Connect with top gaming leaders in Los Angeles at GamesBeat Summit 2023 this May 22-23. Register here.

It’s been a busy summer in the artificial intelligence (A.I.) space, but the most interesting A.I. opportunities may not come from the biggest names.

You may have heard about Tesla’s self-driving cars that made headlines twice, for vastly different reasons — a fatal crash in Florida in which the driver was using the Autopilot software, and claims by a Missouri man that the feature drove him 20 miles to a hospital after he suffered a heart attack, saving his life.

Or you might have heard of Apple spending $200 million to acquire machine learning and A.I. startup Turi. A smart drone defeated an experienced Air Force pilot in flight simulation tests. IBM’s Watson diagnosed a 60-year-old woman’s rare form of leukemia within 10 minutes, after doctors had been stumped for months.

But believe it or not, enterprise IT is also a fertile ground for A.I. In fact, some of the most immediate and profound use cases for A.I. will come as companies increasingly integrate it into their data centers and development organizations to automate processes that have been done manually for decades.


Intelligent Security Summit On-Demand

Learn the critical role of AI & ML in cybersecurity and industry specific case studies. Watch on-demand sessions today.

Watch Here

Here are five examples:

1. Predicting software failures

Research at Harvard University’s T.H. Chan School of Public Health shows it may one day be possible to use A.I. algorithms to evaluate a person’s risk of cardiovascular disease. The study is testing whether these algorithms can draw connections between how patients describe their symptoms and the likelihood that they have the disease. The algorithms could lead to the development of a lexicon to interpret symptoms better and make more accurate diagnoses faster.

In a similar way, A.I. algorithms will be able to review and understand log files throughout the IT infrastructure in ways that are impossible to do with traditional methods and which could be able to predict a system crash minutes or hours before a human might notice anything was wrong.

2. Detecting cybersecurity issues

The list of high-profile cyberattacks over the last couple of years keeps growing and includes the theft of the confidential data of tens of millions of Target customers during the height of the holiday shopping season in 2015, the server breach at the U.S. Office of Personnel Management that compromised the sensitive personal information of about 21.5 million people, and the recent infiltration of the computer network of the Democratic National Committee by Russian government hackers.

Artificial intelligence holds great promise with its ability to learn patterns of networks, devices, and systems and decode deviations that could reveal in-progress attacks. A crop of startups is focused on these approaches, including StackPath — founded by entrepreneur Lance Crosby, who sold his previous company, cloud infrastructure startup SoftLayer, to IBM in 2013 for $2 billion.

The DefenseAdvanced Research Projects Agency, or DARPA (the agency that helped create the internet), recently sponsored a contest in which seven fully autonomous A.I. bots found security vulnerabilities hiding in vast amounts of code. The next day, the winning bot was invited to compete against the best human hackers in the world. It beat some of the human teams at various points in the competition.

A.I. just might hold the key to finally beating the hackers.

3. Creating super-programmers

The fictional superhero Tony Stark in the “Iron Man” movies relies on a powered suit of armor to protect the world. A.I. could offer similar capabilities to just-out-of-college software developers.

We all know Siri. Under the hood, she’s an A.I. neural network trained on vast amounts of human language. When we ask her directions to McDonald’s (not that I would admit that I do that sort of thing), she “understands” the intention behind the English words.

Imagine a neural network trained to understand all the source code stored in GitHub. That’s tens of millions of lines of code. Or what about the entire history of projects as robust as the Linux operating system? What if Siri could “understand” the intention behind any piece of code?

Just as Tony Stark depends on technology to get his job done, ordinary programmers will be able to turn to A.I. to help them do their jobs far better than they could on their own.

4. Making sense of the Internet of Things

Recent research forecasts that A.I. and machine learning in Big Data and IoT will reach $18.5 billion by 2021. It’s no wonder. The idea of devices, buildings, and everyday objects being interconnected to make them smarter and more responsive brings with it unprecedented complexity.

It’s a data problem. As IoT progresses, the amount of unstructured machine data will far exceed our ability to make sense of it with traditional methods.

Organizations will have to turn to A.I. for help in culling these billions of data points for actionable insights.

6. Robots in data centers

Ever seen this cool video of robots working in an Amazon distribution center? The same is coming to large corporate data centers. Yes, physical robots will handle maintenance tasks such as swapping out server racks.

According to a news report, companies such as IBM and EMC are already using iRobot Create, a customizable version of Roomba, to deploy robots that zoom around data centers and keep track of environmental factors such as temperature, humidity, and airflow.

Self-driving cars are far from the only advances pushing A.I. boundaries. The innovations in enterprise IT may be happening behind the scenes, but they’re no less dramatic.

VentureBeat's mission is to be a digital town square for technical decision-makers to gain knowledge about transformative enterprise technology and transact. Discover our Briefings.