Right now, we are in the hype cycle for AI, Gartner calls it the “peak of inflated expectations.” Representations are made and many surprises are unearthed. This is an interview that focuses on definitions, reliability, and automation about applying AI to problems warfighters face.

What is AI?

Years ago, Alan Turing came up with the baseline definition of artificial intelligence – being able to mimic human responses under specific conditions. Some individuals in the episode argue we haven’t even approached artificial intelligence. A statement is made that just because you call it AI doesn’t make it AI.

Is the data reliable?

Years ago, the colloquial phrase was, “garbage in; garbage out.”  Today’s massive data stores reinforce this concept.  However, the phrase becoming more and more popular is, “authoritative data source.”  Each thought leader in the discussion has ideas on cleaning data.

What is the role of automation?

During the interview, Michael Pomatto NAVAIR suggests that today’s interpretation of artificial intelligence has ramifications in the systematic processing of information.  Robotic Process Automation is the first step. When machine learning is added, he calls it hyper-automation.  What does it mean to do mundane office tasks? 

What does the future hold?

Jaime Fitzgibbon, Defense Innovation Unit makes a provocative statement when she asks how technology leaders can plan ten years out when game-changing technology hasn’t been invented yet. 

Future interviews will expand into topics like scaling, the amount of data needed, and bias built into the algorithm.  These experts set the baseline for a fascinating future.

 

Twitter:  @FedInsider

LinkedIn:    https://www.linkedin.com/company/fedinsider/

Facebook:   https://www.facebook.com/FedInsiderNews

 

 

Reference:   https://www.ynetnews.com/article/sjg0qah83