This article tells me what I already understand: Use of cloud services, including compute and storage, is about to inflect upwards, perhaps more than we’ve seen since cloud computing became a thing.
Reporter after reporter has been calling me over the past few months about how cloud providers are suffering due to the downturn. This is déjà vu all over again from the start of the pandemic when everyone was running around in circles waiting for the cloud market to collapse.
As you may recall, that did not happen. Much to the surprise of many reporters and analysts, cloud usage exploded as companies shifted to remote work, and physical data centers became inaccessible and more of a liability.
So, how is this the same? The renewed interest in AI, especially generative AI systems, will lead to some logical conclusions we should be thinking about now:
- It’s obvious that AI is going to grow rapidly.
- AI, certainly generative AI, needs huge amount of storage and compute power.
- The most economical way to consume those resources is through a public cloud provider.
- Public cloud usage (and revenue) will explode.
It’s not much of a logical leap if you ask me. Analysts and press are guesstimating that systems like ChatGPT cost as much as $700,000 a day to operate. This just proves what most people operating and paying for cloud resources already know: AI is costly to own and run, mostly because AI is a cloud resource hog.
Someone is going to make all that money, and it's going to be the public cloud providers that offer AI services and have the infrastructure resources to support these services. These companies are making huge investments in AI, specifically generative AI, that should come back to them as revenue and value almost immediately.
What to do?
We’ve established that the cloud providers will do well during a shift in the market…again. How should the rank-and-file enterprise CIOs be thinking about this shift, and what do they need to understand?
First, the rapid rise in public cloud usage could be a mixed bag in terms of benefits and net negatives. Some cloud providers may drop prices to grab the rapidly growing cloud AI market. In that scenario, I would advise my clients to lock in the lower prices through enterprise agreements, if they are indeed compelling.
However, I can also see the opposite occurring. If demand for cloud computing grows too rapidly, prices go up. We saw this recently with eggs and bacon at the corner store; the same economic reality applies to cloud computing.
Second, no matter if prices go up, down, or remain the same, it’s a good time to get a finops program in place or enhance the one you have. Cloud resources are often wasted if the cloud consumer doesn’t understand how to manage cloud usage effectively. If we expect a rapid expansion in cloud use, all the more important to get better at managing cloud resources so they are able to return an optimized value back to the business.
Finally, get your plan in place for the use of AI and cloud. Most enterprises will tell you that they are likely to leverage generative AI for real business applications (beyond writing thank-you notes or creating creepy pictures). However, they can’t tell you what those are or how much value AI will likely return. You need to answer those questions before you declare that you’re moving to any technology.
Planning seems to have taken a back seat to reactive thinking around the use of technology, including AI. Create a sound plan now that shows you the best path for your business to adopt any technology. Otherwise, you’re likely to spend more than you need and not return enough value, just considering the number of costly mistakes that you are likely to make.
Is exploding cloud usage again a good thing? It has good and bad aspects to it. The best advice is to understand what’s occurring and put together a strategy to take advantage of this market shift.