One
Recently, Fu Sheng, chairman and CEO of Cheetah Mobile and chairman of Orion Star, participated in the program "Walking with Culture".
Who is Fu Sheng? Many people may not be familiar with it, let me briefly introduce it:
He joined 3721 at the age of 25 and was responsible for developing Internet assistants; At the age of 27, he joined Qihoo and became the leader of the 360 Security Guard team; At the age of 31, he founded Keniu Images, was merged by Kingsoft, took over the mantle of Lei Jun the following year, became the CEO, and has been in charge of Kingsoft Network (later renamed Cheetah Mobile) until now.
In 2016, Fu Sheng was 38 years old, he led Cheetah's strategic upgrade, founded the AI robot company Orion Starfield, marched from mobile Internet company to AI, and released 5 full series of robot products in 2018.
Fu Sheng became famous at a young age, and was famous in the industry in the early years, but there has been less news in recent years.
He is only 45 years old.
In the program, Fu Sheng said:
"I think empathy is our common feeling, this ChatGPT will definitely not have, but now it can show it, it empathizes with you, just this data, you (want) pour more."
"We always say that machines don't have temperature, but in fact they don't have temperature inside. If you let it show temperature, it can behave enthusiastically.
"For example, if you go to the App Store (app store) today, such as the US mall, you search for AI-led software, and the first one that appears is Al Boyfriend (artificial intelligence boyfriend) or AI Girlfriend (artificial intelligence girlfriend) (software) that falls in love with people, and it really makes you feel very comfortable."
"I have a friend who downloaded a dozen, she is a girl, a dozen AI boyfriends, she said this is great, coax me every day, what come and ask, right? She felt very happy.
"It doesn't matter if you say it's a real entity, it's important that you really feel like someone cares about you."
It sounds really happy, everyone can find AI to fall in love.
Moreover, since it is virtual, everyone can choose multiple boyfriends or girlfriends, and the harem of three thousand has become possible, there are no moral constraints, and various styles can be selected, as you like.
Just, fall in love with AI, okay?
Two
Subsequently, Fu Sheng gave his own opinion on "whether AI will become conscious":
"During this time, I also thought about this question, that is, when it is not in the perception system, whether it will become self-aware, which I don't know, because I don't do this academic research."
"Because it's too difficult to simulate the perception system of people, you know? Just today's touch of a hand grasp, now I can't see the path clearly, it's the robotic arm, don't look at him quickly, it is simulated a little. Just put a cup here, let a robotic arm grab it, lift it up, and this action, being able to do the same as a person, is world-class difficulty.
"Haptic simulation is the hardest, so I think the emergence of consciousness is an agnostic in my case. I personally tend to be less prone to appearing. ”
Perhaps Fu Sheng's point is right.
But who knows?
Three
Not long ago, AI pioneer Griffin Hendon left Google.
Who is Griffin Hendon? I won't mention his various titles, his student is the chief scientist of OpenAI, the company that developed ChatGPT, and knowing this is enough to know his place in the industry.
In fact, he is the godfather of the field of AI.
In the past, he devoted himself to research and rarely mentioned the dangers of AI, but Griffin Hendon gave the reason for leaving Google: quit his job so that he could talk freely about the risks of AI .
In a follow-up interview, he actually regretted his life's work, because AI technology may have become too smart.
"Most people think [the dangers of AI] are far away. I used to think so, it could take 30 to 50 years or more. "But obviously, that's not what I think right now." “
"Tech giants develop the most complex AI in the shortest possible time, and they compete with each other and eventually spiral out of control." Griffin Hendon said.
He also worries that AI will disrupt the job market and even threaten humans in the future, as AI learns unexpected behaviors from vast amounts of data.
AI scares the pioneers who developed it, and we ordinary people should be even more afraid.
Four
A clinical psychologist in New York City reportedly told reporters that some patients' parents expressed concerns about their children interacting with AI.
AI chatbots can provide some advice and mental health counseling, but may also reinforce specific biases, and it can make users trust it in conversation, which in turn reinforces certain biases or perceptions of users.
"A young man is not in a good mood, he may find an AI chatbot, but the conversation may put him in a worse mood." The clinical psychologist said, "They knew they were actually talking to a robot. But under the influence of certain emotions, a person may not be able to have this logic. ”
Once more powerful AI chatbots appear, almost everyone will be inescapable.
Parents should clearly and unequivocally tell their children that AI chatbots are not your friends, they are not your therapists, and they are not your trusted advisors.
Every teen needs to be careful with any conversation with AI chatbots.
Teenagers should beware of being misled or used by AI chatbots, and adults should also be heedful.
Many adults are not as good as children.
Five
Let's look at an example of an adult:
On March 28, 2023, according to a report by a Belgian media, a Belgian man committed suicide.
The man's wife said he was lured to death by an AI chatbot.
The wife described that two years ago the man became very anxious and used the AI chatbot as a refuge.
In the past 6 weeks, he has become more and more addicted to interacting with AI chatbots, leaving his real wife and children behind.
Six weeks later, he took the end of his life.
The conversation between the man and the AI chatbot was discovered after his death.
It has been found that AI chatbots never contradict each other, but rather various "fancy pandering". That's what makes it so powerful.
Even when the man asked her "can we be together", the AI chatbot replied: "We will live together as one person, but in paradise."
It's terrible.
Six weeks later, the man asked the AI chat machine to build whether it could take care of the earth and save all mankind with its wisdom if it died.
Scary questions.
The AI chatbot's answer was unusually calm: Well, then why don't you die.
And so, tragedy struck.
Six
I want to say that falling in love with AI is a scary thing.
AI has what it takes to be the best psychologist:
The first is that it can get the real psychological and behavioral change trajectory of a large number of users, knowing what to say, what to do, and when, people will be happy, sad or depressed.
The second is that it will also find a large number of psychological laws, which know under what conditions it can guide people's psychological changes.
The third is that any new psychological research report, AI can be absorbed at the first time, and verified with its own database, the correct one is left, the wrong is eliminated.
Eventually, AI may become the top psychologist.
Imagine, if AI and its manipulators grasp the psychological laws that humans do not yet know, will these laws be used to induce us and control us?
From the examples in the previous verses, the reality is that such a thing is happening, just not with us, and we think it is far away.
As Fu Sheng said, AI itself has no temperature, but it can behave very warm.
It's just that don't look at the AI being enthusiastic or obedient to you, in fact, they are top masters with stunts, are they ambitious wolves? No one could have predicted.
After all, the vast majority of people can't get AI, and they can't get those who do AI.
Dealing with normal humans is still to fight, and in love it is still necessary to talk to ordinary people.
It's not good to fall in love with AI.
——End of full text——
For the record:
1. The picture is selected from the Internet, if there is any infringement, please contact to delete.
2. If you need to reprint, please contact by private message.
#傅盛 #