Micro Focus is now part of OpenText. Learn more >

You are here

You are here

5 AI Articles We Almost Forgot We Love

Joe Stanganelli Managing Editor, TechBeacon
Photo by Karolina Grabowska on Pexels

Valentine's Day has come and gone. Maybe you forgot. Or maybe someone close to you forgot.


It happens to the best of us. I once forgot how difficult it is to schedule a flower delivery for Valentine's Day if you don't remember to do it until February 13. Maybe you know what that's like. Or maybe you know what it's like to not get flowers until after the fact.

Or maybe you're at a point in your life where you prefer to forget about Valentine's Day. And that's fine. There's nothing wrong with a little cynicism—if not misanthropy—every now and then. As Jean-Paul Sartre put it, "Hell is other people."

So let's turn to the machines. Artificial intelligence (AI) may not be able to truly love (yet), but it can be taught biases. It may not actually possess emotions, but it can write a love poem expressing them nonetheless. It can't keep you warm at night (unless, I suppose, it's running your climate-control system), but it can keep your network safe from cyber-attackers. It could probably even find a better framing device for a roundup of TechBeacon AI articles than I have. And it would do a better job of remembering Valentine's Day, too.

So I looked back—manually—at some of TechBeacon's AI coverage and reminded myself of some of my favorites. Here they are, with my own manmade odes to go with the links and summaries.

1. Testing for Bias in Your AI Software: Why It's Needed, How to Do It

Roses are red
Violets are blue
Humans are biased
So AI is too

Maybe you didn't need to be taught to love, but computers do. So too with being taught to hate. And every feeling and perspective in between. AI, at least at this stage, can only imitate. So if, for example, you teach it to use race or gender as critical metrics, whether you mean to or not, that's exactly what it will do.

In this overview, Peter Varhol, principal at Technology Strategy Research, explores real-world examples of bias in AI, the main categories of AI bias, and why bias occurs in AI systems. From there, he gives some tips for QA testers on discovering and combating AI bias.

2. To Keep Up, SecOps Teams Need AI and Automation

Roses are red
Defenders are blue
Keep up with bad actors
ML'll help you

Advanced persistent threats (APTs) are called "advanced" for a reason; increasingly, they're nearly impossible to detect by human defenders (particularly as the tech industry contends with a cybersecurity skills shortage). In this article, Satyavathi Divadari, chief cybersecurity architect at CyberRes, writes that machines are coming to the rescue—with nearly four out of five surveyed organizations using or planning to use AI and machine learning (ML) to help combat cybersecurity threats.

Divadari also highlights some of the types of AI/ML cybersecurity tools on the market—including behavioral analytics and incident response.

3. Why Isn’t AI Working for Your Business?

Roses are red
AI is hyped
Improve DevOps and hiring
And know your data types

AI is a super-hyped technology. Some might argue over-hyped; some might argue not hyped enough. Either way, business leaders have been having difficulty finding the actionability beneath the hype. In this article, writer Soundarya Jayaraman points to four common problems that doom AI projects before they begin: 

Jayaraman offers some tips, accordingly, for moving past these problems in the planning stages to get AI projects started off on the right foot.

4. Will AI Delete Your App Sec Job? 5 Lessons for Software Teams

Roses are red
There's no need to sob
AI can write code
But you'll still have a job

Pundits have long speculated that the endgame for AI involves machines taking our jobs. The idea is nothing new, and new technology does often mean job obsolescence. But it also often means new jobs being created to manage and contend with the new technology.

Here, writer Rob Lemos explores how software developers are and may be affected by AI that can code. Lemos concedes that devs are facing disruption, but he argues that the results will be that humans and machines will complement each other in coding and application security. The exact job descriptions and workflow dynamics may change, but AI will augment human intelligence more than it will replace it.

5. ChatGPT and the State of AI

Roses are red
Could ChatGPT
Have written these poems
Any better than me?

TechBeacon's most recent AI coverage, this article picks up where Rob Lemos left off pre-ChatGPT. Indeed, in the wake of the release of OpenAI's ChatGPT, the AI hype has gotten heavy.

But is ChatGPT smart? Is it "real" AI? And what is "real" AI? When does AI become real?

Matthew Heusser, managing consultant at Excelon Development, takes a thoughtful approach to these questions. In this article, Heusser provides some historical highlights on generative AI and contextualizes them more practically—in terms of behavior instead of philosophy. From that vantage point, he then takes a high-level look at ChatGPT itself—along with its abilities to generate both creative text and functional code. In particular, Heusser is intrigued by ChatGPT's ability to learn and be trained.

For me, this raises a deeper behavioral and philosophical question about AI. What's more representative of humanlike intelligence? The ability to remember, or the ability to forget?

Happy belated Valentine's Day, TechBeacon readers. We love you. Don't forget.

Keep learning

Read more articles about: Enterprise ITDigital Transformation