Welcome to the episode number 11 of the Authors of Mass Destruction podcast. My name is Natasha Bajema, aka WMDgirl on Twitter. I’m a fiction author, national security expert and your host for this podcast.

• If you’re interested in science & technology, in reading good fiction, or want to write fiction based on technology, you’re in the right place.
• Each week, I kick the episode off with a technology news headline that has caught my attention that week and answer listener questions.
  • If you’re interested in submitting a question to the show, become a patron for only a few dollars a month at Patreon at www.patreon.com– p a t r e o n / natashabajema
  • The support of my patrons makes this podcast possible. In addition to my time, there are hosting and marketing costs associated with the show. Every little bit helps me continue doing the show.
• Before we get started, a few notes:
  • The views expressed on this podcast are my own and do not reflect the official policy or position of the National Defense University, the Department of Defense or the U.S. Government.
  • The AOMD podcast is proud to be part of the Authors on the Air Global Radio Network. Check us out at www.authorsontheair.com
• I’ve got two news headlines for this week, both about artificial intelligence:
  • “Amazon’s warehouse-worker tracking system can automatically pick people to fire without a human supervisor’s involvement” published on April 25 on businessinsider.com
  • You’re probably aware that more and more companies are turning to machine learning tools to manage their highly complex operations.
  • Amazon uses a machine learning tool to automatically generate the paperwork to fire employees if they’re not meeting targets.
  • “If the system determines the employee is failing to meet production targets, it can automatically issue warnings and even termination paperwork, all without a supervisor’s intervention, although Amazon said that a human supervisor can override the system.”
  • I don’t know about you, but this makes me feel extremely uncomfortable. If I knew that my productivity was being monitored by a computer, I’d become paranoid about taking bathroom breaks, about checking my personal email or making an important phone call to address a mortgage issue, etc.
  • My second headline is related, but takes similar issues to the battlefield: “How AI Could Change The Art Of War” published by Sydney J. Freeberg, Jr on breakingdefense.com
  • This article examines how we are growing increasingly dependent on automated systems on the battlefield and explores what it means to rely on machine learning to make national security decisions.
  • Think about how you use GPS to reach your destination. Do you believe the route chosen? Do you follow the instructions or do you deviate.
  • In reality, we’ve been depending on automated systems in defense for decades:
    • Early warning systems for detecting nuclear missiles
    • Radar
    • The author brings up the Aegis air and missile defense system on dozens of Navy warships, which “recommends which targets to shoot down with which weapons, and if the human operators are overwhelmed, they can put Aegis on automatic and let it fire the interceptors itself.”
  • One of the major challenges/risks in delegating any aspect of our decision-making to machine learning tools is that we don’t understand how the algorithm reached its decision. Nor can we.
  • DARPA is working to develop what it calls “explainable AI”, but as the author suggests, it could be counterproductive to constrain AI in terms that can be understood by humans. He points to examples where machine learning tools have defied human understanding of chess or the Chinese game of Go.
  • He also gives an example from Amazon’s warehouse operations. When the operations are controlled by machines, the more efficient storage principle is “random stow”. The machine stores items randomly in the warehouse but knows where each item is in his database. When a customer order comes in, it calculates the most efficient pathway to pick up all of the items in an order.
  • Humans need things to be stored in an organized fashion to operate efficiently.
• Let’s get to my interview. Today, I’m talking to Craig Martelle who is a bestselling science fiction and fantasy author, a retired U.S. Marine and the co-mastermind of the self-publishing Facebook Group 20Booksto50K.