IS ARTIFICIAL INTELLIGENCE READY TO BECOME THE NEXT “KING” YET? - PART 2

avatar
Robert Massey
Clock icon
12 min read
Calendar icon
July 23, 2024

How AI Can Make Us Safer and More Capable

In 2015, DARPA, in conjunction with the Sikorsky Aircraft Company, partnered on a program known as ALIAS (Aircrew Labor in Cockpit Automation System). To date, the program has resulted in the development of a fully autonomous Blackhawk Helicopter. But here’s the thing – a helicopter that can fly itself is cool and all, but nobody actually wants one.

What gets lost on the average person is in the name of the program itself: “Aircrew Labor.” Think about the number of tasks a pilot may be doing at any one time: maintaining an airspeed, an altitude, heading, monitoring other traffic in the air, communicating with multiple people on multiple radios, navigating, monitoring fuel consumption, avoiding obstacles. Then what happens when there’s an in-flight emergency or you’re trying to land into a degraded visual environment?

ALIAS was developed for Human-AI synergy to do more and do it safer. This is exactly how great organizations hire AI colleagues that work for their mission purpose and yield superior results.

Why Agile?

During the early part of my Army career, hardware was king, as I mentioned earlier. If you brought the better weapon systems to the fight, you won. In the latter half of my career, I saw software emerging as the new king. On today’s battlefield, success goes to the side that can deploy and push software updates to their weapon systems the fastest. This shift towards software dominance has been pivotal. Agile development methodologies have become increasingly crucial in this context. In the commercial sector, agile practices have facilitated rapid innovation and adaptation, outpacing the progress seen in our government partnerships.

Take the Department of Defense’s (DoD) most expensive fighter jet – the F-35. Although it is the most sophisticated fighter jet in the military, the DoD struggles to deliver software upgrades to that program on a monthly basis. Contrast this with companies like SpaceX, which can push software updates daily, up to the moment of launch. This stark difference underscores the agility advantage enjoyed by companies in the commercial space, where rapid software iteration can mean the difference between success and failure in highly competitive markets.

fighter-jet-image

Producing Tangible Results With RLEF

Having witnessed these differences, I believe in having an agile mindset when it comes to delivering innovative solutions. Techolution has been doing this for almost a decade on the commercial side. That said, I’d like to talk about one of the success stories with implementing Reinforcement Learning with Expert Feedback (RLEF). Techolution took up a project that revolutionized the recruitment process for both nurses and healthcare employers in the thick of a nationwide nursing shortage. By integrating the job matching feature into nurse.com, which serves over 200,000 registered nurses with certification, license tracking, and education, the project allowed nurses and employers to interact with each other and more.

Employers were able to invite candidates for interviews, reject profiles with specific reasons, and communicate directly through personalized messages. Even better, nurses were able to receive interview requests, view job details, and indicate their interest in job postings. This is where our RLEF comes in. When a match is rejected, the platform privately collects the reason to personalize future suggestions for each nurse or hospital, similar to how ChatGPT’s memory works. This ensures that similar candidates or positions are not recommended again based on individual preferences.

RLEF is also applied to approved matches, refining the platform’s understanding of whom to consider for similar jobs in the future. This enhances the accuracy and efficiency of the matching process, making it more reliable and effective for both parties involved. Think of it as your very own AI matchmaker, but without the cheesy pickup lines. By continually learning from the feedback of both nurses and employers, the platform evolves, becoming more adept at matching the right candidates with the right positions, ensuring a more satisfying experience for everyone.

That’s Trusted AI, That’s AI Done Right!

Soon, and perhaps we’re already there, we will watch AI reign as the new “king”. It’s hard work, but it’s work worth doing to deliver capability at the speed of relevance.

Human-AI synergy is key to maintaining a high standard of AI integrity and successfully implementing AI solutions. With RLEF, you can trust that AI AI systems to be effective, transparent and reliable. By fostering collaboration between human experts and AI, we can pave the way for AI to become the new “king”, delivering trusted solutions that attain desired outcomes for our customers – who knows, maybe one day, our AI assistants will even understand our cryptic coffee orders. That’s when we can say with a straight face, “We will do more with less”, and we’ll do it better, faster, and safer. That’s what real-world AI can deliver. That’s trusted AI, that’s AI done right!