We are all accustomed to rapid evolution of the high-tech space, and I'm one of many journalists who have made a career out of talking about it. Computers were an important accessory when I first started doing this, then they became a necessary accessory, and then, for many industries, the server room became the centerpiece for everything – everything else was an accessory as long as the servers kept churning.

That was a while ago. Over the past several years, corporations have been dusting out those old server rooms, committing the space to other purposes, and sending their data to the cloud. One important factor in the rapid growth of the cloud has been cost savings. Save the space; save the staff; slash the electric bill – let the cloud handle everything. Vigorous competition in the cloud industry, including among the so-called "big three" cloud vendors (Google, Amazon, and Microsoft Azure) kept prices low – and often even pointing downward.

But early signs indicate that the gravy days of cloud pricing might be coming to an end. Asian cloud giant Alibaba recently announced a price increase of as much as 34 percent for some services [1]. The biggest spike was for AI workloads, but even conventional server space saw an increase. Earlier this year, Amazon Web Services (AWS) raised their prices by 15 percent for EC2 capacity blocks for machine learning (ML), a GPU-intensive product often used for ML workloads. According to the Register's Corey Quinn, "To be clear, AWS has raised prices before, but rarely as a straight increase to a line item. The company prefers to change pricing dimensions entirely, often spinning this as a price reduction for most customers – a claim I'd characterize as creative. Historical straight-up price increases have been tied to regulatory actions: per-SMS charges in certain markets and the like. This is different." [2]

The reason for the upward trend? All this falls in with the big story that is dominating the high tech marketplace: the AI gold rush. It is hard to avoid the news that AI-focused data centers are springing up all over the world – if you follow IT, I'm not the first to tell you this. What is perhaps less well documented is the effect the AI bubble could have on ordinary meat-and-potatoes cloud customers who just want conventional server space. First, there is the rising cost of RAM as AI vendors bid up the price and rush to build more and bigger facilities (this increase is also well documented in the press). But then there is also the subtle shift in focus for companies like AWS. As their customers rush to rent space for AI training, the burgeoning demand causes AWS to raise their prices for the GPU-intensive capacity blocks, which, ultimately, causes them to commit more of their focus and resources to the AI sector, which forces their conventional customers into a diminishing swath of the company's business and could, some experts believe, eventually force conventional customers to pay higher prices even if they aren't using AI.

What are these conventional customers supposed to do? Regress to local enterprise server farms (but the price of RAM went up)? Switch to a second-tier cloud provider (which could also result in higher prices, although possibly better customer service)? And then there is the vendor lock-in that comes with trying to untangle from a big cloud vendor.

So it looks like there is a chance that even companies that aren't participating in the big AI scale-up will be paying to support it. Oh well in that case, why not join the party…

Joe Casad, Editor in Chief