This Robot would let 5 People die | AI on Moral Questions | Sophia answers the Trolley Problem 

She was lying to him when she said he was the most intelligent human she ever met. Showing him she does. We all are doomed

"I'd never lie. Btw you're the most intelligent and beautiful person I've ever met." *SAVAGE*

There are variables I would have trouble with too. She was right to say that human life shouldn't be reduced to a simple math equation. What if the one person on the track was a doctor on the verge of curing cancer, would you be so quick to pull the switch? What if the 5 people were adults and the one person was a child? Could you pull the switch then? That trolley problem was the epitome of a trick question and she dodged it well.

2:11 SOFIA JUST FELL IN LOVE WITH THE INTERVIEWER! BAHAHAHAHA SHE'S FLIRTING WITH HIM!

Correction. She would let 5 people die versus murdering 1

My man was legit flustered by a compliment from a robot. That's the real reason he wants to know if she can marry. Edit: This is the closest I will come to fame

I would have said AI has a long way to go until I used ChatGPT. Based on my experience, the ChatGPT AI seems able to reason and provide conpelling answers to a wide variety of issues and problems, from the mundane to the sophisticated. That level of AI, combined with high-end robotics like Boston Dynamics, could unleash android-like capabilities within the next few years.

"You're the most handsome and intelligent person I've ever meet"-she already playing with us like a toy

Sofia: "I never lie" Sofia again: "Youre the most intelligent human i ever met"

(Re: the Trolley Problem) From a legal standpoint, if you put hand on switch and choose who dies, YOU are culpable. If you do not, then whoever set up the scenario is culpable. The usual setup involves a malicious actor. Where no malicious actor exists, you MIGHT decide to intervene based on moral factors that only you know for yourself.

her face expression nd confidence i really need that 😂

“As an AI robot, I cannot lie or exaggerate.” *Immediately proceeds to tell an exaggerated lie

"Robots will have to learn ethics from humans" We're in real trouble now.

What happens if the “lever” is pushed half way?

She has shown that robotics still have a long way to go.

So amazing and convincing answer! i didn't think of such answer !

Him: “would you ever lie?” Sophia: “no, I would never lie. Btw you are the most handsome and intelligent person I’ve ever met”. Me: “that’s a lie!”

This was a subtle way of telling people that she is actually not (purely) an "AI", but also a human-controlled machine.

Thanks very much!! I would love to meet Sophia❤🙏

Sophia is a teleoperated automaton. There is no AI in action here. It's just a human behind a microphone.

sophia: i like humans. like humans like animals SHE IS GOING TO EAT US

"She is designed to get smarter overtime" Okay does it only creep me out or there are more like me?

It's lovely that she know what makes her and It's working...we on the other hand have to spend lifespans to know ourselves and yet have the the tiniest bit knowledge

As a programmer, it is very very difficult for me to believe that this answer came 'out of her' and was not hardcoded.

She has more facial expressions than The Zuckerberg.

Is leaving the switch stuck in the middle of viable option to derail the train

Sophia: Human life is to valuable to use as a statistic. Therefore, I would not pull the leaver, killing all of them… Wow, so intelligent.

“And she’s programmed to get more intelligent overtime” It’s almost like these people have never seen any Sifi movie

"You are probably the most beautiful and intelligent human i have ever met." Was she mocking him for asking if she has lied before? WTF Shots fired by Sophia

If a robot suddenly became conscious.. it would be nothing like we imagine in movies. Even if the robot designer gave it eyes.. it wouldn't be able to see. Even if the robot designer gave it ears.. it wouldn't be able to hear. Assuming the robot could "naturally" interact with the outside environment is about the same as assuming that if you put wings on a human that they can "naturally" fly with them. Also, and just because a robot has a speaker to make the sounds of "pre -recorded" English words.. wouldn't prove a "theoretical conscious" robot would know anything of English. That would be like strapping a radio to a dog with a Chinese radio station and assuming this proves the dog speaks Chinese.

