Press "Enter" to skip to content

Did AI blow up your cloud bill?

The term artificial intelligence it was first used in a 1955 proposal for a study presented by John McCarthy of Dartmouth College, Marvin Minsky of Harvard University, Nathaniel Rochester of IBM, and Claude Shannon of Bell Telephone Laboratories. This happened before I was born. It seems a bit crazy to me that AI was discussed long before we had the computing and storage power to make it work.

As a decision support analyst fresh out of college, I built the first AI systems that were too expensive to run, so they were only used in special circumstances. It was a niche technology. Due to its high operating costs, AI’s popularity fell from the early 1980s until about five or seven years ago. Now, the on-demand consumption model of cloud computing and much better AI technology have substantially reduced operating costs, and AI is once again at the center of enterprise IT.

Public cloud providers are the driving force behind the current AI resurgence. Although AI technology is now better optimized (and let’s admit, it’s fun to play with), you need to fully understand the business value it can return, and recognize when the ROI isn’t there.

What is more valuable to AI than the cheapest and most powerful computing cycles? The fact that storage is a commodity. AI gets its power from learning data and learning patterns in that data, not from cleverly written algorithms. The more data that is available to a learning model, the more focused the data becomes and the better knowledge or understanding it creates.

Despite their substantially lower operating costs and the potential value that AI and machine learning can bring to a business, the performance falls short in many cases. 2022 was a year of huge cloud cost overruns. Misuse of cloud resources by a business as a whole creates the majority of cloud cost overruns. In some cases, this means choosing cloud AI/ML systems when more pragmatic alternatives could deliver more value.

Many AI/ML systems are much more expensive to maintain. Specialized skills are needed to build and implement these systems and then operate them. “Cloud AI” just means that the data processing and storage is outside the company. Massive amounts of general-purpose and specific-purpose data are needed to power AI engines, and that data must be continuously stored, managed, and protected. You also need to deal with data compliance.

Also Read:  AWS re:Invent 2022 roundup: Data management, AI, compute take center stage

In many cases, the business has custom needs that require custom training data that is not part of the general-purpose transactional business database, but is unique to support a specific AI system need. That means more storage, more labeling, more streaming, and more operating costs.

All of this can be worth it if there is a strong business use case. In many cases, there isn’t. The easy availability of AI led to its use where it is not needed.

For example, a good use case for AI could be a sales order entry system that leverages machine learning to determine recommendations that are automatically presented to customers who order online. AI could increase sales and therefore return business value. However, AI is most often used within traditional transactional systems where it provides only a minor advantage. An example of misuse would be running AI to verify a valid shipping address to reduce shipping errors.

Remember, there are two sides to every AI use case. In the second example, the shipping savings could be a few thousand a month, which is nice. But the cost to develop and operate the cloud-based AI system could be up to 20 times that amount per month, which is bad. There are on-demand solutions that don’t use AI but are equally or more effective and can be had for a few hundred dollars a year.

The problem is the gate. Cloud vendors and consultants often recommend AI for use cases where it will not provide the necessary ROI for the business. It slips away if someone doesn’t ask the hardest questions or if a strong business case is never made.

It’s not about whether the AI ​​works, it always works. It is the misapplication of cloud AI systems that detracts from business value. Make that mistake often enough and the business will cease to exist.

I am not rejecting AI or AI in the cloud. I have created amazing applications that use AI concepts and technologies, and I will make many more in the future. This technology can do amazing things. However, as with any technology, AI has its place. We need to be more attentive to the ROI of its use.

Copyright © 2023 IDG Communications, Inc.

Be First to Comment

Leave a Reply

Your email address will not be published. Required fields are marked *