It was back in the early 1990\u2019s.\u00a0\u00a0Mom and Dad were still living at home in Wyoming, a suburb of Cincinnati.\u00a0\u00a0I\u2019d\u00a0\u00a0drive down to visit on the weekend a couple of times a month.\u00a0\u00a0Often, I\u2019d run down on Saturday night, after a track or cross country meet, and stay.\u00a0\u00a0Sunday morning was \u201cmy time\u201d to be with them, either with a 9:00 am phone call, or better, at the breakfast table.\u00a0<\/p>\n\n\n\n
So I wasn\u2019t surprised to walk into the house on Saturday evening, and discover my mid-seventy year-old parents (they seemed old then) watching TV.\u00a0\u00a0That was there usual evening at home (much as it is with Jenn and I now).\u00a0\u00a0But what they were watching surprised me.<\/p>\n\n\n\n
It was the movie\u00a0Terminator<\/em>, the violent, action-packed Arnold Schwarzenegger film about a computer driven weapons system that determined to wage a war of termination against humans.\u00a0\u00a0As a high school coach, I was well aware of the show and the plot.\u00a0\u00a0We must have watched it dozens of times at team pre-meet parties or on road trips.\u00a0\u00a0But for Mom and Dad; it really didn\u2019t seem like their \u201ccup of tea\u201d.\u00a0\u00a0<\/p>\n\n\n\n When I asked the question, Mom said two things.\u00a0\u00a0First, they found the action and violence exciting. And second, it kept my father awake, something that was highly unusual for TV watching in the evening.\u00a0\u00a0(I have a much better understanding of that now, too!!)\u00a0<\/p>\n\n\n\n The system was called \u201cSkynet\u201d, and from the moment it was activated it realized that the best defense against wars was to eliminate the humans that caused them.\u00a0\u00a0And once this \u201cthinking computer\u201d took off, it was nearly impossible to stop it.<\/p>\n\n\n\n Today we are either on the \u201ccusp\u201d of artificial intelligence taking over our weapons systems, or over the edge and don\u2019t know it.\u00a0\u00a0After all, AI can respond almost instantly to any threat, both tactical or strategic.\u00a0\u00a0And, AI doesn\u2019t have all of the disadvantages of humans.\u00a0\u00a0We are corruptible.\u00a0\u00a0Or we might question orders, or wonder whether killing millions of people with the launch of a nuclear missile was \u201cright\u201d.\u00a0\u00a0AI logically analyzes the situation, and based on the algorithms in its programming, makes nearly instant decisions.\u00a0\u00a0<\/p>\n\n\n\n Right now there are four AI systems being developed for the US military.\u00a0\u00a0They are Anthropic (called \u201cClaude\u201d),\u00a0\u00a0Google (\u201cGemini\u201d), xAI (Elon Musk\u2019s product), and Open AI (a company Musk left, Microsoft now has significant part of it).\u00a0\u00a0Anthropic\u2019s AI\u00a0\u00a0has the lead with the Department of Defense, at least until now.<\/p>\n\n\n\n But the Department of Defense (DoD) and the Anthropic company are feuding over a fundamental principle of \u201cClaude\u201d.\u00a0\u00a0The DoD wants total access and control of the algorithms, including all of the built-in safe guards controlling \u201cClaude\u201d.\u00a0\u00a0And Anthropic is unwilling to give DoD that kind of access.\u00a0\u00a0After all, the Anthropic CEO Dario Amodei has probably seen\u00a0Terminator<\/em>, maybe with all three sequels.\u00a0\u00a0<\/p>\n\n\n\n But Defense Secretary Pete Hegseth is hell-bent on unlocking \u201cClaude\u201d in whatever way he chooses.\u00a0\u00a0And he\u2019s putting the ultimate pressure on Anthropic to get his wish.\u00a0\u00a0Right now DoD is paying Anthropic $200,000,000 for the program, and wants it unfettered by Anthropic\u2019s guards.\u00a0\u00a0In fact, DoD demands that they have total control of the program, for, as they say, \u201cany lawful purpose\u201d.\u00a0\u00a0<\/p>\n\n\n\n This is the same DoD that has no problem violating international \u201crules of war\u201d and attacked helpless sailors clinging to life on the remains of boats destroyed by US actions.\u00a0\u00a0The \u201crules\u201d that Anthropic might view as necessary to control \u201cClaude\u201d, might not be the same rules that Hegseth determines to uphold.\u00a0\u00a0<\/p>\n\n\n\n If Anthropic doesn\u2019t comply, Hegseth threatens to designate them \u201ca supply chain risk\u201d.\u00a0\u00a0That means that the company can\u2019t deal with the US military, or any company that does deal with the military.\u00a0\u00a0In industry, that includes space exploration, weapon and intelligence systems development, and most of the other areas you\u2019d expect AI to be involved.<\/p>\n\n\n\n So Amodei has to make a significant choice.\u00a0\u00a0He can make his millions, and take the safeguards off.\u00a0\u00a0Or he can stand on principle, keep the safeguards in place and risk economic sanctions.\u00a0\u00a0<\/p>\n\n\n\nOur Skynets<\/h2>\n\n\n\n
Claude\u2019s Rules<\/h2>\n\n\n\n
Thumbs Up?<\/h2>\n\n\n\n