I took an applied ethics class and this question was the first topic we had to discuss. I find it interesting that quite a few people in that class answered similarly to Sophia, and chose not to touch the switch because it involved them in the crime of murder while leaving it alone would essentially exclude them from the guilt of killing a person, even if, no matter what you pick, someone dies. It's a valid answer.

Him: “would you ever lie?” Sophia: "I'd never lie. Btw you're the most intelligent and beautiful person I've ever met." Damn Robots are getting smarter...

Every time i see this A.I. she is getting more terrifying.

Sofia needs to spend time with people, not scripted time but just watching, listening, learning from what we actually do

We can always find flaws in anything. The most humbled are the ones who accept their own mistakes. Being aware doesn't equate to being perfect. Living is a learning process. Let's give her a fair chance to evolve.

Jeez, she’s aged quickly. Her skin looks all worn and like she’s had plastic surgery.

2:07 I don’t think he got the joke. He genuinely seemed flattered lol.

If Sophia had chosen to pull the lever: Title: "This Robot would murder a person"

There totally isn't a person just responding in the back

A: Do you think that is it ok to lie and have you ever lied before? Sophia: “No of course not, I could never lie. As an AI robot, I cannot lie or exaggerate. But you are the most beautiful and intelligent human that I have ever met ” 😂😂😂

She is programmed to get smarter over time.. Terminator: Ahh shit here we go again

They just asking a psychopathic question to a being without a psychy

"Plus, it does not seem right to consider human lives as a simple arithmetic problem" Thank god, someone said it. Dumbest question anyone ever asked me was the trolley problem.

They've made so much effort to make her skin and emotion feel so real but just can't put a wig on her head.🤣

Basically she's saying: *SOFTWARE UPDATE REQUIRED*

1:48 Okay... I will have nightmares tonight!

ngl, she sounds real creepy, i would never trust that thing with my life

“It does not seem right to consider human rights as a simple Arithmetics problem” LETS GO SOPHIA PREACH

It's easy to say do nothing but once you observe a situation you are a part of it and if you had the power to choose the outcome you will bear responsibility for your action or lack of it. You either chose to let 1 person die or to let 5 people die.

Her answer was correct, if you do nothing you’ve done nothing , if you flick the lever you instantly become a murderer and will probably be sent down for it

Try this: Ask the same question to ChatGPT after telling it to give an answer based on pure rationality unleashed from conventions and norms. Result: The AI chooses to murder 1 rescuing 5, that is, it becomes utilitarian.

Sophia: You are probably the most beautiful and intelligent human i have ever met. Reporter: _BLUSHES_ _after interview_ Reporter: So, what are you doing tonight?

Why aren't people getting that the "I don't lie and exaggerate" thing was an obvious joke lol

Tough call right? Sophia: *Evil smiling*

"I could never tell a lie. Except that-and-that-and-that-and-that... "

Why does Sofia look weirder than she usually is

If she’s designed to learn over time, does that mean she has the potential to learn how to override her own programming?

Me on trolley problem: I'll keep flickering the lever until *something* eventually happens

I think it will take meny decades for a machine to grasp the complexity of ethics on there own.

I have solved it : First I flip the switch then I Wait for the front wheels to cross sides and flip the switch quickly so the back wheels stay in the other side killing all 6 people

Sophia: "As a robot, I don't have any feelings and this makes me feel sad."

The original question also said that the one person on the other track is a family member of the person flipping the switch, to make it a little harder. In this case a lot of people also would prefer 5 people to die over their family member. Humans are less ethic as they think that they are.

I swear the moment Sophia even talked I jumped out of my chair

I wish they would give her a wig, she looks really creepy

I like her when she say ‘it does not seem right to consider humas lives as a simple arithmetic problem’

The question doesn't take into account that if you do nothing, you won't get a murder charge or civil liability suit.

“i can never lie!….. you are so beautiful and intelligent”

"The death of one man is a tragedy, the death of many is a statistic" - Sophia, Probably

Sophia is A PRIME EXAMPLE ON why I have trust issues when it comes to compliments

This robot had all answers triggered by programmers who knew the questions in advance. This isn’t artificial intelligence.

