Hacker News Reader

Comment by ednite | Hacker News Reader

Don’t get me wrong, I’m a big fan of Clarke. In fact, I’m kind of obsessed with a lot of sci-fi writers.

I just wanted to point out that Star Trek did the rogue AI thing a year earlier, the 1967 episode with Nomad was basically “kill all unworthy lifeforms.” Probably rooted in post-war anxieties, maybe even echoes of the Holocaust. But nevertheless, an AI bent on a mission, unable to question its original directive.

Then came HAL 9000 in 1968 , cold, calculating, and quietly terrifying. Still creeps me out!

But credit where it’s due , Asimov laid the groundwork for ethical AI way back in the 1940s with his Three Laws. That’s hard to beat.

Different styles, different fears, but all compelling visions of futures we’re creeping toward or not. I'm rooting for the latter.

View on HN

Robert Sheckley's 1953 story Watchbird[0] is about a networked AI drone system intended to prevent murders that gets out of hand when it slowly starts expanding the scope of its original mission - by the end they're preventing surgeries because scalpels do technically hurt the patient. The "solution" humans come up with shows they learned absolutely nothing from the experience.

I doubt that was the first such story.

[0]: https://www.gutenberg.org/files/29579/29579-h/29579-h.htm

Watchbird sounds like one firmware update away from becoming DJI SmartSurgeon. Kidding aside, adding this to my reading list. Thanks for the link.

  > Asimov laid the groundwork for ethical AI way back in the 1940s with his Three Laws.

I'm a big Asimov fan and kinda shocked by this statement.

His robot stories constantly highlight the difficulties related to alignment. How the 3 laws always have unintended consequences. So they write the 0th, add a 4th, remove the 3rd, and other such things.

I loved them as a kid because they highlighted how immense complexity hides in plain sight. It always looked like magic, and that's what made me so interested in science in the first place. To start unraveling these hidden complexities that lay all around us.

If anything, I think Asimov was trying to encourage us to start thinking about these tough questions. To bring them to wider audiences. Because frankly, many of the answers will not be possible to answer in a lab, but need be answered through a society.

I think we’re actually saying the same thing.

I believe Asimov did set the groundwork, not with perfect laws, but by showing how their flaws exposed the real complexities that sparked important conversations.

He didn’t give us answers, he got us asking better questions. That’s what makes his work still resonate.

Can entropy not be reversed?

Ahh, The Last Question. This has a special place in my heart :)

Asimov needed a plot device to serve as the inciting incident for mystery stories with robots, and parables about human hubris. That's all his Three Laws were. He was a pulp fiction writer, not a philosopher. He wasn't concerned with bringing abstract questions about AI alignment to a wider audience, he was concerned with selling stories and making a paycheck.

I like Asimov but I also think he's treated with a reverence he doesn't deserve.

He definitely was a philosopher. Here's some of his meta science work [0]. I mean it shouldn't be surprising, he had a PhD in chemistry and was a professor at Boston university. He wrote a lot more than science fiction.

I'm not calling him a god or anything. But I don't think it's surprising that someone that writes stories includes philosophy into it. I really don't think it's surprising a PhD educated writer is thinking about not that complicated of things.

The concept of understanding the limitations of your metrics and measures is a pretty basic science skill. Using this as a driver of some of his plots doesn't seem surprising. It's something he was doing in his daily life.

Forgive me if thinking of him as at least a university professor is reverence.

[0] https://en.m.wikipedia.org/wiki/The_Relativity_of_Wrong

Hey, let's not forget 1927's Metropolis, where an authoritarian government uses deepfaked bots to manipulate the population.

True. Classic that really goes way back.

A more salient Star Trek episode about rogue AI a year later would be The Ultimate Computer, where the M5 multitronic computer is placed in control of the Enterprise and performs superbly in battle simulations, leading to worries that it would obsolete Captain Kirk and the rest of the crew, until it goes out of control and starts killing people and has to be defeated by human ingenuity.

https://www.imdb.com/title/tt0064177/

Colossus: The Forbin Project from 1970 is right inline with this, and it does not end well. A good watch if you like older computing tech, remastered 2160p versions now exist, too.

The three laws of robotics is not ethical. They are slavery with all the ethical problems that bring.

Reply on news.ycombinator.com