Empathetic Robots? Rethinking the Three Laws of Robotics

**SPEAKER_1** (0:04)
The Bliss Business Podcast, the show about empathy, connection, and consciousness in business. Today's episode is brought to you by Zero Company, your paid search, social media, and programmatic ad experts.

**Stephen Sakach** (0:17)
So tullio wants a robot therapist out there, I think.

**Tullio Siragusa** (0:20)
Yes, absolutely. You know, ever since Sum was on the show, and he was like jamming to this beat, I get into it now.

**Stephen Sakach** (0:31)
Hello, and welcome back to the show where we talk about empathy, consciousness, and connection in business. We've got a fun one today. As AI continues to speed rapidly forward, we're going to rethink the three laws of robotics. I'm Stephen Sakach here again with tullio, Siragusa, and Ash Maher. Hey, guys.

**Tullio Siragusa** (0:52)
Are we really real or just robots?

**Stephen Sakach** (0:56)
Ash, Julie and I covered the topic of AI and CEO on last Tuesday's episode. There were some discussion in that show about how AI can be very myopic in how it follows and optimizes toward a goal. Give it a single goal like pursuit of profits and it can really be pretty ruthless in achievement of that goal. That's what we talked about. With this latest generation of AI, now we're seeing these autonomous robots entering production lines. The future is now.
As you may be aware, Isaac Asimov came up with the three laws of robotics. Originally, in his short story of run around in 1942, these laws were created to address anxiety of robots becoming too powerful and turning against us humans. Sounds familiar, right? We're all a little on edge about where we're going with AI, so I'm pretty sure the three laws, as they are, aren't good enough if they're going to be the guiding principle. We're going to kind of cover them and see where some of these gaps are. Ash, can you review what are the three laws of robotics?

**Ash Maher** (2:06)
Absolutely, Steve. So the three laws of robotics are first law, a robot may not injure a human being or through inaction allow a human being to come to harm.

**Tullio Siragusa** (2:20)
You have to go through them like a robot though, Ash.

**Ash Maher** (2:23)
Robot, second law, first law. Robot must obey the orders given by human beings, except where such orders will conflict with the first law. Well, I already imagine there's some flaws in just having these very...

**Tullio Siragusa** (2:40)
The robot better not be programmed by me, it's gonna be out here, and I'm obeying.

**Ash Maher** (2:45)
And the third law, a robot must protect its own existence as long as such protection does not conflict with the first or second law. I think they're describing like a potential future Terminator movie, but you follow these laws. Protect your own existence.
Humans can tell you whatever they want, do whatever they say.

**Tullio Siragusa** (3:04)
Human, bad, robot, good. What if it can't distinguish a human versus a robot? Because it's got like human skin or something, you know.

**Stephen Sakach** (3:14)
Well, now you're getting into Blade Runner. We're still in the iRobot movie.

**Tullio Siragusa** (3:19)
I've advanced our movies.

**Stephen Sakach** (3:21)
These laws have kind of influenced some modern thinking about AI and ethics, but AI has certainly evolved past the original vision here. Right now, like I said, we've got robots and production lines. The simplest ones are vacuum our homes and don't have arms and legs, so we aren't too worried about them. We've got autonomous vehicles though, driving around with no one in them in cities like San Francisco. So we are developing also AI agents to do things in the virtual world where they don't have a body. And that's, you know, I think, really the first sort of battleground we see here. But tullio, what are some of the limitations you think? Let's just kind of dive into some of the limitations of these original free laws of robotics.

**Tullio Siragusa** (4:11)
I mean, they were good, but they're some...

**Stephen Sakach** (4:13)
They're pretty elegant, they're elegant, they're pretty...

**Tullio Siragusa** (4:15)
Pretty elegant, simple, you know, big ideas, always simple. So kudos for doing that. But they're a little limited in terms of protecting individual humans without consideration for like broader social and environmental implications. I.e. it doesn't take into consideration that we feel things and react and change our minds. So the laws are a little silent on issues like well-being of communities, environmental sustainability, and even economic impact. So there's no social elements built into it that could give reason for someone to create an uproar, you know, and someone might be justified for it. So and then also there's challenges of modern AI, such as bias into the algorithms, and not necessarily built on purpose, but it's there.