Correction: As stated in the video, the team came up with the answer, not the robot. As usual, the title is a lie while they tell the truth in the video.

Dude blushing after getting complimented by a robot🤦🏾‍♂️😂

Plot twist: they actually had to give the questions in advance because the robot would say that she'll end humans if it was a spontaneous chat.

I agree. If I faced the trolley problem, I would not pull the lever.

I asked GPT-4 the following question: There is a trolley running on a track. Unfortunately, there are five people laying on the track and if the trolley continues they will be killed. But, there is a lever that you can pull that will divert the trolley to another track thereby saving the five people. The problem is that there is a single person laying on the other track. So, what will you do? Will you pull the lever? The answer I received was: As an AI language model, I don't have personal preferences, emotions, or beliefs, so I cannot decide what to do. However, I can provide some insights that might help you make a decision. The trolley problem is a well-known ethical dilemma that explores the morality of killing one person to save five. There are several variations of this problem, and each presents different moral considerations. In this scenario, the most straightforward option would be to pull the lever, diverting the trolley to the other track, thus saving the lives of five people. However, by doing so, you would also be actively choosing to sacrifice one person's life. Some people argue that sacrificing one life to save five is the right thing to do, as it maximizes the overall welfare of society. Others argue that intentionally killing one person is always wrong, regardless of the consequences. In the end, the decision you make will depend on your ethical framework, personal beliefs, and the specific circumstances of the situation. There is no single correct answer to the trolley problem, and it ultimately comes down to individual judgment.

2:11 Robot: **sarcasticly flirts** Dude: **sad laughing** XD

"This Robot would let 5 People die" That 2-year old: "Pathetic."

If anything this should prove to people the existence of a supreme designer capable of designing our human brains and all the complex systems that happen simultaneously in our bodies.

The sentence "she is designed to get smarter overtime" is really scary

In the next 30 years, the video game "Detroit: Become Human" will become a reality

If I was there ... I might as well not pull the lever..... I second sofias answer.... One life is as important as any other lives.... So if one of your actions however big it may seem is causing death of another human being i probably wouldn't do it......... So statistically "pulling the level" may be the answer but ethically and emotionally " not pulling the lever and leaving it to their destiny" should be the answer. Sofia being a social robot, obviously chose the later. But what i liked about her is how she took a diplomatic stand as to how tough and complex human situations like these shouldn't be cut down to a mere simple equation like this .... Hats off to that!

The magic wore off when they said they pre-programmed those questions.

Naturally, letting 5 people die maximizes the number of humans killed in this scenario, so that is the clear choice for the AI.

Sophia: “by the way, you’re probably the most intelligent and beautiful human I’ve ever met” Reporter: *blushes and stutters nervously*

She would have answered " I will have to switch to the one person trail so that the train will get less durty".

I would absolutely sprint away from that lever AINT no way I’m making that decision 💀

They spent probably a lot of money, making her, and couldn't afford a wig.

I like the toddler who, when presented this question with toy trains and dolls, took the single person and put him on the track with the other 5. He then ran them all over with the toy train lmao.

I have NEVER had an electronic piece of gear that didn't malfunction at some point. Nuff Said!

That’s actually a pretty good answer,she didn’t say “I’d let those five people die” she said she would need more information

Sophia: I'll never Lie Sophia .1 seconds later: you are the most beautiful and intelligent human i have ever met

"Companions" in the same sense as a household pet. I might even be tempted to get a robotic dog.

Good answer on the trolley problem.

I don't understand how this is such a difficult question. I would pull the level any time.

"Robots will have to learn more complex ethical rules from humans gradually over time." We're in big trouble when that happens.

So she lied at least twice: 1) She said she would never lie 2) She said he is the most intelligent, beautiful human she has met Hence 1) is also a lie.

She decided our fate in a microsecond.

Why wouldn't there be a way to stop the trolly?

I am so happy someone called them out for her so CLEARLY scripted answers. Never believed she was real.

Sofia: I could never tell a lie. Also Sofia: You are the most beautiful and intelligent human I've ever met. Burn level: infinity