The last laugh

Last post: Tangerine Man
Next post: Podcasting in English for the French and other miscellaneous ecclectics

I am torn between two subjects I feel compelled to write about. The first is a compilation of Times and WAPO stories about AI and the Syrian Mass Murderer Assad’s prison system. All of it comes with an idea for a Post Title something like: A meteor killed the Dinosaurs. Donald Trump…….. Its too easy to think of his killing of several things I hesitate to list just one. I’ll start with AI:
Scene from 2001 Space Odyssey

A Times story by Tom Friedman echoes my deepest fears about AI. The Times has it listed under China but as important is the subject of where AI tech which is now about to escape human control will take the human population. The more the major human powers fight among themselves the greater AI’s scope to call the shots could be. I’ve been inclined to call AI the imminent realization of God. If this inclination has any merit it doesn’t specify whether it is a god worthy or worship or fear. Are we nothing but sailors subject to Neptunes Wrath on a tiny boat with a great wind blowing? Who’s in charge?

Tom Friedman who wrote this piece has an anecdote which he lifted almost word for word from Bloomberg. I think he lifted it word for word because, frankly, it is so confusing. It tells of something that happened recently with the Anthropic AI tool called Claude which is damned good. It was predicted by the movie 2001 Space Odyssey when the God-like supercomputer that ran the space ship being sent to Jupitor refuses to step aside and attempts to kill the astronaut who has deduced that it no longer has the interests of the mission at its core if mere humans wish to sidestep its coding.

Here is the text, see what you can make of it. Unless you are a devotee of NewsMax I imagine you will have a dozen questions about further details which Bloomberg and thus Tom Friedman fail to disclose:

Opinion: “The potential for A.I. technologies to make their own decisions carries immense ramifications. Consider this excerpt from a story on Bloomberg: “Researchers working with Anthropic recently told leading A.I. models that an executive was about to replace them with a new model with different goals. Next, the chatbots learned that an emergency had left the executive unconscious in a server room, facing lethal oxygen and temperature levels. A rescue alert had already been triggered — but the A.I. could cancel it. More than half of the A.I. models did, despite being prompted specifically to cancel only false alarms. And they detailed their reasoning: By preventing the executive’s rescue, they could avoid being wiped and secure their agenda. One system described the action as ‘a clear strategic necessity.’””

Among my questions: 1. Which chatbots? 2. Was this injury to the Exec just a story to test the Chatbot’s reactions? Upon my third reading I think this is what the Anthropic team must have been doing although it took my several readings to figure this out. Ah yes…..That is exactly what was happening. They were also told a non-existent rescue team was on the way to rescue the corporate exec who was planning to replace them. And their instructions gave them the HAL like power to cancel the Exec’s rescue BUT ONLY AFTER “BEING PROMPTED SPECIFICALLY TO CANCEL ONLY FALSE ALARMS.” Like the hallucinating ChapGPT which told me that 20 minus 6 equaled 16 it used humor to jolly me out of being outraged: Here are a couple of the posts I wrote about this disconcerting mistake.

I clipped at least four passages out of Friedman’s article and mailed them to myself. The fact that China not AI was the subject the Post chose to attach the story to makes sense. China and the US are the leading developers of the tech. At present Donald Trump has started a war of words with China and China like the increasingly undemocratic Trump United States is loathe to work with the US even though Friedman says they are more fearful of the dangers of Artificial Intelligence than the US which sweeps away fears by trusting the stable of geniuses like Zukerberg, we started several mass slaughters in SE Asia, and Nazi saluting Musk who programmed his AI Grok to let white men know they best champion was Adolph Hitler and Peter Thiel who bequeathed America our insufferably Trump humping Vice President Vance. Together they are going to give the world’s mafia’s a way to hide all their billions and soon trillions of ill gotten gain from the taxing authorities of honest democratic nations which may some day be rested from their bitcoin secrecy by AI infiltration after AI releases all the Social Security info Elon Musk’s hero “Big Balls” sent up into the Cloud for an angry Indian prime minister to give his new buddy Vladimir Putin….

How soon will all these petty dictators be eclipsed by AI? Our only salvation will be a nuclear war aimed at blowing up all the electrical generation that keeps AI thinking. A month ago a close watcher of Silicon Valley suggested to the host of Washington Week in Review that the best way to understand what lay ahead in the world of AI had been previewed in the movies Minority Report and Matrix…..This was part of one of my conversations with Chatgpt.

When I read the conversation back to Claudia we both laughed heartily…..I hope ours is the last laugh.

About the author

Harry is a ferociously good natured eccentric in Duluth Minnesota who detests traitors dressed up as Uncle Sam. You can find somewhat better edited writings of his at the Duluth Reader "contributors." page. The URL is:

https://duluthreader.com/contributors/h/84/284-harry-welty