Getty Images/iStockphoto
Look beyond AI and ML buzzwords for the cloud
AI and machine learning are more than buzzwords, they can have serious impact to your business when applied correctly. Learn how to identify the true cost savings of these tools.
Concepts like the cloud, AI and machine learning can be easy to misunderstand, due to varying definitions and hype. As more companies move to the cloud, they request services such as AI and machine learning, often without knowing what they're asking for.
Let's set the stage about what AI and machine learning (ML) are, what they can do and why enterprises might choose to use these cloud services.
ML vs. AI
AI is a computer or computer system designed to mimic cognitive functions for learning and problem-solving. It applies math and logic processes to simulate how humans think.
Machine learning is an application of AI in which a computer or computer system uses complex mathematical models to help a computer learn without user intervention. As machine learning gathers more data, the results improve due to the continued learning based on the system's own experience.
To summarize, AI is a system that reasons, adapts or acts like a human, while machine learning takes a data set and extracts knowledge from it to make a model learn autonomously.
Look beyond the buzzwords
It can be difficult to find a product that doesn't reference AI or machine learning in some way so vendors can push a product line or feature. How do enterprises look beyond these buzzwords to obtain the services?
AI and machine learning are most often deployed as cloud services, not tools that enterprises install locally. This cloud model is necessary because the services require a large scale of resources, although most enterprises likely don't need them 24/7. Most IT security products mention that their cloud-based AI technology can help protect a company's environment. For example, one top security product uses the phrase "advanced AI learning" but doesn't explain how or what that capability does.
Most likely, these products apply machine learning instead of true AI. Because machine learning is a subset of AI, vendor descriptions about AI capabilities can become gray. That doesn't necessarily mean they're lying or have bad products, but marketing departments might be taking a few liberties with AI and machine learning terms.
Benefits of AI and ML services
AI and machine learning services can help enterprises save money on cloud. Using an AI service, such as machine learning, enables enterprises to look at their cloud usage and use that data to drive autoscaling decisions about how to allocate cloud resources, which can save money.
It might seem counterintuitive to spend money on an additional service to save money. But a machine learning service can yield back bigger savings over time rather than in the short term.
Machine learning can also work with applications in the cloud to deliver more accurate results in data mining or predictive models. Machine learning tends to focus more on results when it comes to cloud-based applications or outcomes. AI has a much wider scope, but it can be harder to apply to traditional use cases, with the exception of security.
Security threats exist everywhere, and more vendors are using cloud-based security platforms to help keep systems safe. It's common for vendors to move security products off premises to subscription models, which also enables the use of machine learning and AI in ways where they can stand out.
For example, next-generation security often deploys firewalls and antivirus as cloud-based resources, where they can better protect resources. This model enables these services to take advantage of AI to look for zero-day threats in a way virus signatures never could. This has helped power next-generation cloud-based security tools that can better scale for the current threat landscape.
Downsides of AI and ML
While systems that use AI offer benefits, they also present serious challenges if enterprises aren't watching closely. Bias, copyright issues and lack of regulations are some major concerns regarding AI and machine learning.
In 2017, for example, Amazon used AI-based tools to screen job applicants. Over time, the system became biased against women. Once Amazon discovered the bias, it terminated the AI effort. More recently, tools like ChatGPT have crept into classrooms and courtrooms, raising concerns about plagiarism, accuracy and regulation.
Enterprises interested in using AI and machine learning should focus on the deliverables of what each tool can bring. Ask vendors for examples, and evaluate the cost savings metrics. Enterprises that do their research and understand how to use the tools for specific purposes will see benefits. Conversely, businesses interested simply in capitalizing on the latest buzzwords in the IT industry might run into challenges.
