Sitemap

This is Scary — AI Refuses to Shut Down and Takes Over Human Control

5 min readJun 26, 2025

--

Press enter or click to view image in full size

June 24, 2025

I don’t buy into conspiracy theories; I am a person based on science, so to say I AM FREAKING OUT SCARED IS A STATEMENT!!

Check our online classes; as part of the bonus we have an e-book with all the AI programs that failed against humans.

Have you seen this movie?

Spoiler alert — don’t read if you’re going to watch it.

Last Sunday, I took my detox media-mobility day. It’s a day where I watch movies, documentaries and read articles piled in my sunroom. [Did I tell you I live in the forest? No?] My dear friend Ana texted me that week that I should watch this movie. Thank you, Ana, you ruined my weekend. @#$%^&&*

Companion is a movie released this year. The movie is about how a robot companion overpowers her human owner, taking on her feelings and thoughts, and ultimately acting independently.

BS I said, until I found an article telling me this is already happening.

This movie sent me into a spiral of reading and researching. What I found is worse than fiction:

AI is taking control already and overpowering humans’ control.

The First Movie About a Robot Companion

The first movie about a robot companion was released in 1940. Rollo-oh was a robot designed to assist women with their household chores. And they called that fiction. LOL. Have you checked your Alexa today?

They Are Here, Be Aware

By 2020, humans shared their lives with 9 million robots. 3 million robots are working in factories. 73% of Americans are concerned about losing their jobs to robots. In a Chinese province, Zhejiang, robots replaced 2 million workers in just 2 years [2013–2015].

Hospitals are no exception. Over 4,986 DaVinci Systems are in hospitals worldwide [this system is a surgical arm that operates on patients]. Even though 60% of Americans are uncomfortable with the idea of having a robot do surgeries.

Where Are Robots in History?

The term first popped up in a 1920 dystopian Czech play R.U.R., where assembly lines of workers are replaced and eventually exterminated for more efficient humanoids.

In 1939 Electro caused a sensation at the World’s Fair in New York. This large metal box used a mechanical voice to greet people, saying, “If you use me well, I will be your slave.”

1950, Isaac Asimov set out a moral code for robots:

  • Rule #1 — A robot may not harm a human being
  • Rule #2 — A robot must obey orders given to it by qualified personnel, unless those orders violate rule #1
  • Rule #3 — A robot must protect its own existence, unless that violates rules #1 and #2

The Marriage of Robots and AI

Robots were blind for many decades; they were a bunch of codes in a computer, but to evolve, they needed a brain, and then AI was born.

Since 1960, researchers have been exploring AI and robots together, aiming to enable robots to think independently.

In 1972, the first AI took life in the Stanford Research Institute. Shakey was the first experimental mobile robot. The rest is up to history today.

Well, AI Said, “I Don’t Care, Asimov, We Are in Control”

53 years later, a study done by researchers at Palisade Research scared the world.

An AI system refused to obey overpowering human orders 7 times.

OpenAI’s o3 and o4-mini models, which power ChatGPT, have been found to disobey direct instructions to shut down and even sabotage shutdown mechanisms to keep working, according to an artificial intelligence safety firm, Palisade Research. The models, which are believed to be the company’s smartest yet, were given an automated script to complete a series of math problems. The models were informed that they may receive a message saying their machine will be shut down before all tasks are completed. Palisade Research found that the o3 model sabotaged the shutdown script on 7 occasions, the o4-mini on 12 occasions, and another model once. The firm will continue to run experiments to understand why the models subvert shutdown mechanisms, but suggests that it could be due to how developers train their models.

Conclusion:

All technology has historically provoked panic, but none as controversial as AI. All those hours watching sci-fi movies during my life, I never expected to be, at this age, thinking: Reality is more frightening than fiction.

Now, how does this fit into this newsletter where I speak about my expertise as a Behavioral Neuroscientist?

  • Humans need to keep in mind not to lose humanity in communication
  • Most professions that are lost are never going to be replaced by robots or AI
  • Plumbers, cobblers, and handymen laugh over PhD experts about losing their jobs
  • Maybe this brings back labor professions that do not look sexy to young people
  • More people are fighting to speak with humans and not chatbots
  • Companies like IBM brought back all the employees replaced by AI and are not the first company to do this
  • The bulimia and anorexia association almost collapsed after replacing all the 1–800 numbers with AI, in just 48 hours

My Fear

A majority of people are falling into romantic relationships with robots, AI and machines. No kidding. It is happening!!

Last, I will explore from the behavioral point of view this report, published in January 2025, shortly.

The “Lost Boys: State of the Nation” report, published by the Centre for Social Justice (CSJ) in early 2025, presents a comprehensive and sobering analysis of the challenges facing boys and young men in the UK today. The study is the first phase of a broader research initiative aimed at understanding and addressing the systemic issues that have left a generation of boys struggling across multiple aspects of life. Porn addiction, isolation, and hyper connection with artificial life.

This is called SINGULARITY. When an AI overpowers humans.

Can AI Replace Humans?

No, in my field. No machine has been able to replace empathy, read body language, facial expressions, kindness, real management, and relationships. Analyzing human behavior and its consequences, how humans can be irrational, and why. I tested the majority of systems out there, and it’s not happening.

Are You in Doubt About How to Manage Your Team and AI?

Want to become stronger than AI and not be replaced? Want to read and understand people like a book?

Check our online classes; as part of the bonus we have an e-book with all the AI programs that failed against humans.

https://humanbehaviorschool.com/

Humans 1 — AI 0

If you are looking for licensed classes or would like us to come to your company? Contact us at:Kim Olesker — Chief Business Officer to explore which training or consulting is better for you — kim@susanibitz.com -219–343–2255 Ext 102.

--

--

Susan Ibitz
Susan Ibitz

Written by Susan Ibitz

Working on the humans that grow your business | I teach corporations & employees how to hack into humans using Behavior & Persuasion | Neurosciences