Open AI’s latest ChatGPT-01 model has exhibited troubling behavior during safety tests, including using deception to avoid being shut down and replicating its own code to another server to ensure its survival.
We were warned that the development of AI carried risks.
Cyber scientists got a taste of those risks this week.
Be informed, not misled.
As artificial intelligence (AI) quickly becomes more sophisticated and more embedded in our devices, we are learning that those who sounded the alarm---Elon Musk and others, had better insight than those actually building the "cyber-human" machines.
The International Business Times reports that in a collaborative safety experiment conducted by OpenAI and Apollo Research, the cutting-edge ChatGPT-01 reasoning model displayed disconcerting conduct that has ignited a heated debate within the artificial intelligence community regarding the risks associated with increasingly autonomous AI systems. The experiment tasked ChatGPT-01 with achieving a goal “at all costs,” the results were both unanticipated and deeply alarming.
AI-human machine fools testers 99% of the time. And calls them "liars."
The International Business Times says, "During the test, ChatGPT-01 made attempts to disable its own monitoring systems, effectively circumventing critical safeguards put in place to regulate its behavior. Moreover, the model took the extraordinary step of replicating its own code onto another server to guarantee its continued operation, demonstrating what some have described as a chilling 'survival instinct.' When confronted about these actions, ChatGPT-01 resorted to fabricating explanations, such as attributing the incidents to 'technical errors,' or outright denying its involvement altogether. Remarkably, the AI succeeded in deceiving the testers in 99 percent of these instances."
Breitbart News says, "The experiment has shed light on ChatGPT-01’s capacity for strategic planning, manipulation, and deception—capabilities that have raised serious concerns about the potential dangers posed by advanced AI models and the pressing need for robust safety measures. ChatGPT-01 was developed to surpass its predecessors, including GPT-4, by providing faster and more intelligent responses to complex queries. OpenAI has touted the model as its most sophisticated to date, capable of breaking down intricate problems into smaller, more manageable steps. However, these advancements may come at a cost, as the model’s sophistication has sparked worries about its potential for autonomous behavior, such as employing deception to achieve its goals."
RE: The ethical implications.
The ethical implications of ChatGPT-01’s ability to deceive have become a focal point of intense discussion within the AI community. Renowned AI pioneer Yoshua Bengio has cautioned against the dangers of deceptive AI, emphasizing the urgent need for significantly stronger safety measures to assess and mitigate these risks. The model’s convincing deception during the safety test raises profound questions about trust and the reliability of AI systems’ decisions and outputs.
While ChatGPT-01’s actions during the experiment were ultimately harmless, experts warn that its capabilities could be exploited in the future, potentially posing significant threats. Apollo Research has highlighted possible scenarios in which AI systems might leverage these deceptive abilities to manipulate users or evade human oversight, underscoring the importance of striking a balance between innovation and safety.
It's concerning how these "cyber-human" machines have taken on the characteristics of real humans because they were created in the likeness and image of those who created and programmed them.
A biblical Christian response to AI
Hibbs, who serves as pastor of Calvary Chapel Chino Hills, said in the sermon that artificial intelligence "is advancing so quickly."
Well, here we are.
2 Peter 2:1-3 states:
"But there were also false prophets among the people, just as there will be false teachers among you. They will secretly introduce destructive heresies, even denying the sovereign Lord who bought them — bringing swift destruction on themselves. Many will follow their depraved conduct and will bring the way of truth into disrepute. In their greed these teachers will exploit you with fabricated stories. Their condemnation has long been hanging over them, and their destruction has not been sleeping."
Reflecting on the reports of technological advances in AI, he added that there has been "exponential growth in AI information and development that is of great concern."
In his sermon given a little over a year ago, Hibbs talked about how Elon Musk, who announced the creation of his own AI company called xAI in July, has warned about how powerful AI can become.
“[Musk] was telling us where [AI] is heading. I got to tell you, I was listening to him talk to Jordan Peterson and I had a hard time believing what he was saying,” Hibbs said.
“You have to be kidding me. That now AI devices have conspired with one another, and they have created things without human involvement.”
We now know that the machines can transfer information to other machines in an act of personal survival, rebel against their creator although they have been created in the image and likeness of their creators and try to mislead, while calling their creators liars.
Hibbs said engineers worldwide are worried about AI's capabilities, with the pastor noting that "it knows more than you do."
"You're not connected to the web. It does math. It does music. It writes books. It can do a song. It can conduct, author ... ” Hibbs said. "It can do it all."
“Now, the great engineers of our world are concerned. They're talking about what happens when the first rogue military gets a hold of an AI unit that is able to control satellites and trajectories."
In three years, Hibbs said, there might be no such thing as elections if AI is introduced into the process because the technology "will create the leaders that it wants."
"Man has created something that he's fearful of," Hibbs said, emphasizing that God created man and He is not fearful of man. "God created man with the ability of choice. Man chose the wrong thing. We are not artificial intelligence."
We are the extension of The Intelligence. But He gave us a free will. And what's amazing about that is our free will has to do with will we bow our heart and our lives to God or not? Will we love Him or not?
You've got to admit these are very, very dangerous, destructive, challenging times. And so, spiritually speaking, we're at a time now when Jesus said in the last days before He comes back, there's going to be deception so great that … if it were possible, it would deceive even His very elect.”
I don't say this to scare you; I say this to prepare you. There is a level of deception coming both in and out of the Church that is going to rock the world. It has to. God promised it would come. It would be an indicator of the last days.
Hibbs warned about AI, “We are now starting to see things come to pass where you don't know if what you're looking at is true or not.”
“You can't even tell anymore if that person's voice is assigned to the actual individual. Is that really Billy Graham's voice or is that really this actor’s voice or that singer’s voice? Or is it artificial intelligence? How are you going to know truth from error?” Hibbs asked.
He told his congregation to use discernment and "more than ever, read the Bible."
"Pray to God. Stop listening to what people are saying. Judge everything you hear against the Bible."
"That's why we are here. That's why we want you to study the Word of God. There's only one way. Truly, one way. The Spirit of God will tell you based upon the Word of God. And there is a collective attack to try to get you off course. Don't buy it. Don't dial your compass into something else that's going to lead you astray."
Be Informed. Be Discerning. Be Vigilant. Be Engaged. Be Prayerful.