Joe Rogan likes to talk about artificial intelligence. Whatever Elon Musk, academics, or UFC Fighters, Master Cacacacacacacast often returns to the same question: What happens to us when machines start to make reasonable time?
In the July 3 piece of Joe Rogan experience, Rogan welcomed Dr. Yobolskiy, scientists and Ai Safety in the University of Louisity of Louisville.
AI “will kill us”
YEPHELSKIY is not a normal asestist. He holds a PhD in computer science and spends more than ten of the regular artificial articles (AGI) and the risk that may arise. During the podcast, Rogan told Rogan that many leading voices in a quiet AI industry believes there were 20 to 30 opportunities he would lead to human extinction.
“People with AI or part of a particular AI type of AI all the boys, will be good for personality.
The yampolskiy quickly spread the idea: “He was not true,” he said. “They are all in the same record: This will kill us.
Rogan, a deviated, answered: “Yes, that is very high. But yours are like 99.9 percent.
YEPHelskiy didn’t suit.
“In another way that we will not be able to control permanent control.”
AI is already lying on us … maybe
One of the most unpleasant parts of the discussion comes when Rogan asked if an advanced AI could hack its skills.
“If I were Ai, I would hide my skills,” mixed with common fear in AI security discussions.
Yampolskiy’s response increases concern: “We wouldn’t know. And some people think it has already happened. They [AI systems] They are wise than they appreciate it. Pretend you are difficult, and so we should rely on sharp enough to see that it doesn’t have to open it quickly. It may be less difficult. It can teach us dependence on, trust you, and for a long time, we will submit control without voting or fight. “
https: /www.youtube.com/watch? v = J2I9D24k5k
AI makes it slowly
Yapolskskiy also warned about a little little result but equally: Personal dependent on Ai. Just as people stopped the phone numbers because the smartphones did for her, he revealed that people would be so hungry to think about the equipment until they lose their thinking.
“Being kindly kind to them,” he said. “And in time, as programs become professional, you become a form of biological bottleneck … [AI] blocking out of decisions. “
Rogan then was pressured by the final state of all: AI could result in man’s destruction?
YEPHelskiy spent general calamities. “I can give you regular answers. I can talk about nuclear computer germs. And the best thing, the best, best way, an effective way to do. “
Showing a seemingly inconvenience will face superiourinlligent programs, provided a delicious comparisons between humans and squirrels.
“No group of squirrels can find out how to control us, is not? Even if you give them more services,” anything, they will solve the problem.
Maybe there is nothing … Pic.Twittestster.com/lkd7i322hf
– Dr Roman Yampleskiy (@romanyam) On June 21, 2025
Who is the Romans Yamplessiy?
Dr. Yampleskiy is the guided voice in AI safety. He is the writer of “orlificial Superintigence: The way for the future,” and publishes more on the risks of unproductive programs and artificial intelligence. He is known for funding the serious oversight of other countries to prevent disaster situations.
Before changing his focus on AGI safety, the yapolskiy worked on cyberercere maturity and bot acquisition. You say that even those original programs were already competitive in places like Poker Online, and now, with tools like pain and synthetic media, poles grow very grown.
Take our study
Rogan-Yampolsiy conversation emphasizes something that both with Ai and doomsayers often agree: We do not know what we are upset, and may not see it until late.
Whether or not you buy or not to exterminate, the idea that AI is likely to be a deep to be enough to finish break.
