-based Internet, I thought that the days of computer industry monopoly might be over, but after watching the sheer volume of news emanating from Amazon’s re:Invent conference this week, I’m no longer so sure. The volume and scope of announcements was breathtaking, with new services rolling out in database, storage, business intelligence, data transfer, security and even Internet of things. AWS topped things off with a joint venture with Accenture, the world’s largest consulting firm, to offer migration services for enterprises.

Amazon is so far ahead in public infrastructure-as-a-service that it’s almost scary, particularly considering how much growth the market has left in it. One analyst firm estimated that Amazon’s processing infrastructure is 10 times as large as that of its next 10 largest rivals combined. In an already price-competitive market, it continues to drive down costs by leveraging the scalability model of software while still earning a tidy profit on a $7 billion business that is estimated to be growing at 50 percent annually. Speaking on theCUBE Amplify Partners’ General Partner Mike Dauber said AWS has practically become the new OS.

Sure looks like a budding monopoly to me.

Amazon would argue (correctly) that it has plenty of competition. So did Microsoft and IBM at one point, but over time the competitors became merely spoilers. Computing platforms tend to spawn monopolies and duopolies because customers don’t like an abundance of choice. Dominant platforms quickly become “safe,” and then customers come running. Once they have a critical mass of customers on board, monopoly companies drive up the cost of switching to consolidate their control.

Amazon is in the process of doing this on several levels. It has been extremely active in the database market over the past year, launching the Aurora homegrown relational database, adding support for MariaDB and adding QuickSight to its analytic offerings. It’s reported to be working on a super-fast analytical database as well.

Its multi-tier storage options now cover everything from flash-based performance storage to the archival Glacier offering. In a stroke of brilliance this week, the company added a service that uses a physical appliance for high-volume data transfers.

The pattern is clear: make it as cheap and easy as possible for customers to move large amounts of data to the Amazon platform. Once their data is safely tucked away in Amazon databases, the cost of switching becomes prohibitively high. Once Amazon has the hearts and minds of enterprise customers, the applications will follow.

Wikibon co-founder David Vellante notes that because Amazon supports multiple operating systems, it needs to lock in customers at the database level. That explains a lot of the company’s recent hyper-activity there.

You can expect Amazon to continue to build out infrastructure services such as security, directory and network management that cover all the checkpoints of an enterprise on-premise data center. Whether the company gets into applications in a big way is an open question. My guess is it will stick to its core market – where there is plenty of growth for many years – and court developers to join its Marketplace. That latter strategy would be straight out of the playbook Microsoft used to build such a formidable ISV blockade in the 1990s.

Monopolies aren’t illegal, and there’s no reason to believe that Amazon’s hegemony will limit customer choice or drive up costs. In fact, it’s in Amazon’s best interest to continually lower its prices. Microsoft was able to blunt the Justice Department’s case in the late 1990s by pointing out that PC price performance continued to improve despite its dominance.

Customers should be cautious, though. Once they’re in the Amazon cloud, the vendor has no incentive to make it easy for them to go elsewhere. Amazon has notably declined to support open platforms like Cloud Foundry or OpenStack, because doing so would be against its best interest. It will continue to build out its proprietary platform-as-a-service offering, aggressively introduce new infrastructure capabilities and press its advantage as hard as it can.

Those with a long memory will recall the 1991 breakout moment when Bill Gates dissolved Microsoft’s partnership with IBM and went its own way with Windows. Amazon stands at a similar crossroads. Its momentum is so great right now that it can almost write its own ticket. I, for one, will watch in awe, tinged just a bit by fear.

Photo by Unsplash via Pixabay